Luminescent Thermometry: Applications and Uses 3031285158, 9783031285158

Table of contents :
Contents
Introduction to Luminescence Thermometry
1 What is Luminescence Thermometry?
2 Luminescence
3 Methods Used in Luminescence Thermometry
3.1 Time-Integrated Methods
3.2 Time-Resolved Methods
3.3 Other Luminescence Thermometry Methods
4 Analysis of the Performance of Luminescent Thermometers
4.1 Thermal Sensitivity
4.2 Temperature Resolution
4.3 Spatial and Temporal Resolutions
4.4 Repeatability and Reproducibility
4.5 Temperature Operation Range of a Luminescent Thermometer
5 Calibration of a Luminescent Thermometer
6 Materials in Luminescence Thermometry
6.1 Lanthanide-Doped Luminescent Thermometers
6.2 Metal–Organic Frameworks (MOFS)
6.3 Quantum Dots as Luminescent Thermometers
6.4 Organic Dyes
6.5 Luminescent Thermometers Based on Polymers
6.6 Transition Metal-Doped Nanoparticles
6.7 Metal Nanoparticles
6.8 Carbon-Based Materials: Nanodiamonds and Carbon Dots
6.9 Biomaterials
7 Conclusions
References
New Strategies to Improve Thermal Sensitivity and Temperature Resolution in Lanthanide-Doped Luminescent Thermometers
1 Introduction
2 Multilevel Thermal Coupling
3 Excited State Absorption Based Luminescent Thermometers
4 Sensitization of Ln3+ by the Transition Metal Ions
5 Negative Expansion Coefficient
6 Structural First Order Phase Transition Derived Luminescent Thermometers
7 Conclusions and Perspectives
References
An Overview of Luminescent Primary Thermometers
1 Introduction
2 Non-luminescent Primary Thermometers
2.1 Optical
2.2 Acoustic Gas Thermometers
2.3 Electric Thermometers
3 Performance of Luminescent Thermometers
4 Luminescent Primary Thermometers
4.1 Thermometric Parameters Based on Emission Features
4.2 Thermometric Parameters Based on Excitation Features
4.3 Thermometric Parameters Based on Excitation and Emission Features
5 Conclusions and Perspectives
References
Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles
1 The Fundamental Principle of Lanthanide Ions Doped Luminescent Thermometers
2 Bright Luminescence Intensity of Lanthanide Ions Doped Nanoparticles Induced by Heavily Doping
2.1 Core–Shell Structure
2.2 High Power Excitation
2.3 High Temperature Treatment
2.4 Effects of Heavily Doping on the Radiative Transition Rates
3 Lanthanide Ions Heavily Doped Nanothermometers
3.1 Thermally-Coupled-Levels Nanothermometers
3.2 Non-thermally-Coupled Nanothermometers
3.3 Lifetime-Based Nanothermometers
4 Effects of Heavily Doping on the Lanthanide Ions-Based Nanothermometry
4.1 Effects of Heavily Doping on the Thermal Sensitivities
4.2 Laser Heating in Lanthanide Ions Heavily Doped Nanoparticles
4.3 Self-Monitored Photothermal Therapy Using Lanthanide Ions Heavily Doped Nanoparticles
References
Metal–Organic Frameworks for Luminescence Thermometry
1 Introduction
1.1 Metal–Organic Frameworks: History, Synthesis and Applications
1.2 Ratiometric Thermometry in Metal–Organic Frameworks
2 Various Strategies to Develop MOFs Luminescent Thermometers
2.1 Eu-Tb Mixed Metal–Organic Frameworks
2.2 Nd-Yb Mixed Metal–Organic Frameworks
2.3 Eu3+-Based Metal–Organic Frameworks
2.4 Others Metals
2.5 Host–Guest Molecule as Second Emitter
3 Conclusion and Perspectives
References
Luminescent Nanothermometers Operating Within Biological Windows
1 Biological Windows
2 Luminescent Nanothermometers Within Biological Windows
3 Performance of Luminescent Nanothermometers
4 Thermometers Operating in the I-BW
4.1 Nanodiamonds
4.2 Fluorescent Gold Nanoclusters
4.3 Quantum Dots
4.4 Carbon Quantum Dots
4.5 Transition Metal Doped Materials
4.6 Lanthanide Doped Materials
4.7 Mixed Luminescent Materials
5 Thermometers Operating in the Second Biological Window
5.1 Quantum Dots
5.2 Transition Metal Doped Materials
5.3 Lanthanide Doped Materials
6 Thermometers Operating in the Third Biological Window
6.1 Er3+
6.2 Ho3+, Tm3+
7 Applications of Thermometers Operating Within the Biological Windows
8 Conclusions
References
Luminescence Thermometry for in vivo Applications
1 Introduction
2 Semitransparent Organisms
3 Insect Models
4 Murine Models
4.1 Control of Photothermal Therapy
4.2 Control of Magnetic Hyperthermia
4.3 Study and Diagnosis of Tissues
5 Conclusion
References
Luminescence Lifetime Nanothermometry for Accurate Temperature Measurements In Vivo
1 Temperature Detection
2 Luminescence Lifetime Thermometry
2.1 The Advantages of Luminescence Lifetime Thermometry
2.2 Performance Evaluation Index of Luminescence Lifetime Thermometers
2.3 Luminescence Lifetime Imaging
3 Luminescence Lifetime Thermometry Applied In Vivo
3.1 Luminescence Lifetime Thermometry Using Carbon Dots
3.2 Luminescence Lifetime Thermometry Using Long-Lived Luminescent Nanocapsule
3.3 Luminescence Lifetime Thermometry Using Rare-Earth-Doped Nanoparticles
4 The Development Prospect of Luminescence Lifetime Thermometry
References
Contactless Luminescence Nanothermometry in the Brain
1 Introduction
2 Brain Temperature
2.1 Brain Temperature Distribution and Fluctuations
2.2 Thermal Modulation of Neural Activity
2.3 Techniques to Study Brain Temperature
3 NIR-II Luminescence Thermometry of the Brain
3.1 Light-Tissue Interaction
3.2 Ag2S Nanoparticles in Brain Thermometry
4 Limitations and Challenges
4.1 Measurement Type and Accuracy
4.2 Delivering Nanoparticles to the Brain
5 Conclusion and Future Perspectives
References
Optical Trapping of Luminescent Nanothermometers
1 Introduction
2 Fundamentals of Optical Trapping
3 How the Field Started
4 Optimization of the Optical Trapping of Luminescent Nanothermometers
5 Desing of the Experimental Setup
5.1 The Optical Trap
5.2 Imaging and Nanoparticle Detection
6 Luminescence Thermometry with Optically Trapped Dielectric Nanoparticles
6.1 Optical Trapping of Luminescent Nanothermometers
6.2 Spinning of Dielectric Nanoparticles for Temperature Sensing
7 Thermometry with Optically Trapped Metal Nanoparticles
7.1 Optical Trapping of Gold Nanoparticles
7.2 Spinning of Metal Nanoparticles for Temperature Sensing
7.3 Dark Field Spectroscopy for Temperature Sensing
7.4 Anti-Stokes Spectroscopy for Temperature Sensing
8 Perspectives
References
Critical Analysis of the Recent Advances, Applications and Uses on Luminescence Thermometry
1 Introduction
2 Novel Approaches in Luminescence Thermometry
2.1 Luminescent Primary Thermometers
2.2 Multiparametric Approach
2.3 Luminescence Nanothermometry Below the Diffraction Limit
3 Engineering Novel Materials
3.1 Heavily Doped Lanthanide Nanoparticles
3.2 Metal–Organic Frameworks for Luminescence Thermometry
3.3 Persistent Luminescence Materials
4 New Strategies to Improve Thermal Sensitivity and Temperature Resolution in Luminescent Thermometers
4.1 Optical Trapping of Luminescent Nanothermometers
4.2 Influence of the Pumping Regime on the Temperature Resolution for Luminescence Thermometry
4.3 Impact of Noise and Background on Temperature Uncertainty in Luminescence Thermometry
4.4 Novel Approaches to Achieve Better Thermal Sensitivities
5 Novel Applications
5.1 Biological Applications for Luminescence Thermometry
5.2 Contactless Luminescence Nanothermometry in the Brain
5.3 Benefits of Lifetime-Based Luminescence Thermometry for Biological Applications
5.4 Luminescence Thermometry for the Internet of Things
5.5 Self-assessed Photothermal Therapy
5.6 Luminescence Thermometry in Engineering Applications
6 Final Reflections: Current Limitations of Luminescence Thermometry and Future Challenges
6.1 Current Limitations
6.2 Future Challenges
References

Citation preview

Joan Josep Carvajal Martí Maria Cinta Pujol Baiges   Editors

Luminescent Thermometry Applications and Uses

Luminescent Thermometry

Joan Josep Carvajal Martí · Maria Cinta Pujol Baiges Editors

Luminescent Thermometry Applications and Uses

Editors Joan Josep Carvajal Martí Departament de Química Física i Inorgànica Universitat Rovira i Virgili Tarragona, Spain

Maria Cinta Pujol Baiges Departament de Química Física i Inorgànica Universitat Rovira i Virgili Tarragona, Spain

ISBN 978-3-031-28515-8 ISBN 978-3-031-28516-5 (eBook) https://doi.org/10.1007/978-3-031-28516-5 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Contents

Introduction to Luminescence Thermometry . . . . . . . . . . . . . . . . . . . . . . . . . J. J. Carvajal and M. C. Pujol New Strategies to Improve Thermal Sensitivity and Temperature Resolution in Lanthanide-Doped Luminescent Thermometers . . . . . . . . . L. Marciniak, W. M. Piotrowski, M. Szymczak, M. Pieprz, and K. Trejgis

1

69

An Overview of Luminescent Primary Thermometers . . . . . . . . . . . . . . . . 105 Joana C. Martins, Carlos D. S. Brites, Albano N. Carneiro Neto, Rute A. S. Ferreira, and Luís D. Carlos Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Lu Liu and Jianzhong Zhang Metal–Organic Frameworks for Luminescence Thermometry . . . . . . . . . 193 Thibault Amiaud and Hélène Serier-Brault Luminescent Nanothermometers Operating Within Biological Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Albenc Nexha, Maria Cinta Pujol Baiges, and Joan Josep Carvajal Martí Luminescence Thermometry for in vivo Applications . . . . . . . . . . . . . . . . . 269 Erving Ximendes Luminescence Lifetime Nanothermometry for Accurate Temperature Measurements In Vivo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Lijun Wu and Guanying Chen Contactless Luminescence Nanothermometry in the Brain . . . . . . . . . . . . 299 Blanca del Rosal

v

vi

Contents

Optical Trapping of Luminescent Nanothermometers . . . . . . . . . . . . . . . . . 315 Lucía Labrador-Páez and Patricia Haro-González Critical Analysis of the Recent Advances, Applications and Uses on Luminescence Thermometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Maria Cinta Pujol Baiges and Joan Josep Carvajal Martí

Introduction to Luminescence Thermometry J. J. Carvajal and M. C. Pujol

Abstract In this chapter, the fundamentals of luminescence thermometry are reviewed, concerning the different mechanisms on which luminescence thermometry relies, and the main performance parameters that should be given for a particular luminescent thermometer so that it can be compared with others. Finally, the different families of materials on which luminescent thermometers have been developed are reviewed, giving key examples of how each class of materials can be adapted to work as temperature sensors. Keywords Luminescence · Luminescence thermometry · Temperature sensors · Methods · Performance · Figures of merit · Materials

1 What is Luminescence Thermometry? The finality of luminescence thermometry is to measure temperature. So, what is temperature? As a simple definition, the temperature is the measure of the hotness or coldness of a body expressed in any temperature scale (Fahrenheit, Celsius, Kelvins) [12]. In physics, the temperature is a physical quantity that measures the average kinetic energy of the atoms or molecules in the system [8]. So, in other words, the temperature is the external manifestation of the thermal energy existing in any system. The lowest theoretical temperature is the absolute zero. At this temperature, labelled as 0 K, there is no thermal energy in the matter. Besides, by measuring the temperature difference between two media, one can know in which direction the heat (energy) will spontaneously flow, always from the area at which a higher temperature is detected to the area in which a lower temperature is present [8]. From these definitions, one can deduce the importance of temperature in several fields J. J. Carvajal (B) · M. C. Pujol Departament Química Física i Inorgànica, Universitat Rovira i Virgili, Campus Sescelades, Marcel·lí Domingo, 1, 43007 Tarragona, Spain e-mail: [email protected] M. C. Pujol e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_1

1

2

J. J. Carvajal and M. C. Pujol

such as biology, chemistry, metallurgy, climate… a long list that does not need to be discussed here. As mentioned above, thermometry has the aim of measuring temperature. As it is known, using the terminology of the metrology field, a measurement is the determination of the magnitude of something [48]. This measurement needs a comparison of the unknown quantity with some standard quantity of equal nature, known as measurement unit. Then, a physical measurement is defined by the act of obtaining quantitative information about this physical magnitude in a body by the comparison with the standard [48]. When defining a physical measurement, it is important to identify and understand the principle of measurement: which is the physical phenomenon in which the measurement is based. In luminescence thermometry the principle of measurement will be the change of any parameter of the luminescence generated by a material affected by the temperature at which this material is exposed. On the one hand, it is worth to define here the meaning of the error of the measurement, which will be related to the accuracy and reproducibility of the instrument used for measuring. However, the error is related to the measurement of the magnitude itself, and not to the instrument used for measuring it. In mathematics, the error of a measurement can be defined as the difference between the observed value of a variable and the true, but unobserved, value of that variable. Furthermore, when discussing about the instrument used for the measurement (here in this book the luminescent thermometer) it is important to determine, as in all measurement instruments: (i) the resolution (related to the accuracy); (ii) the uncertainty; and (iii) the operational range. In the NIST handbook [NIST/SEMATECH e-Handbook of Statistical Methods, http://www.itl.nist.gov/div898/handbook], we can find a list of helpful definitions when dealing with measurements; here some of them, which terms will be used repeatedly along the book: • Resolution is the ability of the measurement system to detect and faithfully indicate small changes in the characteristic of the measurement result. The resolution of the instrument is δ if there is an equal probability that the indicated value of any artefact, which differs from a reference standard by less than δ, will be the same as the indicated value of the reference. • Uncertainty is a measure of the ‘goodness’ of a result. Without such a measure, it is impossible to judge the fitness of the value as a basis for making decisions. • Operation range (also measurement range, measuring interval, working interval) Set of values of quantities of the same kind that can be measured by a given measuring instrument or measuring system with specified instrumental uncertainty, under defined conditions. The measurement of the temperature in our case is done through the instrument called thermometer. Thermometers are calibrated in various temperature scales. The most known scales are the Celsius scale, also called in the past centigrade scale, in which the unit of temperature is the centigrade degree (°C), and based on defining the 0 °C for the freezing point of water and 100 °C for the boiling point of water and dividing this interval between the defined points in 100-degree steps. From another

Introduction to Luminescence Thermometry

3

side, one can also encounter the Fahrenheit scale, in which the unit of temperature is the Fahrenheit degree (°F), and based, originally, on defining the 0 °F for the freezing temperature of a brine solution composed by a mixture of water, ice and ammonium chloride that forms a eutectic system which stabilizes its temperature automatically to a single value. The other limit established is the estimate of the average human body temperature at 96 °F. During the twentieth century, the Fahrenheit scale was defined by two fixed points with a 180 °F separation. The first one, as in the case of the Celsius scale, is the temperature at which water freezes at sea level under standard atmospheric pressure, and located at 32 °F, and the second one corresponds to the boiling point of water, defined to be 212 °F, with a separation between these two points of 180 °F. Finally, it is also important to mention here the Kelvin scale of temperatures, also designated as the absolute thermodynamic temperature scale, in which the unit of temperature is the kelvin (K), that is by convention the international system unit for temperature. One kelvin corresponds to the change in the thermodynamic temperature that results in a change of thermal energy kT by 1.380649 × 10–23 J, corresponding to the value of the Boltzmann constant k. In 1953, Bradley [11] studied the use of thermographic phosphors mixed with binders and ceramic materials to measure surface temperature in aerodynamics. A thin layer of temperature-sensitive phosphor particles deposited on the surface of a flat plate showed the distribution of temperature on this surface when exposed to the conditions of a supersonic flow. The calibration of the phosphor was described in that case using a photograph in which the brighter parts show surface parts that were cooler by about 2 K than the darker parts of the image (see Fig. 1.1). It was also possible to show that the temperature of the surface of the flat plane was about 5 K colder by the action of the supersonic flow. By the use of a photomultiplier tube, Bradley was also able to determine the temporal response of the temperature evolution. This study may be considered the first applicative study of the, new at that time, luminescence thermometry method. Fig. 1.1 Photograph of the half-wedge of a wing model surface covered with temperature sensitive ZnCdS phosphor after 2 s of being introduced in a wind tunnel under the conditions of a supersonic flow. Adapted with permission from Ref. [15] after Ref. [11]

4

J. J. Carvajal and M. C. Pujol

2 Luminescence Luminescence is the spontaneous emission of light by a substance that is not the result of its heating process; this is why it is also called “cold light”. Luminescence should not be confused with incandescence, which is the process of light emission by a substance as a result of overheating. The process of emission of luminescence occurs after a suitable material has absorbed energy. Depending on the source of excitation used or the type of energy absorbed, the different types of luminescence are classified. For example, if an optical radiation was used as the excitation source, then this phenomenon would be called photoluminescence. This energy absorbed makes the material to pass to an excited state by elevating its electrons to a higher energetic state. Then, as an excited state is an unstable form of energetic state, the material undergoes another transition relaxing its electrons to the ground energy state, and some of the absorbed energy is released in the form of light, hence generating photons. The physical characteristics of these photons depend on the properties of the electronic states involved in their emission. Luminescence can be found in multiple applications and devices in our daily lives, such as neon and fluorescent lamps, television screens, radars and X-ray fluoroscopes, in organic substances such as the luminol highly used in forensic applications, or in luciferins that are responsibles for the generation of light in bright worms, in certain pigments used in outdoor advertising, and also in natural electrical phenomena such as lightning and the northern lights. The term “luminescence” was introduced in 1888 by Lum [64]. As mentioned above, depending on the source of excitation of the material and the type of energy that it absorbs, we can find different types of luminescence. In this book the one about which we will be talking fundamentally is photoluminescence. Photoluminescence is the result of the absorption of photons by the material. It can be divided in fluorescence that is the photoluminescence resulting of the electronic relaxation of singlet–singlet electronic transitions. In that case, the lifetime of a typical emitting excited state is of the orders of nanoseconds. Another form of photoluminescence is the phosphorescence that is the photoluminescence resulting either of a triplet–triplet electronic relaxation or the persistent luminescence. In that case the average lifetime of the emitting excited state goes from microseconds to hours. There are other types of luminescence that we want to mention, although they are not directly involved in the concepts treated in this book, but still, they are important in the context of materials science. For instance, when considering aspects related to the electrical properties of the material, we encounter the electroluminescence, that is the result of the pass of an electrical current through a material. We can also find the cathodoluminescence, in which the excitation energy comes from the impact of a beam of electrons on the material. When considering aspects related to the mechanical properties of the material, we can talk about mechanoluminescence, that is the result of a mechanical action on a solid; triboluminescence, generated when the bonds of the atoms in a material are broken when it is scratched, crushed, or

Introduction to Luminescence Thermometry

5

rubbed; or piezoluminescence, produced by the action of the application of pressure or a strength on the material. There are other types of luminescence, more specific ones, and related with the composition of the material. For instance, we can talk about radioluminescence, that is the result of the bombardment of the material with ionizing radiation. We can also find the thermoluminescence that is the re-emission of the energy absorbed when a substance is heated, or the cryoluminescence, that is the emission of light when a material cools down. Finally, a special mention should also be made to the crystalloluminescence, in which the initial energy comes from the crystallization process, and that is a special case of triboluminescence due to a chemical process, in that case the recombination of ions to form a non-dissociated salt. Thus, the most used type of luminescence in luminescence thermometry is photoluminescence. Photoluminescence is a process that, in general, proceeds in two different steps. The first one is the excitation of the emitter, i.e., the absorption of one or more photons of the excitation source by the material. This causes the electrons of the luminescent material (either molecule or ion) to jump to higher electronic states. From here, the second step takes place, when the electrons of the luminescent material relax, returning back to the ground energetic state or to an intermediate electronic state, this extra energy will be emitted in the form of light (if a radiative process takes place) or dissipated in the form of heat (if a non-radiative process is occurring) [94]. The Jablonski energy level diagram shown in Fig. 1.2 illustrates this luminescence process, including the absorption of energy by the electrons of the luminescent material, the internal energy conversion processes, and the luminescence phenomena that might occur. The properties of the light emitted through these processes depend on the properties of the electronic levels involved, which are influenced, among other parameters, by the local temperature of the system. When a change in temperature is produced, many of the parameters describing the light emitted are modified, including the intensity of the luminescent bands emitted, the shape of the emission spectra, or the luminescence decay time, among others. Many of these parameters can be exploited to determine the temperature of a particular event if the physical mechanism that produces the change in the considered parameter is known. Thus, we can define that luminescence thermometry consists of correlating the changes in these parameters of the light emitted with the variation of temperature produced.

3 Methods Used in Luminescence Thermometry Methods used in luminescence thermometry can be classified in two big groups: (i) time-integrated methods; and (ii) time-resolved methods. For time-integrated methods, luminescent materials should be illuminated with a constant intensity of light during the period of detection of the luminescence generated, and then, the signal integration, or the spectral changes of the emission spectra generated, can be analyzed.

6

J. J. Carvajal and M. C. Pujol

Fig. 1.2 Schematic representation of the luminescence process through a Jablonski energy level diagram. The arrows looking up represent the absorption of energy that might occur through the absorption of light. The straight arrows looking down represent the emission of light. The twisted and discontinuous arrows represent the internal conversion processes and the non-radiative processes (including vibrational relaxation and quenching processes) which hamper the emission of light, including vibration relaxation mechanisms

For time-resolved methods, a pulsed illumination source should be used, while the observation of the emission takes place during the period in between pulses, allowing analyze the temporal changes of the emission spectra. These time-resolved methods can also be developed by modulating periodically the illumination intensity while the signal analysis is produced in the frequency domain. In general, it has been observed when comparing their performances, that timeintegrated methods provide higher thermal sensitivity values than time-resolved methods [15]. Also, it has been observed that time-resolved approaches are more precise and reliable than time-integrated methods for determination of temperature in biological samples [76]. Recent investigations seem to indicate that accurate determination of temperature in biological systems using time-integrated approaches is nontrivial, as the optical transmission of biological tissues changes with temperature, distorting the emitted luminescence spectral shape before reaching to the detector [117]. To solve this problem, time-resolved methods should be used. However, the most used time-integrated techniques require monitoring two different emissions to achieve self-referencing techniques, while measurements on time-resolved methods require only the observation of one emission band. Also, we must consider that measuring temporal emission changes can be done with much more accuracy than emission intensity measurements. Nevertheless, reduced uncertainties compensate for lower sensitivities, and therefore time-integrated methods

Introduction to Luminescence Thermometry

7

show this better performance. Also, regarding the complexity and the cost of the equipment used in time-resolved methods, time-integrated approaches are more favorable, since they require simpler measurement set-ups and cheaper equipment. Here we present both methods, since they have been used to determine temperature by analyzing different parameters in luminescence thermometry.

3.1 Time-Integrated Methods There are three main parameters that define the emission bands in the luminescence spectrum generated by a substance that can be used in the time-integrated methods: • The band position: it represents the wavelength at which the maximum intensity for the emission is encountered. • The bandwidth: it represents the energy or wavelength interval for which the intensity of the emission takes half of its maximum value. • The intensity: it is the variation of the recorded signal as a function of the wavelength or energy. These three parameters might change with temperature due to different processes, including: • Population redistribution over the different electronic levels of the active ion or molecule according to the Boltzmann statistics: changes in temperature would activate the population redistribution among the different energy states. • Quenching mechanisms activated by temperature: an increase of temperature would activate mechanisms of cross-relaxation between electronic levels, while luminescence quenching centers would reduce the luminescence intensity until eliminating it completely. • Non-radiative relaxations or deactivations: through these mechanisms, electrons relax from the excited states towards the ground or intermediate states dissipating the energy in the form of heat instead of emission of light at a different wavelength. • Auger conversion processes assisted by phonons: as temperature increases processes of Auger electrons ejection would be activated, that influence the parameters of the corresponding spectra. • Dilatation or contraction of the crystal lattice: due to temperature changes, the crystal lattice in which the active ions or molecules (emitters) are embedded will expand or contract. That might produce changes in the position of their electronic levels, and consequently, changes in the luminescent properties of the material. • Changes in the refractive index of the media in which the emitters are located: again, due to the expansion or contraction of the material as the temperature changes, or to changes in its polarizability that are also a consequence of the temperature changes, the luminescent parameters might change with temperature. Based on the three main parameters that define the emission bands in the luminescence spectrum generated by a substance (band position, bandwidth and intensity),

8

J. J. Carvajal and M. C. Pujol

we can define five different methods of operation in the time-integrated approaches for luminescence thermometry: i. ii.

iii. iv. v.

intensity-based luminesce thermometry: here the variation of the intensity from a single emission band as a function of the temperature is measured; intensity ratio (or band shape) thermometry: in this technique the variation of the intensity ratio between two emission bands as the temperature changes, that also affects their band shape, is quantified; spectral bandwidth luminescence thermometry: for this approach the variation of the bandwidth of an emission band when the temperature change is determined; spectral position luminescence thermometry: in that case the spectral shift of a particular emission band as a function of the temperature is considered; spectral polarization luminescence thermometry: in this particular technique the variation of the polarization of a particular emission band as the temperature changes is analyzed.

Figure 1.3 presents, schematically, how can we observe the changes of these parameters induced by the change in temperature in an electromagnetic spectrum. These different methods of operation in the time-integrated approaches for luminescence thermometry are described in detail in the following sections.

Fig. 1.3 Time-integrated possible effects caused by an increase of the local temperature on the properties of the luminescence emission spectrum of a given emitter

Introduction to Luminescence Thermometry

9

Intensity-based luminescence thermometry It has been observed that, in general, the intensity of the emission of any material is reduced when the temperature increases until it is totally quenched. This decrease of intensity, especially in the dominant band of a spectrum, can be easily observed, even visually. For this reason, this became one of the first mechanisms used to sense temperature through luminescence thermometry [11], even in the application of this thermometric technique to biomedical sciences [55]. Intensity-based luminescence thermometry has been analyzed in different systems, including quantum dots [39, 105], organic dyes [63, 83], lanthanide-doped systems [32, 115], polymers [97], and gold nanoclusters [92]. However, luminescent thermometers based on following the evolution of the intensity of a single emission band as the temperature changes are substantially influenced by the measurement conditions, since the emission intensity depends also on other parameters. Among them, it is important to mention power fluctuations or illumination oscillations of the excitation source, the signal-to-noise ratio and instabilities in the detection set-up system, the absorption and scatter cross-sections of the emitters, the variation of the concentration of emitters in the sample where the temperature wants to be measured, or the inhomogeneity on the distribution of the emitters in the luminescent thermal probes. Thus, it requires recursive calibration procedures that are not compatible with friendly end-user applications. As an example of application of this class of luminescent thermometers, it can be described the use of fluorescence gold nanoclusters as intensity-based luminescent thermometers. For that, lipoic acid protected fluorescent gold nanoclusters, with ultrasmall sizes (1.6 ± 0.3 nm in diameter), with excellent colloidal stability in biological media, and good biocompatibility, were used as luminescent thermometers operating under the intensity-based technique [92]. These properties are crucial for the potential application of these luminescent probes in biological media especially. The emission intensity of these gold nanoclusters, which changed considerably over the physiological range of temperatures [288 K (15 °C)–318 K (45 °C)] was used as the thermometric parameter under the intensity-based luminescence thermometry technique. These gold nanoclusters emitted bright fluorescence in the near-infrared region as a single peak centered at ~700 nm, in the so called first biological window where the optical transparency and scattering of light of biological tissues is reduced, as it is described more in depth in Chap. 6, after being excited at 580 nm with an excitation power density of 2.8 kW cm−2 . Figure 1.4 shows the pronounced temperature dependence of the steady-state fluorescence emission spectra of the gold nanoclusters suspended in phosphate-buffered saline (PBS). In the figure it can be seen how the intensity of the emission of the gold nanoclusters decreased substantially from the lowest temperature at which the spectrum was recorded (283 K, 10 °C), in the upper part of the figure, to the highest temperature analyzed (318 K, 45 °C) in the lower part of the graph. This decrease in intensity represented a 67% of reduction upon raising the temperature. The inset in the figure shows how the maximum intensity of the emission band decreased as the temperature increased.

10

J. J. Carvajal and M. C. Pujol

Fig. 1.4 Dependence with temperature of the fluorescence emission intensity of lipoic acid-capped gold nanoclusters dispersed in PBS, after excitation at 580 nm. Adapted with permission from Ref. [92]

This graph indicates that an intensity change of 0.5% K−1 is required to resolve a temperature difference of 0.1–0.3 K. Intensity ratio or band shape luminescence thermometry The intensity ratio luminescence thermometry (also called band shape thermometry in some sources since it is depicted in fact by a change on the shape of the spectra) can be used when the luminescence spectrum of a given system consists of several emission bands or lines, and the intensity of at least one of them changes significantly as the temperature increases to respect the intensity of the others. So that, it exploits the relative change in the intensity ratio between two electronic transitions. At present, this is, by large, the most used method in luminescence thermometry, since it constitutes a self-referencing method. This means that the second emission serves as an internal standard to calibrate the response of the luminescent probe. The intensity ratio luminescence thermometry technique has demonstrated that it can provide a high thermal sensitivity and has been used for imaging with better thermal resolution than the intensity-based luminescence thermometry technique. It also mitigates most of the problems that have been identified for the intensity-based luminescence thermometry technique, especially those referring to changes in the measurement conditions affecting the signal obtained. This is because the intensity ratio luminescence thermometry technique determines the ratio between two absolute intensities. In the intensity ratio luminescence thermometry technique, we have to distinguish between two different cases for which even different theories to model the defined intensity ratio between two different emission lines must be used: • Single-center thermometers: this situation can be considered when the emission bands/lines used to build the luminescent thermometer are generated by a single luminescent center. Here, the changes in the band shape of the spectrum are caused by the redistribution of the electronic population among the different energy levels

Introduction to Luminescence Thermometry

11

of the emitting center caused by changes in temperature. Although simpler conceptually that the following case, even here different situations must be considered, depending on whether the emissions used to calculate the intensity ratio are generated from thermally coupled levels (known as TCL) or not, as we will discuss in the following section. • Dual-center thermometers: this situation can be considered when the emission bands/lines used to build the luminescent thermometer are generated by two different emitting centers. Here, the changes in the band shape of the spectrum are induced by the different thermal quenching ratios of each luminescent center, or by changes in the energy transfer rates between the two emitting centers, caused in all cases by changes in temperature, as we will see more in detail later. The intensity ratio luminescence thermometry technique has been reported in a wide variety of luminescent systems, such as quantum dots [44, 103], organic dyes [28, 77], and especially in lanthanide-doped systems [88, 89, 102]. However, the intensity ratio luminescence thermometry technique suffers from an important drawback when it is applied to biological systems, as will be detailed later in Chap. 8. In fact, the optical transmission of biological tissues is temperaturedependent and varies from one type to another (it is not the same the optical transmission of muscle tissue than that of the epithelial tissue, for instance). This makes the shape of the emitted luminescent spectra to be distorted by the biological tissue before being recorded by the detector. Thus, important precautions must be taken into consideration when using this technique to measure temperature in biological system, despite it is widely extended in the literature, or alternatively time-resolved techniques should be used. Single-center luminescent thermometers. Regarding the use of single-center luminescent thermometers, the ratiometric approach that constitutes the intensity ratio luminescence thermometry technique, usually exploits the ratio between the intensity of two emissions originated from two different excited states, closely separated in energy, that are considered to be in thermal equilibrium. For instance, in lanthanidedoped systems it can be considered that two different electronic energy levels are in thermal equilibrium when the energy difference between them is in the range 200–2000 cm−1 [104]. This relatively small energy difference allows the promotion of electrons from the low energy level to the high energy level using only thermal energy [81]. This situation constitutes a particular case when analyzing single-center luminescent thermometers that is referred to as the fluorescence or luminescence (depending on the bibliographic source) intensity ratio (FIR or LIR, respectively). In these systems we can apply Boltzmann’s statistics to describe the distribution of electronic population between the two electronic levels involved in this process. The electronic population of these two electronic levels is regulated according to the Boltzmann’s distribution law as:   −ΔE g2 (1.1) N1 · exp N2 = g1 kB T

12

J. J. Carvajal and M. C. Pujol

where N 2 and N 1 are the number of electrons in the higher and lower energy levels, respectively; ΔE is the difference of energy between the two electronic levels involved in the process, defined as the energy gap between the barycenters’ of the 1 → 0 and 2 → 0 emission bands; k B is the Boltzmann’s constant; gi are the degeneracies of the two energy levels; and T is the absolute temperature. According to that, FIR or LIR is defined using the emission intensities corresponding to the 2 → 0 and 1 → 0 transitions, where 0 denotes the ground state or an electronic level with lower energy than 2 and 1 [104]:     −ΔE I2 A02 hϑ02 N2 g2 A02 hϑ02 −ΔE = B · exp F I R(L I R) = = = · exp I1 A01 hϑ01 N1 g1 A01 hϑ01 kB T kb T (1.2) where A0i are the total spontaneous emission rates from the high energy levels to the ground or lower energy state, υ0i are the frequencies corresponding to the 1 → 0 and 2 → 0 transitions; and h is the Plank’s constant. To determine the values of B and ΔE/k B the natural logarithm of Eq. 1.2 can be used, since it will show a linear dependence with the inverse of temperature, and then, can be calculated from the line-slope and the intercept with the y-axis, respectively, of the function obtained. For instance, different emission lines can be used to calculate the FIR of different Er3+ , Yb3+ -doped materials that constitute the broad emission band observed at around 1500 nm, associated to the radiative transition between the 4 I13/2 first excited state and the 4 I15/2 ground state of Er3+ (4 I13/2 → 4 I15/2 ). Figure 1.5a shows the emission spectra of Er3+ at around 1500 nm when doped in different materials, including simple oxides, complex oxides, fluorides and oxyfluorides, always accompanied by Yb3+ , at room temperature and at 333 K [88, 89]. In this figure, it can be clearly seen how the shape of the spectra changes as the temperature increases. Spectra were recorded at different temperatures in this interval, and by comparing the absolute intensity of two lines in the different spectra for each material as they change when the temperature increases (in that case the emission lines located at 1534–1535 nm and at 1553–1555 nm were considered), the intensity ratio between these two lines could be calculated. Figure 1.5b shows the trend of these intensity ratios as the temperature increases, in which different slopes can be observed depending on the type of material considered. The higher the slope of the trend of the intensity ratio with the temperature, the more sensitive the luminescent thermometer is. In that case, oxyfluoride materials exhibited the highest change of the intensity ratio with temperature, and thus constitute the most sensitive luminescent thermometer of the range of materials selected, despite the emitting ion is the same in all cases (Er3+ ) and the excitation conditions were kept the same for all the materials analyzed. This indicates the importance of a correct selection of the host in which a particular lanthanide ion is embedded to optimize the sensitivity of the luminescent thermometer. From the FIR graph (Fig. 1.5b), the B parameter can be calculated, fitting in a graphic representation software the different lines plotted for the different materials with Eq. 1.2.

Introduction to Luminescence Thermometry

13

Fig. 1.5 a Representation of the Er3+ emission at around 1500 nm in different Er3+ ,Yb3+ co-doped nanoparticles for different materials (single oxides, complex oxides, fluorides and oxyfluorides) taken at room temperature and at 333 K to show the different shape of the spectra as the temperature increases. b FIR between the emission lines located at 1535 and 1555 nm of Er3+ in the different materials at different temperatures, showing the trend evolution of the intensity ratio with temperature. The highest the slope, the more sensitive the luminescent thermometer is. Adapted with permission from Ref. [88, 89]

There is still another way of determining the B parameter using the Judd–Ofelt theory [49, 72]. For this, the integrated coefficient of spontaneous emission of the J → J, transition must be considered: ⎡  ⏋ 2 64π 4 e2 ni3 n n 2 + 2 Sed + n 3 Smd AJ J, = (1.3) 3hc3 9 where e is the electronic charge, n is the refractive index of the medium, and Sed and Smd Wλ are the electric and magnetic dipole strengths, respectively, given by: Sed =

Σ |⟨ ∥ ∥ ⟩|2 1 Wλ | J , ∥U (λ) ∥ J 2 | (2J + 1) λ=2,4,6

(1.4)

h2 |⟨ 0∥L + 2S∥i⟩|2 16π 2 mc2

(1.5)

Smd =

where Wλ with λ = 2, 4, and 6 are the Judd–Ofelt intensity parameters, and m is the electron mass. U (λ) and L + 2S. are reduced matrix elements tabulated by Carnall and Crosswhite [17], where the L and S angular operators are in units of h. The W2,4,6 parameters in Eq. 1.4 can be obtained from the absorption spectra

14

J. J. Carvajal and M. C. Pujol

of the corresponding lanthanide ion [36], or if Eu3+ is used, and only in this case from its emission spectra [16]. This allows calculating the coefficient of spontaneous emission, and thus, the B parameter. As an example, and for the particular case of Er3+ and its emissions in the green arising from the 2 H11/2 → 4 I15/2 and 4 S3/2 → 4 I15/2 induced electric-dipole transitions, the B parameter can be calculated as follows [57]: n4 β2 B = 24 n1 β1 ≈

Σ

∥ ⟩2 ⟨ ∥ Wλ I ∥U (λ) ∥ H ∥ ⟩2 ⟨ ∥ W6 I ∥U (6) ∥ S

λ=2,4,6

n42 β2 0.7158W2 + 0.4138W4 + 0.0927W6 0.2225W6 n41 β1

(1.6)

where I stand for the 4 I15/2 , H for the 2 H11/2 , and S for the 4 S3/2 electronic levels. Apart from FIR (LIR), for which the electronic levels from which the emission bands arise need to be in thermal equilibrium, i.e., thermally coupled, we can also determine an intensity ratio between two emission lines arising from a single-centre luminescent thermometer whenever their intensities change at different rates when the temperature changes. This is the case, for instance of Tm3+ ions sensitized by Yb3+ , in which, after pumping at 980 nm, the energy absorbed by Yb3+ ions is transferred to Tm3+ ions that generate emissions in the blue and the deep red spectral regions of the electromagnetic spectrum, arising from the 1 G4 → 3 H6 and 3 F3 → 3 H6 transitions, respectively [87]. It can be seen in Fig. 1.6, how the relative intensity of these transitions changes when the temperature increases in the case of Yb, Tm: GdVO4 @SiO2 core–shell nanoparticles. At room temperature the blue emission, located at 475 nm, is dominant, while at 473 K (200 °C) the deep-red emission, located at 700 nm, reaches an intensity comparable to that of the blue emission, and becomes clearly dominant at 673 K (400 °C). Thus, a luminescent thermometer can be built based on the changes induced thermally in the intensity ratio of deep red and blue UC emissions, derived from the different temperature dependence of radiative and non-radiative relaxation rates of the 1 G4 and 3 F3 emitting levels, as well as of the temperature dependent energy transfer rate between them, providing an internally calibrated signal for the thermometer [45, 104]. The mechanism proposed for this luminescent thermometer can also be seen in Fig. 1.6, in which at room temperature, the population of the blue emitting 1 G4 electronic level can be explained by the sequential absorption of three photons from an energy transfer process from excited Yb3+ . In a first step, the excited Yb3+ 2 F5/2 level transfers part of its energy to the 3 H5 electronic level of Tm3+ , from which electrons relax in a very fast way towards the 3 F4 electronic level of Tm3+ . Here, a second energy transfer process from Yb3+ takes place, promoting the Tm3+ electrons to the 3 F2 energy level, which relaxes at its time, populating the 3 F3 and 3 H4 levels. Finally, the third energy transfer process occurs, promoting Tm3+ electrons in the 3 H4 level to the 1 G4 level. It is from here that the blue emission centred at 475 nm is generated, arising from the 1 G4 → 3 H6 radiative transition. The progressive quenching of the blue emission

Introduction to Luminescence Thermometry

15

as the temperature increases, and the simultaneous development of the deep-red emission arising from the 3 F3 level point towards the existence of some efficient crossrelaxation processes, labelled as CR1 in Fig. 1.6, from which both 1 G4 depopulation and 3 F3 population take place. Additionally, the enhanced population of the 3 F4 level of Tm3+ can also contribute to feed the 3 F3 level via energy transfer from Yb3+ to Tm3+ , labelled as CR2 and CR3 in Fig. 1.6. Furthermore, the 3 F3 level is only separated by ΔE = 900 cm−1 from the 3 H4 level [59], so that it can also be populated by this last level when the temperature increases [68, 111], indicated in Fig. 1.6 with orange circular arrows. These processes are evident for high Tm3+ concentrations. When we plot the intensity ratio between these two emission bands as a function of temperature (see Fig. 1.6), it can be seen that it is strongly temperature dependent, and that the plot follows an increasing exponential function that can be fitted to the

Fig. 1.6 Example of an intensity ratio based single-center luminescent thermometer involving two emission lines generated by two electronic levels that are not thermally coupled: a Evolution from room temperature up to 673 K (400 °C) of the up-conversion spectra under 980 nm excitation recorded for 1% Tm3+ , 15% Yb3+ :GdVO4 @SiO2 core–shell nanoparticles. b Schematic representation of the Yb3+ and Tm3+ electronic energy levels and proposed mechanisms responsible for populating Tm3+ levels generating the blue and deep red emissions involved in the luminescent thermometer proposed. c Calibration curve of the intensity ratio [R(T)]® between the emission peaks located at 475 and 700 nm (I700 /I475 ) as a function of temperature. Error bars reflect the reproducibility of the calibration curve after the analysis of several spectra for each measured temperature. Adapted with permission from Ref. [87]

16

J. J. Carvajal and M. C. Pujol

following equation: R(T ) =

I700 = A + B · exp(C T ) I475

(1.7)

where A, B and C are fitting parameters and R(T ) defines the intensity ratio, indicated with a different symbol than FIR (LIR) since the emissions do not arise from two thermally coupled levels. In that case, however, no equation of state governs the intensity ratio, and thus, it is not possible to develop a primary thermometer with this approach, as can be seen in Chap. 3. Dual-center luminescent thermometers Dual-center luminescent thermometers are those in which the emissions arise from two different luminescent centers. Different schemes have been proposed for this kind of luminescent thermometers: • The same particle or molecule contains the two luminescent centers that can be excited with a single wavelength. • The same particle or molecule contains the two luminescent centers, but they must be excited at two different wavelengths simultaneously. • The two luminescent centers are distributed in two different particles or molecules, but they can be excited at the same time with the same wavelength. • The two luminescent centers are distributed in two different particles or molecules, and they must be excited at two different wavelengths simultaneously. In all these cases a phenomenological equation describing the dependence of the luminescence with temperature is established, and thus, an equation of state cannot be established, and thus, primary luminescent thermometers, as those described in Chap. 3 cannot be developed using dual-center luminescent thermometers. Spectral position luminescence thermometry This method is based on the analysis of the position of the emission lines in the spectrum. The position of the lines depends on the energy separation between the electronic levels involved in their generation. Such differences in energy between the electronic levels are a function of several temperature dependent parameters, such as the refractive index of the material and the inter-atomic distances among emitting centers due to the expansion of the crystal lattice as the temperature increases, for instance [67]. Thus, a correlation between temperature and the spectral positions of the different emission bands or lines of the spectra can be established to use them as luminescent thermometers, provided that these spectral shifts are sufficiently large to allow for accurate measurements. The main advantage of using this method to determine temperature is that the measurements are not affected by luminescence intensity fluctuations due to changes in the concentration of the emitting centers, or fluctuations of the power of the excitation source, neither by the shadowing effects or the movement of the sample in which the temperature is being determined.

Introduction to Luminescence Thermometry

17

Spectral position luminescence thermometry has been developed principally using quantum dots (QD), where spectral shifts with temperature are relevant [65, 67]. Nevertheless, other materials, like lanthanide-doped nanoparticles have also been explored for this kind of luminescence thermometry [82]. However, the shift position observed in the spectra for the different emission lines was much smaller than the one observed in the case of QDs, of the order of 0.1 cm−1 K−1 . From another side, the fact that the FWHM of the emission bands of lanthanide ions is substantially narrower than that observed in QDs, makes easier the determination of the position of these emission bands. As an example, Maestro et al. [65] used CdSe quantum dots for spectral position luminescence thermometry, since the peak emission wavelength of these QDs is temperature sensitive. Figure 1.7 shows the variation of the emission peak position for these CdSe QDs as the temperature increases under one-photon (excitation in the blue at 488 nm with a continuous laser) and two-photon excitation (excitation in the near-infrared, 800 nm, with a femtosecond pulsed laser) conditions. The two-photon excitation conditions lead to a higher confinement of the luminescence generated by these nanoparticles, interesting for achieving higher spatial resolutions, even below the wavelength of excitation. Figure 1.7a, b show how the position of the luminescence band and the emitted intensity depend on the temperature. Thus, an increase in temperature results in a shift of the emission band towards longer wavelengths (red shift), while the emitted intensity is also reduced due to a reduction in the optical conversion efficiency. This latter phenomenon is more evident in the case of the twophoton excitation since near-infrared (NIR) light is used to generate visible light, a less efficient mechanism than pumping above the emission wavelength of these nanoparticles, as it happens with the one-photon excitation mechanism. Figure 1.7c the spectral shift and the variation of the emitted intensity as a function of temperature for these CdSe QDs that can be used for an accurate calibration of the spectral parameters to be used later as luminescent thermometers. A different way of using spectral position luminescence thermometry, in that case reported for lanthanide-doped materials, is through the comparison of their excitation spectra. Excitation spectra provides an alternative way of determining the position and the strength of absorption features in a material. This kind of spectra are obtained by measuring the intensity of an emission band while the excitation wavelength changes [3]. These spectra are formed normally by two different structures: (i) a strong and broad absorption band called charge transfer (CT) band, due to electron transfer processes between the host material and the doping ion [25], whose position in largely influenced by temperature; and (ii) several narrow bands due to the electronic transitions within f electrons in the doping lanthanide ions, whose position is almost temperature independent. Thus, the position of this CT band can be used as a thermal sensor. Figure 1.8 shows the change in the position of the CT band in the excitation spectra of Eu:Y2 O3 particles as the temperature increases from 114.5 to 350.0 °C. In the same spectra, it can be seen how the position of the small bands attributed to the f-f transition of Eu3+ ions are almost insensitive to temperature, while the broad and highly intense CT band changes substantially its position to longer wavelengths as the temperature increases, at a rate of 0.6 nm/°C.

18

J. J. Carvajal and M. C. Pujol

Fig. 1.7 Emission spectra of CdSe QDs dispersed in a phosphorus buffered saline (PBS) solution at three different temperatures showing the displacement of the position of the emission band and the decrease in intensity under a one-photon and b two-photon excitation conditions. c Temperature variation of the position of the emission band (top) and integrated emission intensity (bottom) recorded for the dispersion of the CdSe QDs in PBS. Adapted with permission from Ref. [65]

Fig. 1.8 Temperature dependence of the excitation spectra of Eu:Y2 O3 , in which it can be clearly seen how the charge transfer band, very broad, with a high intensity and located at shorter wavelengths changes substantially its position, while the position of the low intensity bands attributed to the f-f transitions of Eu3+ are almost insensitive to the changes in temperature

Introduction to Luminescence Thermometry

19

Spectral bandwidth luminescence thermometry Whenever temperature increases above 0 K, higher energy phonon levels are populated in a material, generating the intrinsic vibrations of the structural lattice. The population of these higher energy phonon levels leads to a broadening of the absorption and the emission bands. This is due to the fact that the excited electronic sublevels can also be populated, and they contribute to the absorption and emission processes. This phenomenon is known as the homogeneous broadening of the absorption or emission bands. From another side, the presence of different optical centers, as well as the presence of defects in the host, might generate also a broadening of the absorption and emission lines, known as inhomogeneous broadening. Since these broadening effects of the absorption and emission bands depend on temperature, they can be used to generate a luminescent thermometer. The homogeneous broadening effect is very sensitive to temperature changes, while the inhomogeneous broadening effect is less sensitive to temperature variations. The change in the broadening of the absorption or emission band can be quantified as [42]: / ω(T ) = ω0 coth



hW 2k B T

 (1.8)

where ω0 is the FWHM of the absorption or emission band at 0 K and hW is the energy of the vibration of the structural lattice or the energy of the phonon or phonons interacting with the electronic transition. Thus, as the temperature increases, the width of the absorption or the emission band becomes wider, due to the contribution of thermal vibrations to the luminescence processes, but also due to the contribution of thermal vibrations arising from the neighboring atoms and/or molecules. However, this luminescence thermometric technique has a major disadvantage, and it is that the broadening change is small, and it can only be observed in systems that present narrow absorption and emission bands such as those found in lanthanide ions [6]. Figure 1.9 shows the broadening of the emission bands of Nd3+ in YAG corresponding to the 4 F3/2 → 4 I9/2 transition and the evolution of the FWHM as the temperature increases, as well as with changing the excitation power, that results in an increase of temperature. Nevertheless, this luminescence thermometry technique has also been implemented when using QDs as luminescent thermometers [75]. This change in the width of the absorption and emission bands can be made more evident by generating electron–phonon coupling effects through the careful selection of emitting centers and hosting materials, as it has been demonstrated for instance in Tm3+ doped TiO2 particles [116] or when embedding Eu3+ in a coordination compound containing ketoprofen [54]. Despite this disadvantage, it seems that spectral bandwidth luminescence thermometry offers a higher thermal sensitivity and a lower thermal resolution than the most commonly used FIR (LIR) technique [43]. In spite of this, since discriminating the changes in the bandwidth of absorption and emission bands is a time-consuming procedure, this luminescence thermometric technique has

20

J. J. Carvajal and M. C. Pujol

Fig. 1.9 a Emission spectra of Nd:YAG at around 940 nm obtained at two different pumping powers of the excitation source emitting at 800 nm, resulting in two different temperatures (the higher the pumping power, the higher the temperature). b Full width at half maximum (FWHM) of one of the emission lines corresponding to the 4 F3/2 → 4 I9/2 transition of Nd3+ in YAG as a function of temperature. Reproduced with permission from Ref. [6]

been reported as a possibility for the determination of temperature, but not practical application has been reported for it up to now. Spectral polarization luminescence thermometry Spectral polarization luminescence thermometry is based on the changes in the polarization of the luminescence generated by the material using the property of the anisotropy of the polarization depending on the direction of observation. Thus, this luminescence thermometric technique can only be used in anisotropic materials. The degree of anisotropy in these materials depends on the temperature, as an increase in temperature accelerates the Brownian rotational motion of the emitting entities. This higher rotation speed of the emitting entities when the temperature increases causes that a higher number of photons loose the memory of the incident light polarization during their luminescence lifetime. Thus, the degree of anisotropy of the luminescence changes as the temperature changes [23]. In practice, when a group of emitting centers embedded in different bodies is illuminated with linearly polarized light, their emission will be partially polarized due to the random orientation of the different bodies. This partial polarization is due to the polarization anisotropy of the light generated, and can be expressed as:

Introduction to Luminescence Thermometry

r=

21

I parallel − I per pendicular I parallel − 2I per pendicular

(1.9)

where I parallel is the intensity of the emission band polarized parallel to the polarization of the incident illumination light, and I per pendicular is the intensity of the emission band polarized perpendicularly to the polarization of the incident light. Since this polarization anisotropy is due to the rotational diffusion induced by the molecular Brownian dynamics, that at its time depends on temperature, this correlation can be estimated by the Deby-Stokes–Einstein equation: τR =

V μ(T ) kB T

(1.10)

where τ R is the rotational lifetime; μ(T ) is the viscosity, that depends also on temperature; and V is the hydrodynamic volume of the emitting entity. By using Perrin’s equation, and considering Eqs. 1.9 and 1.10, the polarization anisotropy ban be correlated with the rotational lifetime [24]:   τL 1 1 1+ = r r0 τR

(1.11)

where τ L is the luminescence lifetime and r0 is the polarization anisotropy in the absence of any molecular motion. In general, as explained above, when the temperature increases the emitted photons will lose the memory of polarization of the incident light, since the Brownian rotational motion of the emitting particles is accelerated. Thus, the higher the temperature, the faster the Brownian rotational motion of the particles and thus, the larger the number of emitted photons that will lose the memory of the polarization of the pumping light. This is why an increase of temperature leads to a decrease of the degree of polarization of the emission band, and thus, to the polarization anisotropy. This luminescence thermometry method has the advantages that is insensitive to fluctuations in the intensity of the light emitted produced by photobleaching effects, or fluctuations of the illumination intensity, as well as changes in the concentration of the emitting centers, since it is based on a ratio of intensities. Thus, it constitutes again a self-referenced method, and after the appropriate calibration process, it gives the temperature of the medium in which the temperature probe is embedded. However, this method suffers from some severe drawbacks for practical applications. The first one is that it requires a complex measurement set-up, especially for thermal imaging purposes, since it requires the use of a polarization beam splitter, and that the luminescence is recorded simultaneously on two avalanche photodiodes. The second one is that, since the rotation Brownian motion is very fast, viscous systems, like mixtures of glycerol and water must be used to allow the measurement of the polarization degree. It is also possible to measure this parameter by increasing the hydrodynamic volume of the emitting molecule, using for instance proteins. This luminescence thermometry method has been used to map the internal temperature

22

J. J. Carvajal and M. C. Pujol

Fig. 1.10 Fluorescence polarization anisotropy (FPA) measured in a living C. elegans to determine the temperature. a Bright-field image of the living organism overlapped with the fluorescence intensity arising from the green fluorescent protein (GFP) located in the neurons of the C. elegans. b– d GFP luminescence intensity recorded at three different temperatures (24.5 °C, 29 °C and 35.5 °C). e–g Mapping of the FPA at these three temperatures. h Calibration performed for the luminescence thermometer developed by using the spectral polarization luminescence thermometry technique, in heating and cooling cycles. Reproduced with permission from Ref. [24]

of HeLa cells when heated externally [23] and map the internal temperature of a living Caenorhabditis elegans (C. elegans) modified genetically to express the green fluorescent protein (GFP) [24]. This, together with the fact that a low acquisition speed is required (it takes between 0.5 and 2 min to map the distribution temperature of the C. elegans), questions if the temperatures mapped can be considered as realtime temperatures. Figure 1.10 shows this kind of measurements taken on a living C. elegans.

3.2 Time-Resolved Methods Time-resolved methods, or lifetime luminescence thermometry, as it is commonly reported in the literature, estimates the temperature of the thermal probe from the temporal dependence of its emission. Three different moments or periods can be distinguished in general in this temporal evolution of an emission, as can be seen in Fig. 1.11. In the first instant, the emission lifetime follows the time dependence of the pulse train used for excitation. Then, a rise in the emission intensity is observed, due to the accumulation of electrons in the excited level. Finally, in the third period, the intensity of the emission decreases due to interactions with phonons and internal conversion and non-radiative processes. This last period is longer than the two

Introduction to Luminescence Thermometry

23

Fig. 1.11 General shape of the temporal evolution of the luminescence emission intensity generated by a thermal probe. Reproduced with permission from Ref. [27]

previous ones, and it is the one used normally to determine temperature in lifetime luminescence thermometry. Decay time luminescence thermometry Decay probabilities from electronic levels depend on many parameters, some of which are temperature dependent, including phonon-assisted energy transfer processes and multiphononic decay processes. In general, lifetime luminescence thermometry has an important advantage when compared to intensity-based luminescence thermometry: it does not depend on the local concentration of luminescent probes. Thus, this solves problems related to non-controllable spatial fluctuations of the fluorescence intensity due to an uneven distribution or concentration of the thermal probes on the sample to be thermally analyzed. Also, problems associated with the uncontrolled motion of the thermal probes (by themselves or because they are being carried in a fluid, for instance) are also avoided. Finally, the distribution and shading of the light generated by the sample, that might also mask some characteristics of the recorded spectra can be minimized. Consequently, decay time luminescence thermometry is a self-referencing and robust method. It also allows for very fast time resolutions of the order of picoseconds to milliseconds, depending on the thermal probe used. With this technique, temperature can be determined at the time interval of the luminescence lifetime that is independent on light scattering, reflection, and intensity fluctuations of the excitation source. Also, this luminescence thermometry method has been used for high temperature measurements, avoiding the undesired contribution of the blackbody radiation [40]. An additional advantage is that this technique does not require recording the complete luminescence spectrum, saving time during temperature measurements. This is especially important when low signal levels from the thermal probes are generated, in cases, for example, where high spatial resolution is required, or when living organisms are being analyzed, allowing also for thermal imaging in real time.

24

J. J. Carvajal and M. C. Pujol

These hypothetical short measuring times would avoid, or at least minimize, problems related with local heating of the system under investigation by long lasting illumination with the pumping source. Normally in decay time luminescence thermometry time domain techniques are used to measure the emission decay times. These techniques require using a pulsed excitation source, together with long illumination and acquisition times, which at the end limit the use of this method, although the recent technological advances simplified its use and reduced the costs of this kind of excitation sources. Nevertheless, the use of these techniques has been widely used in lifetime luminescence thermometry. Figure 1.12 shows the luminescence decay curves of the green emission of Er3+ in NaYF4 and NaY2 F5 O particles, always co-doped with Yb3+ , at two different temperatures. It can be easily seen that the lifetime of this luminescence emission is more sensitive to temperature changes for NaY2 F5 O than for NaYF4 particles, due to the activation of phonon-assisted processes and of multiphonon decay processes as the temperature increases. The reason for these changes is the higher phonon energy exhibited by the oxofluoride compound when compared to the fluoride one, as can be seen in the Raman dispersion spectra shown in Fig. 1.12e. This results in a more sensitive decay time luminescent thermometer when using Er,Yb:NaY2 F5 O particles. Rise time luminescence thermometry As indicated above, a rise of the intensity of the emission is observed shortly after the thermal probe is illuminated with an excitation pulse. This is due to an accumulation of electrons in the excited level of the thermal probe because of fast non-radiative transitions ending to this level and generated from higher energy levels. Also, energy migration processes from nearby excited ions must be considered. All these processes are much faster than the radiative transition of the thermal probe to the ground state or to a lower energy state [27]. This technique, however, has been much less used for luminescence thermometry than the decay time technique. For the particular case of lanthanide ions, the number of electrons in the emitting level as a function of time can be approximated to [73]: ⎡  ⏋ −t N (t) = N0 + N1 · 1 − exp τr

(1.12)

where N0 is the number of electrons directly excited within the ion, N1 is the number of electrons accumulated by energy migration from nearby excited ions, t is time and τr is the rise time. Thus, the general temporal dependence of the emission intensity, considering it proportional to the number of electrons in the excited level, might be expressed as: 

⎡



−t I (t) = A + B · 1 − exp τr

⏋



−t · exp τd

where τd is the decay time and A and B are fitting parameters.

 (1.13)

Introduction to Luminescence Thermometry

25

Fig. 1.12 Example of a decay time luminescent thermometer based on the green emission of Er3+ in a NaYF4 and b NaY2 F5 O. The luminescence decay curves were recorded at 25 and 60 °C. Calibration of the luminescent thermometer based on the lifetime measurements as a function of temperature for c Er,Yb:NaYF4 and d Er,Yb:NaY2 F5 O, showing how the oxofluoride compound has a higher slope, allowing for the development of a more sensitive thermometer. e Raman spectra of Er,Yb:NaYF4 and Er,Yb:NaY2 O5 O particles, showing that the reason for having a more sensitive thermometer with NaY2 O5 O is due to the higher phonon energy of these particles

26

J. J. Carvajal and M. C. Pujol

Fig. 1.13 Example of a rise time luminescent thermometer constituted by Eu:SrY2 O4 particles showing a the temporal evolution of the emission, and b the evolution with temperature of the rise time and the time to which the maximum intensity of emission was achieved. Adapted with permission from Ref. [61]

As an example of the application of this technique has been the use of the emission rise time of Eu:SrY2 O4 particles, in which it was observed that after the pulse excitation, the energy transfer between the nearby Eu3+ ions caused a notable emission rise from the 5 D0 level before the decay period started. It was also observed that as the temperature increased, the emission rise time decreased linearly. This allowed demonstrating that the time passed between the pulse excitation and the moment at which the emission reached its maximum intensity can be used to sense temperature [61]. Figure 1.13 shows the temporal behavior of the emission of Eu3+ in SrY2 O4 recorded at different temperatures between 20 and 200 °C. It also shows the linear decrease of the rise time and the time to which the maximum intensity of emission was achieved as the temperature increased. One of the advantages of using rise time luminescence thermometers is that they can be used at low temperatures, a thermal region in which the decay times of lanthanide ions are almost insensitive to temperature changes. However, the evaluation of rise time values using Eq. 1.13 is quite complex, making difficult the thermal imaging with this method.

3.3 Other Luminescence Thermometry Methods Apart from the methods described up to here to determine temperature by using the characteristics of the emission spectra, other methods have been reported exploiting the interactions between luminescent centres and temperature.

Introduction to Luminescence Thermometry

27

Luminescence thermometry based on the Förster resonant energy transfer (FRET) process In dual emission systems constituted by different emitting entities, energy transfer processes among these emitting bodies that affect to their emission intensity can also be exploited to determine temperature if this emission is sensitive to temperature. For instance, the molecular design of dual-emitting luminescent nanoparticles constituted by semiconducting quantum dots and quantum rods with their surfaces functionalized with organic dyes allowed the demonstration of a new class of luminescent thermometers taking advantage of the Förster resonant energy transfer (FRET) process between these luminescent entities [2]. In a FRET process a donor chromophore, initially in its excited electronic state, transfer energy to an acceptor chromophore through non-radiative dipole–dipole coupling. Figure 1.14a shows the mechanism of the FRET process in a Jablonski diagram. The efficiency of this energy transfer process depends on the distance between the donor and the acceptor. Albers et al. [2] used quantum dots and quantum rods acting as donors, and organic dyes acting as acceptors. In a practical example, CdSe quantum dots and CdS quantum rods, passivated with an amphiphilic polymer shell to which cyanine dyes were linked were used as luminescent thermometers with an average separation distance of around 7.5 nm. While quantum dots and quantum rods exhibit giant extinction coefficients and bright temperature-sensitive luminescence, the organic dyes show temperaturedependent luminescence quantum yields, but no wavelength shifts. Together, they constitute FRET pairs sensitive to temperature changes, ideal for ratiometric temperature determination using luminescence. Figure 1.14b shows how the spectra of these FRET pairs change with temperature. The shift of the maximum of the emission band observed for quantum dots and quantum rods towards longer wavelengths was accompanied by a modest decrease in the emission intensity. Instead, for organic dyes the intensity of the emission band, located at longer wavelengths, substantially decreased as the temperature increased, but the position of the band did not shift. The ratiometric thermometer was build up by dividing the integrated intensity of the emission band of the quantum dots and quantum rods by the integrated intensity of the emission band of the organic dye, as can be seen in Fig. 1.14c, d. Luminescence thermometry based on the presence of defects in materials Defects existing in inorganic host materials might generate color centers and altered emissions that can be used also to determine temperature by luminescence procedures. For instance, nitrogen-vacancy color centers generated in diamond nanocrystals have been used as luminescent thermometers [53]. Nitrogen vacancy color centers in diamond are point-defects consisting of a nitrogen atom occupying a substitutional site adjacent to a vacant lattice site. The excellent thermal conductivity of diamond ensures that all nitrogen-vacancy centers within the nanocrystal are in thermal equilibrium with temperature of the environment in which they are surrounded. In that case, the quantum mechanical spin associated to the nitrogen-vacancy color centers constituted the temperature reading method. In its electronic triplet ground state, each nitrogen-vacancy center constitutes a spin-characteristic system. These spin states

28

J. J. Carvajal and M. C. Pujol

Fig. 1.14 a Jablonski diagram showing the general mechanism of a FRET process. Reproduced with permission after Alex M Mooney - Own work, CC BY-SA 3.0, https://commons.wikimedia.org/ w/index.php?curid=23197114. b Example of how the FRET process can be used in a luminescent thermometer to determine temperature. In that case the FRET pair was constituted by CdSe quantum dots and CdS quantum rods, with a broad emission band located at around 620 nm, linked to Alexa647 cyanine dye, with a broad emission band located at around 670 nm, through an amphiphilic polymer that covers the surface of the quantum dots and quantum rods. c Ratiometric response of the luminescent thermometer shown in (b) by calculating the intensity ratio (R) as the quotient between the integrated intensity of the emission band of the quantum dots/quantum rods in the spectral region 630–640 nm, and the integrated intensity of the emission band of the dye in the spectral region 664–674 nm. d Changes of the quantum yield (Q, circles) and FRET efficiency (E, squares) for the dye, and shift of the emission band of the quantum dots/quantum rods (S, triangles) as the temperature increased. Adapted with permission from Ref. [2]

can be coherently manipulated using microwave pulses and efficiently activated and detected by means of excitation with laser light. In the absence of an external magnetic field, the precise value of the transition frequency between the triplet ground and its Zeeman split states generated when applying an external magnetic field, depends on the temperature due to thermally induced lattice strains. These electronic states of the nitrogen-vacancy color centers lie within the band gap of diamond. It is a transition dipole allowed transition that can be excited by illuminating with green light, generating an emission at longer wavelengths located between 637 and 800 nm. The operational principle of a nitrogen-vacancy-based luminescent thermometer relies on the accurate measurement of this transition frequency, which can be optically detected with high spatial resolution. Figure 1.15 shows the simplified level scheme for the nitrogen-vacancy color center with its corresponding transitions generating these emission bands. The change in temperature in that case was determined by recording the luminescence spectra at four different microwave frequencies as:

Introduction to Luminescence Thermometry

29

Fig. 1.15 a Simplified level scheme of nitrogen-vacancy color centers in diamond. Green arrows indicate excitation using the green laser light, red arrows represent the spin conservative luminescence, and grey arrows represent non-radiative processes. b Luminescence thermometry using nitrogen-vacancy centres in nanodiamonds as a function of the echo evolution time used to cancel unwanted external magnetic field fluctuations. c Luminescence as a function of temperature for two different echo times after conversion to population by normalizing to two reference measurements where spin values were known (0 and –1). Adapted with permission from Ref. [53]

    1 f + f2 − f3 + f4 δω     δT = dΔ/dT f 1 − f 2 + f 3 − f 4

(1.14)

where f i are each one of the fluorescence recorded for each microwave frequency, ω is the microwave frequency applied in each case, and dΔ/dT is the dependence of the value of the transition frequency of the spin states with temperature. Figure 1.15 also shows the measured luminescence as a function of the echo evolution time used with a damped cosine function generated by two distinct frequencies to cancel unwanted external magnetic field fluctuations, and as a function of temperature after luminescence is converted into population by normalizing to two reference measurements where the spin values are known. However, the mode of operation of this type of luminescent thermometer is not simple. It is necessary to decouple the nitrogen-vacancy electronic spin from the fluctuating external magnetic field. For that a microwave pulse at a particular frequency has to be applied to create a coherent superposition that eliminates the magnetic-fieldinduced shifts of the spin states, allowing for accurate temperature sensing. Also, pure diamond spinless 12 C isotope (99.99%) had to be used to further reduce the magnetic

30

J. J. Carvajal and M. C. Pujol

field fluctuations originating from the intrinsic 13 C nuclear spin bath. Even when using these precautions, the authors observed a characteristic low frequency beating of the luminescence signal that changed from defect to defect, due to locally fluctuating charge traps. Nevertheless, for 30 s of integration time, a temperature resolution of 1.8 ± 0.3 mK was achieved, yielding a very robust luminescent thermometer.

4 Analysis of the Performance of Luminescent Thermometers The luminescence thermometric methods described previously, except for the primary thermometers that use an equation of state, require a calibration procedure prior to being used. This calibration procedure must be performed against an independent thermal probe, so that it will be possible to correlate the value of the parameter determined by the luminescent thermometer, i.e., absolute intensity, intensity ratio, spectral position, spectral bandwidth, polarization, decay time or rise time, or other operation forms, with the temperature determined by the independent thermal probe. Once the luminescent thermometer is in operation, different parameters can be used to determine its performance, and compare it with that of other thermometers, knowing if this particular luminescent thermometer determines better the temperature for a particular application. In fact, the quantitative comparison of the performance of any temperature probe is a critical parameter to evaluate its use and its applicability, allowing to compare it with other thermometric techniques. The performance of a luminescent thermometer can be evaluated based on its thermal sensitivity, temperature uncertainty or resolution, repeatability, reproducibility, spatial resolution and temporal resolution, that can be also considered as the figures of merit of luminescence thermometry. In this section, we will discuss how the performance of the luminescent thermometers should be quantified.

4.1 Thermal Sensitivity Thermal sensitivity is one of the most important parameters to evaluate the performance of a luminescent thermometer. Thermal sensitivity describes the variation rate of the designed thermometric parameter at small changes in temperature. Two different thermal sensitivities can be defined that are normally used in luminescence thermometry: the absolute thermal sensitivity (Sabs ) and the relative thermal sensitivity (Sr el ). Absolute thermal sensitivity The absolute thermal sensitivity is defined from the first derivative of the thermometric parameter (Δ) considered to respect the temperature:

Introduction to Luminescence Thermometry

Sabs =

31

dΔ dT

(1.15)

Absolute thermal sensitivity represents the amount of change induced thermally in the spectral or temporal characteristics of the signal recorded for the luminescent thermometer. It is meaningless to compare quantitatively the absolute thermal sensitivity among thermometers of different nature operating by different physical principles, or even though they use the same physical principle, if they operate using different materials. It can be used to compare the influence of the synthesis method or the different geometry of the material, including size, shape, etc. on its thermometric performance. Another use of the absolute thermal sensitivity is to know how changes in the concentration of the emitting centers in a particular luminescent thermometer affect its performance. However, this parameter cannot be used to compare the performance of luminescent thermometers operating under different methods, or to compare luminescence thermometry against other thermometric techniques, since it is influenced by the characteristics of the sample and the experimental setup used to record the evolution of the thermometric parameter with temperature [14]. In the particular case of the intensity ratio luminescence thermometry when the emitting levels are thermally coupled levels, the most used luminescence thermometry method at present, Sabs takes the form: Sabs

  ΔE dF I R −ΔE ΔE = FIR = exp =B 2 dT kB T kB T kB T 2

(1.16)

As an example, on how the absolute thermal sensitivity can be used for comparison of the same material with different concentration of dopants, Savchuk et al. compared this parameter for Tm,Yb:GdVO4 @SiO2 core–shell nanoparticles containing different concentrations of Tm3+ . Figure 1.16 shows the normalized FIR as a function of temperature for the Tm,Yb:GdVO4 @SiO2 core–shell nanoparticles containing different concentration of Tm3+ and their absolute thermal sensitivity. As expected from the different slopes of FIR obtained, different Sabs values were obtained for the different concentrations. Thus, according to these results, the concentration of doping ions plays a relevant role in defining the performance of the luminescent thermometers. Ciric et al. [20] proposed to calculate the absolute thermal sensitivity for Er3+ doped materials from the Judd–Ofelt parameters, the refractive indices of the host material, the Slater integrals, the spin–orbit parameters, the reduced matrix elements, and the energy differences between the emitting levels used for FIR luminescence thermometry. The tendency they predicted for different host materials considering this theoretical approach matched the tendencies observed experimentally, as can be seen in Table 1.1. So that, molybdates, tungstates, vanadates and simple oxides show the highest values of absolute thermal sensitivities, while fluorides show significantly lower values [70]. For lifetime luminescence thermometry based on measuring the decay-time, the absolute thermal sensitivity has been traditionally substituted by the normalized

32

J. J. Carvajal and M. C. Pujol

Fig. 1.16 Example on how the absolute thermal sensitivity can be used to compare the performance of a luminescent thermometer based on Tm,Yb:GdVO4 @SiO2 core–shell nanoparticles containing different amounts of emitting centers. a Normalized fluorescence intensity ratio as a function of temperature, and b absolute thermal sensitivity calculated using Eq. 1.16

Table 1.1 Comparison of the theoretical values calculated for Sabs by Ciric et al. [20] and those encountered experimentally as listed in [70] for different Er3+ -based luminescent thermometers operating under the FIR technique

Host material

Theoretical

Experimental

Tungstates

0.0324 for LiLa(WO4 )2 0.0144 for KY(WO4 )2

0.0149 for SrWO4 0.0107 for BaWO4 ) 0.006 for NaY(WO4 )2

Y2 O3

0.0181

0.0050–0.0194

Molybdates

0.0173 for LiLa(MoO4 )2 0.0073 for Gd2 (MoO4 )3

0.0235 for YMoO4 @SiO2 0.0227 for BaMoO4

Vanadates

0.0140 for YVO4

0.0116 for YVO4 0.0125 for GdVO4 0.0101 for GdVO4 @SiO2 0.0057 for LuVO4 @SiO2

LiNbO3

0.0122

0.0075

YAG

0.0028

0.0017

Y2 SiO5

0.0012

0.0070

thermal coefficient [40]: | | nor m | dt (T ) || | at = | | dT

(1.17)

where t nor m (T ) is the emission decay time measured at a particular temperature T , normalized to the luminescence decay time measured at room temperature. However, as can be seen by comparing Eqs. 1.17 and 1.16, the normalized thermal coefficient

Introduction to Luminescence Thermometry

33

corresponds to the absolute thermal sensitivity, since it is the derivative of the luminescence thermometer parameter considered, the emission decay time in this case, to respect the temperature. In any case, it should be noted that both Sabs and at depend on temperature. So that, if only a value is given for these parameters instead of a function or a graph, the temperature at which they have been calculated should be given. Relative thermal sensitivity Relative thermal sensitivity has been proposed as a parameter that facilitates the quantitative comparison among different thermometers. It was originally proposed in 1998 in the context of optical fiber point temperature sensing [21]. It represents the relative change of the thermometric parameter considered per degree of temperature change. It is defined as [104]: Sr el =

1 dΔ Δ dT

(1.18)

Normally, the relative thermal sensitivity is expressed in units of percentual change per Kelvin of temperature change [% K−1 ]. Again, Sr el is a function of temperature, as was happening with Sabs . Thus, if only a value for this parameter is given, it should be accompanied of the temperature at which it has been calculated. Sometimes, its maximum value is denoted as Sm , occurring at a temperature denoted as Tm [15], although this way of expressing it is not very extended in the literature. Brites et al. [13] proposed it to compare the performance of different thermometers independently of their nature (luminescent or not) operating at the nanoscale. Since then, it has converted in the commonly used figure of merit to evaluate the performance of any new luminescent thermometer developed, even if they operate by different mechanisms and are based on different types of materials. When compared to Sabs , the relative thermal sensitivity has the critical advantage of being independent of the nature of the thermometer used, allowing a direct and quantitative comparison between different materials, techniques, etc. For that, it constitutes a powerful tool when different thermometric techniques and methods are considered. In the particular case of the intensity ratio luminescence thermometry, Sr el takes the form: Sr el =

1 1 dFI R = Sabs F I R dT FIR

(1.19)

Thus, taking the value of Sabs determined in Eq. 1.16, the relative thermal sensitivity for intensity ratio luminescence thermometry can be defined as: Sr el =

ΔE 1 ΔE FIR = FIR kB T 2 kB T 2

(1.20)

34

J. J. Carvajal and M. C. Pujol

In that case, according to Eq. 1.20, the relative thermal sensitivity depends on the energy difference between the thermally coupled excited states involved in the transitions used to determine the thermometric parameter. This means that by using larger energy differences between these thermally coupled states, at the same temperature, the relative thermal sensitivity should be higher. Thus, the relative thermal sensitivity is an inherent property of the thermal probe used. When considering lanthanide ions as the thermal probes (see Chap. 4), their ΔE values are almost independent of the host in which they are embedded. This means, that the relative thermal sensitivity that can be achieved with a certain lanthanide ion will be pretty much the same independently of the host material in which this lanthanide ion is embedded. In general, we consider that two electronic states in a lanthanide ion are thermally coupled when ΔE is found in the range 200–2000 cm−1 . During a long time, and even now, in the literature the terms of Sabs and Sr el have been used indistinctly, independently of their definition. So, it is important that the reader check carefully the definition given for the thermal sensitivity in a particular publication to verify the extent of its applicability.

4.2 Temperature Resolution The temperature resolution, also called temperature uncertainty or thermal resolution, and symbolized by δT , is another important parameter used to analyze the performance of a luminescent thermometer. The temperature resolution provides information about the smallest temperature change that can be resolved by a particular luminescent thermometer. This parameter depends on one side, on the inherent properties of the material that forms the luminescent thermometer, and on the other side, on the experimental detection setup and the acquisition conditions used to record the response of the luminescent thermometric parameter, as well as on the signal-to-noise ratio (SNR) in the operating conditions. The temperature resolution can be determined experimentally by analysing the fluctuation with time of the thermometric parameter when it is at a certain temperature, measured by an independent thermal probe. For that, first the temperature at each measured point using the previously established calibration curve must be calculated. Then, the mean value and the standard deviation of the measurements from this distribution of values can be calculated. Finally, we can correlate the temperature resolution of the luminescent nanothermometer with the standard deviation of the resulting temperature histogram [87]. Figure 1.17 shows the result of this procedure for Er,Yb:GdVO4 nanoparticles coated with an inert shell of SiO2 . For that, 60 consecutive emission spectra were collected while the temperature was fixed at 313 K, and controlled with a thermocouple. This results in a histogram reproducing a Gaussian distribution of temperatures centred at 313.5 K with a FWHM of 0.4 K, as can be seen in Fig. 1.17 that provides an estimation of the temperature resolution achieved with these nanoparticles under these specific operating conditions.

Introduction to Luminescence Thermometry

35

Fig. 1.17 Example on how the temperature resolution can be determined experimentally by analysing the fluctuation with time of the intensity ratio at a fixed temperature. Reproduced with permission from Ref. [87]

This experimental way of determining the temperature resolution of a particular luminescent thermometer allows estimating potential effects of other parameters, like the fluctuations produced on the excitation source, for instance. Thus, when determining the temperature resolution there is always an implicit correlation between the fluctuation of the readout of the thermometric parameter used and δT , as can be seen in Fig. 1.18. It should be noted that when using luminescent thermometers exhibiting a higher relative thermal sensitivity, the temperature resolution obtained will always be smaller, as can also be seen in Fig. 1.18. This experimental determination of δT allows also estimating the effects of the excitation wavelength or pump power on the temperature determined by the thermometer. If a drift on the temperature readout is observed, it might be an indication

Fig. 1.18 a Correlation between the readout fluctuations of a luminescence thermometer parameter and the temperature resolution. b The temperature resolution depends on the thermal sensitivity. When the relative thermal sensitivity increases, temperature resolution can get smaller values. Reproduced with permission from Ref. [14]

36

J. J. Carvajal and M. C. Pujol

Fig. 1.19 Time evolution of the intensity ratio calculated for bare and silicon oxide coated Er,Yb:GdVO4 nanoparticles, calculated by dividing the integrated intensity of the green emissions arising from the 2H 4 11/2 and S3/2 electronic levels of Er3+ when relaxing to the ground state. Adapted with permission from Ref. [87]

that the wavelength or the pump power used to generate the emission are heating the temperature probe, and thus that the temperature determined might not be correct. Figure 1.19 shows an example of this situation. When Er3+ and Yb3+ co-doped GdVO4 bare nanoparticles are illuminated with a 980 nm emitting laser at a power of 100 mW, the readout of temperature increases constantly with time, indicating that the laser excitation is heating the luminescent thermometer. This effect can be prevented when coating these nanoparticles with a layer 5 nm think of SiO2 [87]. Thus, temperature resolution depends to a high degree on the uncertainty on the intensity measurements. This uncertainty is mainly determined by the detector used, as well as by the thermal sensitivity that can be achieved. Photomultiplier tubes are the most sensitive detectors in terms of intensity discrimination, followed by CCDs and photodiode array detectors. Thus, by using photomultiplier tubes, one can get a smaller temperature resolution. However, although photomultiplier tubes are the detectors offering the fastest response time, their incorporation in experimental setups using a set of gratings and moving optics, results in acquisition times of the order of minutes for a single spectrum. So, this might make that determining δT by recording a set of temperature readouts with statistical significance is not a feasible possibility. In these situations, the temperature resolution can be calculated from the relative thermal sensitivity. Thus, by assuming that δT of a luminescence thermometer is a result only of the changes in the thermometric parameter used, then the temperature resolution can be determined from the Taylor’s series expansion of the variation of the thermometric parameter with temperature [14]: δT =

1 d2T dT 1 dn T 2 δΔ + + · · · + (δΔ) (δΔ)n dΔ 2! dΔ2 n! dΔn

(1.21)

where δΔ is the uncertainty in the determination of Δ, that can be approximated to the standard deviation of the distribution of values of the thermometric parameter.

Introduction to Luminescence Thermometry

37

By considering that the Taylor’s series expansion determining δT is dominated by the first term, and taking into account Eq. 1.18, then this performance parameter can be expressed as: 1 δΔ dT δΔ = δT ∼ = dΔ Sr el Δ

(1.22)

δΔ/Δ is a parameter that depends on the experimental setup used to record the spectra, the detector used, and the integration time invested to record the thermometric parameter. Brites et al. [14] determined that for the typical portable detectors we can find in a common luminescence laboratory, the minimum value of δΔ/Δ that can be reached is 0.1%. Under these conditions, and using luminescent thermometers that might allow getting a value of Sr el of the order of 1–10% K−1 , the temperature resolutions that can be reached are in the range of 0.01–0.1 K. If detectors that are more sensitive are being used, such as photomultiplier tubes or CCDs, δΔ/Δ can take values one order of magnitude higher, of the order of 0.03%, resulting in δT values even smaller, of the order of 0.003 K. δΔ/Δ can also be improved by adjusting the SNR of the emission spectrum by increasing the integration time and/or averaging several consecutive measurements. The temperature resolution values that can be achieved under these conditions are impressive, much better than those obtained with wired thermistors and thermocouples (0.01 K) and non-contact infrared cameras (1.0 K). However, the detectors used normally in luminescence thermometry operate above these minimum values of δΔ/Δ. They operate typically in the range 0.5–2.5%, since a compromise between reducing δT and increasing the spectra acquisition time should be reached. This implies that the achievable values of δT will increase, although temperature resolutions in the sub degree range can be easily achieved with these common detectors. Nevertheless, as will be discussed in Chapter 11, this parameter gives an incomplete description of the potential performance of the luminescent thermometer considered, since the precision of a temperature measurement depends not only on the temperature resolution, but also on the luminescence strength compared to the measurement noise and background signal [100]. To determine exactly the δΔ/Δ value at which the detector used is operating, the root mean square values of the deviations in the measurement of the intensities of the emissions must be taken into account [14]. When using the FIR model it can be expressed as: δΔ = Δ

/

δ I1 I1

2

 +

δ I2 I2

2 (1.23)

where δ Ii /Ii (I = 1, 2) are estimated using the signal-to-noise ratio values of the detector. δ Ii /Ii can also be estimated by recording the readout fluctuations of the baseline and the maximum intensity value of the spectra recorded and dividing one by the other. If I1 and I2 do not differ substantially, this means that the detector response is similar for both emission lines, and the determination of δΔ/Δ can be

38

J. J. Carvajal and M. C. Pujol

simplified to: δΔ √ δ I = 2 Δ I

(1.24)

4.3 Spatial and Temporal Resolutions Spatial and temporal resolutions are two other relevant parameters used to analyze the performance of a luminescent thermometer. These two are important parameters to evaluate the applicability of a particular luminescent thermometer for dynamic temperature determination purposes, i.e., when it is necessary to know how different temperatures at different spatial positions along the sample are, and when it is necessary to know how these temperatures evolve with time. The spatial resolution (δx) determines the minimum distance for which two different temperatures can be resolved when the temperature is measured at different spatial positions. One should note that the value of δx should be higher than the thermal resolution determined for a particular luminescent thermometer, otherwise these temperature measurements would not be reliable. Under this consideration, δx can be determined as [14]: δT | δx = | | → | |∇T |

(1.25)

max

| | | → | where |∇T is the maximum temperature gradient observed in the measurement | max of temperature. As can be deduced from Eq. 1.25, the spatial resolution can be improved by decreasing δT by using detectors with better signal-to-noise ratio values. Alternatively, δx can also be improved by increasing the temperature gradient in the sample. Another way of determining the spatial resolution in the particular case of the confocal regime when a microscope objective is used to focus the excitation source into the sample, and assuming that spatial filters were not used, is [38]: δx =

0.5λem NA

(1.26)

where N A is the numerical aperture of the microscope objective and λem is the emission wavelength of the luminescent thermometer. This is an easy way of obtaining submicrometric spatial resolution using conventional microscope objectives (with N A 0.6). From its side, the temporal resolution (δt) determines the minimum time interval for which two different temperatures can be resolved. Again, it should be higher than δT to have a reliable measurement of the temperature. It can be expressed as [14]:

Introduction to Luminescence Thermometry

39

Fig. 1.20 (left) Scheme of a luminescent thermometer (Er,Yb:PbF2 nanoparticle) incorporated at the end of the scanning tip of a scanning thermal microscope, together with the emission spectra of the luminescent thermometer in the green recorded at low (25 °C) and high (100 °C) temperatures. (right) Temperature map of a Ti microstripe heater fabricated on a Si substrate having an oxide thickness of 100 nm, in which the Ti microstripe has regularly spaced constrictions with a size of 250 × 250 nm2 , in which the electrical current density is about five times higher than in the wider parts of the microstripe, thus creating hot spots. The top-left image is an SEM picture of the structure. The top-right curve is a cross section showing that the measured vertical FWHM of the hot spot is smaller than 400 nm. Adapted with permission from Ref. [85]

δt =

δT |dT /dt|max

(1.27)

where |dT /dt|max is the maximum temperature change per unit of time. A small spatial resolution in luminescence thermometry, of the order of ~25 nm, have been proved when incorporating a luminescent probe to the scanning tip of a scanning thermal microscopy setup, as shown in Fig. 1.20 [1, 85]. From another side, by using luminescent thermometers, temporal resolutions of the order of milliseconds can be achieved by using lanthanide-doped materials [87– 89]. In fact, the temporal resolution in this kind of luminescent thermometers will be limited by the radiative lifetime of the active ions used, since two consecutive temperature measurements cannot be taken at shorter times than the radiative decay-time of the emissions used. Otherwise, the temperature measurements would not be reliable. However, it must be considered that the temporal resolution is also determined by the integration time of the detector used. For instance, Fig. 1.21 shows an example of the evolution of temperature in an aqueous dispersion of Tm,Ho:KLu(WO4 )2 nanoparticles, used as luminescent thermometers to measure the temperature in this suspension [88, 89]. The laser used for excitation simultaneously heats and excites the sample providing temperature dependent luminescence. By using this temporal dependence of the temperature, the thermal resistance of the luminescent thermometers could be determined [88, 89], which indicates the potential of luminescence thermometry to determine parameters that otherwise are very difficult to measure.

40

J. J. Carvajal and M. C. Pujol

Fig. 1.21 a Scheme of the experimental setup used to measure the real-time evolution of the emission spectra of Tm,Ho:KLu(WO4 )2 nanoparticles upon 808 nm excitation. The excitation laser used simultaneously heats and excites the sample providing the temperature dependent luminescence spectra. A beam splitter was used to redirect a part of the emitted signal to the portable spectrometer to simultaneously record the emission spectra. b Temperature increase induced by different excitation power densities using luminescent nanoparticles containing 5 at. % (black squares) and 15 at. % of Tm3+ (green circles) while the concentration of Ho3+ was kept at 1 at. % for both particles. c Evolution with time of the increment of temperature with time in the dispersion of these nanoparticles in water recorded at an excitation power density of 318 × 106 W m−2 . The solid line corresponds to the best fit to experimental data, from which the thermal resistance of the luminescent nanoparticles can be calculated. Reproduced with permission from Ref. [88, 89]

4.4 Repeatability and Reproducibility Repeatability and reproducibility are also two important parameters to be determined in a luminescent thermometer that will condition its applicability. They can be considered as the two components of precision in a measurement system. These parameters are especially important for sensor engineering, since it is critical to achieve the same response under the same external stimulus when monitoring continuously industrial and scientific applications.

Introduction to Luminescence Thermometry

41

The repeatability, also known as test–retest reliability, determines the variation among repeated measurements made under identical conditions. A temperature measurement is repeatable when different measurements taken with the same measurement setup or method over a certain period of time coincide. Under these circumstances, if variability in the measurements is observed when performing a repeatability study, it can be ascribed only to errors due to the measurement process itself. Thus, repeatability describes the capacity of the luminescent thermometer to generate the same results when operated under the same conditions during several heating and cooling cycles. In fact, it indicates the degree of coincidence of the evaluation of temperature of the luminescent thermometer when compared to the measurement performed by an independent reference temperature probe. The only restriction that must be considered is that each measurement is taken when the luminescent thermometer is in thermal equilibrium with the thermal probe of reference. As stated by the British Standards Institution, an instrument can be considered as repeatable, in our case a luminescent thermometer, when the deviation relative to the average measured temperature is smaller than two times the standard deviation of the data. Experimentally, the repeatability of a thermal probe can be estimated by inducing controlled heating and cooling cycles during which the temperature is measured with this thermal probe, and this measurement is compared to the measurement taken with an independent thermal probe of reference. Under these conditions, the repeatability of a luminescent thermometer can be determined as [14]: | | max |Δ − Δi | · 100 (1.28) R =1− Δ where Δ is the mean thermometric parameter value extracted from the calibration curve, and Δi is the value of each independent measurement of the thermometric parameter. Note that the maximum absolute deviation value in the measurement of the thermometric parameter must be higher than δT , if not the calibration procedure, and/or δT are not correct. Figure 1.22 shows an example of how repeatability can be determined experimentally [90]. In that case, the fluorescence intensity ratio of the emission intensity of the two bands generated by Tm,Yb:GdVO4 nanoparticles coated with a shell of SiO2 , located at 700 and 800 nm, was calculated in 10-heating–cooling temperature cycles, allowing to demonstrate a repeatability higher than 99%. The reproducibility measures the fluctuation of the value of temperature determined from the thermometric parameter, when recorded under different experimental conditions. These different experimental conditions can be generated by using different measurement equipment, different luminescence thermometry methods, performed by different observers, changes in the pH or ionic strength of the environment surrounding our thermal probe, etc. [14]. Thus, reproducibility represents the ability of the thermometer to reproduce the same results, even when different detectors are used, or the temperature measurements are made on different days. Normally this parameter is determined from a statistical analysis. By doing so, it

42

J. J. Carvajal and M. C. Pujol

Fig. 1.22 FIR recorded in 10 heating–cooling temperature cycles based on the emissions at 700 and 800 nm of Tm,Yb:GdVO4 nanoparticles coated with a shell of SiO2 , allowing to determine a repeatability of 99%. Adapted with permission from Ref. [90]

can be determined if different calibration procedures are significantly different, and thus would lead to different temperature values. As a result, if the same calibration curve can be obtained in different measurements, within the resulting experimental uncertainty of the fitting parameters, then it can be concluded that the luminescent thermometer produces reproducible readouts under the tested conditions. Nevertheless, this parameter is rarely reported in the literature presenting new luminescent thermometers.

4.5 Temperature Operation Range of a Luminescent Thermometer Whenever the performance of a luminescent thermometer is analyzed, by determining any of the parameters presented in the previous sections, it must be considered that they will only be valid in the range of temperatures in which the luminescent thermometer has been tested. Thus, extrapolation of the performance parameters of the luminescent thermometer outside of this range would lead to wrong conclusions, since it is not known how the thermal probe would behave in those regions. Nevertheless, the range in which measurements of temperature can be performed with a particular luminescent thermometer can be determined. In fact, it is governed by the properties of the host material in which the luminescent emitter is embedded and the luminescent emitter itself. It should be considered that if temperature measurements are performed close to the limits of this range, the measurement uncertainties will increase considerably. The low temperature limit will be determined by the minimum temperature able to populate the higher energy state involved in the electronic transitions from which the thermometric parameter is calculated. So, by taking as example the lanthanide ions as luminescent emitters operating under the FIR (LIR) method, Dramicanin proposed

Introduction to Luminescence Thermometry

43

to determine the low temperature limit by isolating the temperature in Eq. 1.2, and considering the minimum value of the FIR parameter [27], and calculate it through the following expression: Tmin =

ΔE [ ⏋ k B log(B) − log(F I Rmin )

(1.29)

According to Eq. 1.29, the low temperature limit is directly proportional to ΔE and inversely proportional to B, that depends on the spontaneous emission rates of the different electronic levels involved in the transitions generating the emissions from which FIR is calculated, their frequencies and their degeneracies. Thus, small ΔE values and large B values would allow reducing the low temperature limit of the luminescent thermometer. Remember that ΔE is the difference of energy between the positions of the thermally linked electronic levels of the lanthanide, with values in the range 200–2000 cm−1 . So, to reduce ΔE, , lanthanide ions with electronic levels that do not differ too much in energy should be used, like those arising from the Stark splitting of the electronic levels [88, 89]. Since B is more related to the host in which the lanthanide ions are embedded, then materials design and engineering can be used to increase the value of this parameter. Figure 1.23 shows the dependence on ΔE and B of the low temperature limit of a luminescent thermometer operating under the FIR method with a hypothetical FIRmin value of 0.01, calculated according to Eq. 1.29. From its side, the upper temperature limit at which a luminescent thermometer can operate is determined by the phonon spectrum of the host material and the type of

Fig. 1.23 Dependence of the low temperature limit of a luminescent thermometer operating under the fluorescence/luminescence intensity ratio method on ΔE for different values of B supposing a detection limit of 0.01 for FIR/LIR. Reproduced with permission from Ref. [27]

44

J. J. Carvajal and M. C. Pujol

luminescent emitter used. Again, for the case of lanthanide ions, it can be concluded that in general, higher upper temperature limits can be achieved when material hosts with low phonon energy values are used, combined with active lanthanide ions involving thermally coupled electronic levels with large ΔE. However, there are other factors affecting this upper temperature limit that must be considered, like for instance any increase in the blackbody radiation arising from the material itself at very high temperatures. This blackbody radiation might interfere with the emissions generated by the luminescent thermometer, and thus the ratio of signals determined from them would not correspond to the pretended FIR, since it would include part of the blackbody radiation. Despite it would be interesting when characterizing a new luminescent thermometer to report their temperature operation range, this is a parameter that normally is not reported in the literature.

5 Calibration of a Luminescent Thermometer Some of the mechanisms involved in the development of a luminescent thermometer might allow to generate a primary luminescence thermometer, as it is reviewed in more detail in Chap. 3. These primary thermometers are based on temperature determination through the knowledge of a thermodynamic law [15]. This situation contrasts with that of thermometers in which the temperature is determined after comparison with a reference thermal probe, called secondary thermometers. In such thermometers a calibration procedure is required against an external temperature measurement reference, such as that taken with a thermocouple, a pyrometer or an infrared thermal camera, for instance, to compare the temperature of the two thermal probes. Secondary thermometers are the thermometers that are normally used in our daily life. Those are based, for instance, on the calibration of the electrical resistivity of a metallic wire that depends on temperature, such as in thermistors like platinum resistance thermometers; or the variation of the volume of a liquid or a gas, like mercury or alcohol, in a glass tube as a function of temperature, as it happens in the classical mercury-in-glass thermometers. Other secondary thermometers are those based on the production of a temperature-dependent voltage as a result of the thermoelectric effect, as it happens in thermocouples; or the temperature-dependent forward voltage of a silicon diode, used in electronic equipment, to name a few [7]. These thermometers are used whenever the physical properties or quantities that are chosen to be measured to determine temperature cannot be correlated to any equation of state. In these cases, the temperature must be determined by correlating it with other measurable physical quantities, and comparing the temperature measured by this thermometer with that measured with another thermal probe for which the temperature calibration is well known. In secondary thermometers, recurrent calibrations are required, especially when the thermometers are used in a

Introduction to Luminescence Thermometry

45

different media than the one in which they were calibrated. Of course, this constitutes a time-consuming and tedious task that is not always possible to execute. For instance, performing the calibration of the thermometer inside living cells or inside living animals is almost impossible. In fact, in most of the secondary thermometers it is assumed that a unique calibration relation is valid, independently of the medium, a procedure that is potentially inaccurate. This is especially true whenever the thermometer is located in media with altered ionic strength, pH, pressure, ions in a local neighborhoods, or atmosphere compositions that might disturb the response of the thermometric parameter, making no longer valid the calibration relation previously determined. This ad hoc assumption is a fundamental bottleneck of secondary thermometers. All the luminescent thermometers presented in this chapter are considered to be secondary thermometers, and thus, for all of them, a calibration procedure is required to be performed prior to their use to determine temperature in a specific medium. For that, the more common practice is to change the temperature in a controlled manner, measure it with a reference thermometer, like a thermocouple or an IR camera, and associate the spectrum recorded for the luminescent thermometer to this particular temperature. If this procedure is repeated for several known values of temperature, then a function correlating the thermometric parameter calculated from the information associated to each respective spectra and the temperature measured with the independent thermal probe can be extracted. This mathematical expression would allow that afterwards, by recording a spectrum in the temperature range for which the calibration has been performed, and determining from it the corresponding thermometric parameter, the temperature can be determined by introducing the value of the thermometric parameter in the calibration equation.

6 Materials in Luminescence Thermometry Different materials have been used as luminescent thermometers, including quantum dots, organic dyes, metallic nanoparticles, luminescent polymers, nanodiamonds, carbon dots, fluorescent proteins, DNA and lanthanide-doped dielectric materials. Depending on the application envisaged for a particular luminescent thermometer, different requirements are needed to fulfill. For instance, the materials requirements for the development of a luminescent thermometer to be used in biomedical applications are exhaustive. They include materials exhibiting a strong luminescence and good photostability, materials that are biocompatible, with sizes that allow clearance from the organism at the right time. It is important to ensure the overall complete clearance of the bioprobes to avoid potential long-term toxic effects, but at the same time, if this clearance happens too rapidly, this will hinder the probe’s accumulation in the organ or the tissue under observation and would difficult the determination of the temperature. The potential luminescent thermometer for biomedical applications must possess also a good dispersibility in biological media, exhibit a good chemical stability, and when required, show specificity and selectivity towards targeted

46

J. J. Carvajal and M. C. Pujol

tissues or cells. Furthermore, they should have suitable optical properties that allow for deep-tissue penetration of the applied excitation and emission wavelengths, while the excitation radiation does not generate autofluorescence in the biological samples [41]. However, most of them need to be excited using ultraviolet (UV) or visible light. Since UV and visible light show a low penetration depth in biological tissues, as will be discussed more in detail in Chap. 6, this limits their use in biological applications. Furthermore, especially when using UV light, autofluorescence from biological tissues will be generated. This will also affect the detection of the signal emitted by the thermal probes, limiting their performance [86]. This is just an example of the materials requirements for a particular application. For other applications, other considerations will have to be taken into account. In this section we will give a general overview to the materials used in luminescence thermometry.

6.1 Lanthanide-Doped Luminescent Thermometers Lanthanide ions are regarded as luminescent ions weakly influenced by the environmental conditions. This property is attributed to the fact that the 4f electrons responsible for their optical absorption and emission activity are electrostatically shielded by the fully occupied 5d orbitals [42]. However, their luminescence properties depend on temperature. The mechanisms at the basis of these temperature modifications can be classified into: • Temperature-induced population redistribution: many lanthanide ions present electronic states very close in energy in such a way that they are thermally coupled. The population distribution in thermally coupled electronic states strongly depends on temperature, and it can be considered to be governed by Boltzmann statistics. So that, any slight modification in temperature causes relevant modifications in the electronic population of the thermally coupled states. As the emitted luminescence intensity generated from an electronic state is proportional to its electronic population, temperature variations can be monitored by analyzing the relative emission intensities generated by the thermally coupled states, thus allowing for measuring temperature using a ratiometric method, such the FIR (LIR) method. • Temperature-induced variation of energy transfers probabilities: when a given material is co-doped with two or more different lanthanide ions, energy transfer (ET) becomes possible [101]. When the material is optically excited at a wavelength absorbed by one of these lanthanide ions (donor), this could partially transfer the absorbed energy to another lanthanide ion (acceptor), which would release energy in the form of luminescence. The probability of the ET process depends on the concentration of lanthanide ions, their distance inside the host in which they are embedded, and on the spectral overlap between the donor emission and the acceptor absorption bands. This spectral overlap, in turn, depends

Introduction to Luminescence Thermometry

47

on the shape of these absorption and emission bands and on the phonon energy and density in phonon-assisted ET processes. Both the line shape and the phonon density are temperature-dependent parameters, so ET probability to occur between lanthanide ions in a material is a temperature dependent process. Therefore, the luminescence of a material co-doped with different lanthanide ions would also be temperature dependent. In ET-assisted luminescence thermometry, the temperature readout might be achieved by monitoring the intensity ratio between donor and acceptor emissions, among other possibilities. Previous works have demonstrated that the ratiometric thermal sensitivity of lanthanide doped nanoparticles is based either on the temperature dependence of energy transfer efficiency between different lanthanide ions or on the different temperature quenching rate of the intensities of the different luminescent emissions generated by the lanthanide ions [110]. Another interesting case involving FIR (LIR) when using lanthanide ions is that of the up-conversion (UC) luminescent nanoparticles. In such nanoparticles light at shorter wavelengths is generated despite pumping is done at longer wavelengths. Chapters 2, 3, 4 and 5 shows a good bunch of examples of the development of luminescent thermometers using lanthanide-doped ions and their potential applications. Thus, the reader is asked to read these chapters to get more details on how these materials can be used for luminescent thermometry practices.

6.2 Metal–Organic Frameworks (MOFS) Metal–organic frameworks (MOFs) are solid crystalline materials formed by metal ions or clusters coordinated to organic ligands. One of the principal characteristics of this family of materials is that they contain voids that provide a porous structure. MOFs have been investigated traditionally for the storage of gases, in applications in catalysis and as supercapacitors [30]. In the context of luminescence thermometry, they have been explored mainly when the central metal is a lanthanide ion [14]. Chapter 5 provides a deep analysis of the luminescence thermometry potentiality of this kind of materials. The thermometric mechanism mainly used in MOFs is based on an energy transfer process between ions within the polymeric framework that makes the emission from the metal ion to be sensitized by the organic ligands. The main advantage of these luminescent thermometers when compared to molecular thermometers is that their ordered structure provides a highly reproducible thermal sensing response. However, their working range is determined by the triplet energy state of the organic ligands, which corresponds to the range from 10 to 330 K. Also, it has been observed that their S rel are normally smaller than those obtained with Ln3+ -doped nanoparticles.

48

J. J. Carvajal and M. C. Pujol

6.3 Quantum Dots as Luminescent Thermometers Quantum dots (QDs) are semiconductor materials characterized by the quantum confinement of the electrical charge carriers (electrons and holes) in the three dimensions of space due to the reduced sizes of these particles. This happens when the particle size is equal to or smaller than the exciton (electron–hole pairs attracted by Coulomb interaction) size in the bulk crystal or the exciton Bohr radius [5]. The exciton Bohr radius depends on the effective masses of electrons and holes, the effective mass of the excitons, the Bohr radius of the elements that form the quantum dot, and the dielectric constant of the semiconductor. Thus, this exciton Bohr radius, and thus the size of the nanoparticles below which these quantum confinement effects can be visualized, is different for each semiconductor. This quantum confinement effect leads to different electrical, electronic and optical properties than those expected for bulk semiconductors, which can be controlled to a certain degree by a structural and compositional engineering design. This is because when charge carriers are spatially confined, their energy states become restricted, leading to discrete energy levels, whose energy positions, and thus their separation, depend on the size of the quantum dots, i.e., on the size of the spatial confinement. In fact, the separation between these energy levels increases as the size of the quantum dots decreases, because the degree of confinement also increases. A main consequence of this quantum confinement effect is that the luminescence properties of semiconductor materials are strongly modified in the quantum dots, in such a way that the color of their emissions depends on their size [71]. So, for instance, such effects become observable in CdS nanoparticles with sizes around 5 nm, but they can be observed in PbSe quantum dots for sizes as big as 90 nm [46]. But for a particular particle size, the luminescence properties of the quantum dots, including luminescence intensity, luminescence decay time, emission peak position and emission bandwidth, depend strongly on the temperature, making it possible to use them as highly sensitive luminescent thermometers [58, 68, 103]. This is due to several temperature-related effects, such as the thermal expansion of the crystalline lattice, temperature-induced changes in the confinement energy, and temperatureinduced mechanical strength and electron–phonon coupling [46]. The shift of the position of the emission peak of QDs with temperature is mostly related to the variation of the position of the first exciton absorption peak. This again depends on the size of the QDs and is affected by the thermal expansion of the lattice structure of the semiconductor that in its turn affects the exciton energy, since it implies a lesser confinement. Another important factor that affects the shift of the emission peak position is the effect of temperature on the position of the energy levels in these quantum confined structures again due to the thermal expansion of the lattice structure. This is a result of the temperature-induced changes in the electron–phonon coupling arising from the non-discrete nature of the energy levels of the QDs, which are broadened and shifted as a result of the lattice vibrations. Another important parameter to consider when dealing with QDs is the emission efficiency. QDs with large emission efficiencies, known as bright QDs, are especially

Introduction to Luminescence Thermometry

49

suitable for luminescence thermometry. This emission efficiency, or luminescence efficiency, is given by the quantum yield, that can be defined as the ratio of the number of photons emitted to the number of the photons absorbed per unit time. The quantum yield depends also on the sizes of the QDs [66] and is the result of an equilibrium between the radiative and non-radiative deexcitation rates from the excitonic excited state. Bright QDs are achieved when non-radiative deexcitation rates are minimal. Temperature is also responsible for an important decrease of the intensity of the luminescence generated by QDs. This reduction in intensity cannot be only ascribed to a pure thermal quenching mechanism, due to photoionization of the charge carriers that are activated as the temperature increases, so that they can be more delocalized along the material and reach a nearby non-fluorescence state that acts as a trap. In QDs, surface states, generated by defects and/or impurities encountered in the immediate surroundings of the quantum dots, act usually as trap quencher states. Consequently, the environment in which the quantum dot is embedded has a tremendous influence on the quenching of the luminescence due to thermal effects. Therefore, for practical applications, and to avoid these thermal quenching effects, most of the quantum dots used, especially in biological and biomedical applications, have a core–shell structure, as can be seen in Chaps. 6, 7, 8 and 9. However, the formation of these core–shell structures might also have some adverse consequences for luminescence thermometry purposes. It has been shown that when this class of quantum dots are heated above a certain temperature, the thermal quenching is irreversible, and the emission from the QDs cannot be recovered. This is due to the formation of thermally induced structural changes, like defects induced by a mismatch between the materials that constitute the core and the shell, that create permanent trap states. Thus, the thermal quenching processes depend to a great extent on the particular QD structure (core, core–shell, core-multishell) and on the particular environment where the dot structures are hosted. Thus, when using QDs as luminescent thermometers operating under the mechanism of variation of the intensity of a single emission band, the calibration curve might be modified during sensing due to undesirable changes in the QD environment. The average emission lifetime decay of small QDs is also sensitive to changes in temperature in their environment, related to the thermal quenching induced by their environmental solvent and surface coating [40], constituting another interrogation parameter that can be used for luminescence thermometry purposes with these materials. Because of this combination of outstanding properties, QDs have emerged as reliable luminescent thermometers, providing high-resolution thermal readings. QDs have been used as luminescent thermometers to measure the internal temperature in living cells, and in ex-vivo and in-vivo experiments (see Chaps. 6, 7, 8 and 9). They have also been used in the thermal characterization of microelectronic devices, to determine the temperature generated by an electrical microheater in its surroundings [58]. Single QDs can be used as luminescent thermometers from the spectral analysis of the luminescence generated by a single QD or by a collection of them. This is a valid approximation when temperature is extracted from the position of the emission

50

J. J. Carvajal and M. C. Pujol

peak or when the luminescence decay lifetime is used for the same purpose, as they give, after proper calibration, an absolute temperature value without requiring any additional reference. However, when temperature readings are obtained from the thermally induced luminescence quenching, temporal and spatial variations of the concentration of QDs could cause intensity variations not related to temperature, and so leading to erroneous measurements and interpretation. To mitigate these limitations and error sources related to intensity measurements, a luminescence reference should be used simultaneously, so that these intensity fluctuations can be corrected. In this way, the variation of the intensity ratio between the emission generated by the QDs and that generated by the luminescence reference would provide again an absolute temperature value after proper calibration. This could be achieved by developing complex luminescent nanostructures composed of a QD and another luminescent material that would be used for reference purposes. For instance, Albers et al. used red emitting CdSe QDs embedded in CdS quantum rods, generating a semiconductor heterostructure passivated with an amphiphilic polymer shell, which was appended with far-red-emitting cyanine dyes [2], as can be seen in Fig. 1.24. In that case, the intensity of the emission of the quantum dot-quantum rod heterostructure, located at around 620 nm, decreases substantially as the temperature increases, and its position changes. Instead, cyanine dyes show a temperature-dependent luminescence quantum yield but not wavelength shift of the band located at around 670 nm was produced. When both are put together in the same structure, FRET resonant energy transfers processes [33], are activated. Consequently, the overall luminescence generated by this heterostructure consisted of both the emission band at 620 nm generated by the quantum dot-quantum rod core–shell structure and the emission band at 670 nm generated by the dye. When the temperature increased above 293 K (20 °C), the intensity of both bands decreased monotonically. Due to the different mechanisms involved in each case, the intensity ratio between these two emission bands varies monotonically with temperature (see Fig. 1.24), exhibiting a pseudolinear response with a fully reversible behavior. A relative thermal sensitivity of 2.4% K−1 could be achieved with this structure, which allowed obtaining a temperature resolution of 0.2 K. These luminescent thermometers were used to determine the temperature inside HeLa cells. In another approach, Lee et al. proposed to use QDs interconnected with metallic nanoparticles as luminescent thermometers through a polymer that acted as a molecular spring [56]. The nanosized heterostructure consisted of a core Au nanosphere covered by a poly(ethylene glycol) (PEG) film with a thickness of a few nanometers, conjugated to CdTe QDs with a size of 3.7 nm. The Au nanosphere showed a surface plasmon resonance at around 633 nm, which activated the emission of the QDs at 550 nm when the heterostructure was illuminated at the former wavelength. The efficiency of this plasmon-excitation energy transfer depends strongly on the distance between the Au surface and the QDs. This distance, that depended on the thickness of the PEG film, was modulated as a function of temperature, since this polymer expands drastically in the 293 K (20 °C)–323 K (50 °C) temperature range. This generated a decrease of the intensity of the luminescence generated by the QDs as the temperature increased.

Introduction to Luminescence Thermometry

51

Fig. 1.24 Heterostructures involving quantum dots used to improve their performance as luminescent thermometers. (left) Schematic representation of a dual-emitting hybrid luminescence nanothermometer consisting of a CdSe QD embedded in a CdS quantum rod, covered at its time by a temperature-responsive dye in a sort of core–shell-shell structure. (right) Luminescence spectra generated by this heterostructure, obtained at different temperatures featuring the emission of the CdSe@CdS quantum dot-quantum rod structure, located at ~ 620 nm, and that of the cyanine dye, located at ~670 nm. Reproduced with permission from Ref. [2]

6.4 Organic Dyes Organic dyes are aromatic organic compounds, which give color to a substrate in which they are supported by a process of selective absorption of light [26]. These compounds exhibit strong luminescence in the visible after being excited, typically with UV or blue light. The emission and absorption properties of organic dyes depend on their structure and the chemical environment in which they are encountered [118]. In many organic dyes their luminescence emission properties depend also on temperature [93]. These characteristics, combined with the short emission lifetimes they exhibit that allows getting a high temporal resolution, the high spatial resolution that can be achieved with them due to their relatively small sizes (smaller than nanoparticles), and the fact that most of them are biocompatible, make of organic dyes excellent candidates to develop luminescent thermometers. In fact, these compounds are among the earliest materials used for luminescence thermometry in in vitro applications. The luminescence intensity and the emission lifetimes of organic dyes are parameters affected, normally, when the temperature changes [28, 29, 91]. However, only some organic dyes can be used for luminescence thermometry purposes. The reasons for these limitations include [113, 114]: • The luminescence spectra of the organic dyes should undergo a significate change with temperature to provide a good thermal sensitivity. • Their photostability should be good enough to ensure temperature determination precision. • Organic dyes should not be toxic, especially if in vitro and in vivo uses are intended.

52

J. J. Carvajal and M. C. Pujol

• The luminescence quantum yields exhibited by the organic dyes should be high enough in the temperature range in which they want to be used. Several families of organic dyes have been tested as potential luminescent thermometers, including rhodamines, fluorescein, pyranine, 7-nitrobenz2-oxa-1,3-diazol-4-yl (NBD), 6-dodecanoyl-2-dimethylamino-naphthalene (Laurdan), triarylboron compounds, byspyren dyes, acridine yellow, 2,5dihexyloxy-4-bromobenzaldehyde (Br6A), perylene/N-allyl-N-methylaniline, and bis/benzoxazolyl) stilbene (BBS), some of them even embedded in polymers to be used as solid thermometers. The luminescence of rhodamine family dyes, like Rhodamine B, Rhodamine 101 and TAMRA (5-carboxytetramethylrhodamine), who exhibit high quantum yields, has been known to be temperature sensitive for a long time. However, as the rest of the organic dyes, they suffer from photobleaching, that is the photochemical alteration of the dye such that it is permanently unable to generate luminescence. It can be caused by the cleavage of the covalent bonds in the dye caused by transitions from a singlet state to the triplet state of the fluorophore, or by non-specific reactions between the dye and the surrounding molecules. Photobleaching can be solved by using a high-intensity light source before taking the measurements of temperature, but this of course, might damage living cells or living organisms, if these organic dyes are used as thermal probes in in vitro and in vivo experiments. Since organic dyes generate only an emission band, luminescent thermometers based on them work under the intensity-based luminescence thermometry method, which is affected by fluctuations in the excitation source, not allowing the absolute determination of the temperature. To solve these problems, a reference thermal probe can be used together with organic dyes, as pointed out before in the case of the QDs. This reference thermal probe in that case can be another organic dye. In that case, the intensity ratio that can be calculated, being also a function of the temperature, would not be affected by the fluctuations in the excitation source and would improve the precision of these luminescent thermometers. Thermal sensitivities as high as 2.7% K−1 can be achieved with such organic dyes, which resulted in temperature resolutions below 0.5 K [69]. However, an inhomogeneous dye loading and a different photostability between the two dyes might result in imprecision and inaccuracy. Also, luminescence lifetime decay in a time scale of nanoseconds can be used in some cases, with thermal sensitivities around 2% K−1 [47]. For that, rhodaminelabelled DNA oligomers were used, in which the rhodamine-G dye was separated from the main chain by a sequence of 6 methylene units. The main idea of using dye-labelled DNA as a luminescent thermometer is that DNA bases quench the luminescence generated by the dye with variable efficiency. While adenine did not quench the luminescence of the dye, cytosine, guanine and thymine do in an increasing order. Therefore, the spacing between the organic dye and the designed sequence of DNA bases can be modulated by conformation changes of the DNA chain, which allowed to modulate the possibility of the dye molecules to generate luminescence. With this quenching mechanism, the higher the temperature, the faster the conformational changes of the DNA chain, and the shorter the luminescence decay time of the

Introduction to Luminescence Thermometry

53

Fig. 1.25 Luminescence intensity plotted in a logarithmic scale against time for a rhodamine-G dye attached to a DNA oligomer at 2 μM concentration in deionized water at various temperatures (squares 15 °C, circles 20 °C, up-side triangles 25 °C, down-side triangles 30°C and diamonds 35 °C). The inset shows the average luminescent lifetime constant as a function of temperature. Reproduced with permission from Ref. [47]

emission arising from the organic dye. Figure 1.25 shows the decay of the luminescence intensity with time for the rhodamine-labelled DNA oligomers at different temperatures, illustrating this decrease of the lifetime as the temperature increases. The fluorescein family of dyes are also highly efficient luminescent probes with a luminescence quantum yield of almost 100%. As for rhodamines, the luminescence intensity of fluorescein decreases as the temperature increases. They also have good long-term stability and a fast equilibrium response. NBD and Laurdan, from their side, are twisted intramolecular charge transfer (TICT) compounds, that usually exhibit conformation changes in different excited states [113, 114] that can be accessed through different temperatures. These kinds of compounds can maintain their total luminescence intensity, or even enhance it from lower to higher temperatures. Furthermore, if the excited states are emissive, a luminescence color change might be observed when the temperature changes. In fact, the quenching of the luminescence with temperature of these compounds has been used in intensity-based luminescence thermometry. In the case of NBD the changes observed in the intensity of the emission are due to electronic changes induced by an increase of the temperature leading to a TICT state, together also with a decrease in the luminescence lifetime decay [31]. Despite luminescence thermometry with organic dyes has been used to determine the temperature in living cells, especially on the cell membrane, and in in vivo experiments in rat tendon and brain samples, due to their liquid nature, their main application has been in microfluidics [34, 50, 69]. Figure 1.26 shows the temperature maps obtained when mixing hot and cold streams at the intersection of a Ychannel fabricated in a PDMS/glass microfluidics chip using rhodamine B as the

54

J. J. Carvajal and M. C. Pujol

Fig. 1.26 Scheme of the experimental setup used to map the temperature generated when mixing hot and cold streams at the intersection of a Y-channel fabricated in a PDMS/glass microfluidics chip using rhodamine B as the luminescent thermometer. The two temperature maps were recorded 5 min apart by measuring the emission intensity of the organic dye. Reproduced with permission from Ref. [34]

organic dye to determine temperature using the intensity-based luminescence thermometry technique. They have also been used for mapping the temperature generated in micro/nanowires [63], or to measure the temperature of an imprint resist film [52]. Organic dyes, as well as polymers, however, generally display poor photostability and a pronounced cross-sensitivity to oxygen [2].

6.5 Luminescent Thermometers Based on Polymers Structural phase transitions in polymers can be exploited as a mechanism to enhance the temperature dependent luminescence response of solvatochromic dyes, since these dyes will change their luminescence properties in response to changes in the local environment [95]. Thus, such luminescent thermometers are formed by a thermoresponsive polymer that experiences a sharp phase transition at a particular temperature, and an organic dye whose luminescence is sensitive to this structural change and that can be correlated to the temperature. Thermoresponsive polymers can be encountered as hydrophilic molecularly dissolved structures or hydrophobic dehydrated collapsed globule structures depending on the temperature [84]. These

Introduction to Luminescence Thermometry

55

changes in the structure of the polymer, that also change the polarity of the environment, generate different absorption and/or emission properties in the organic dye that can be correlated to temperature [99]. The strong dehydration of the internal part of the collapsed polymeric globules produced during the phase transition induces a decrease in the polarity of the environment. This change in polarity, from its site, generates an uneven stabilization of the ground state or the excited state of the organic dye that changes their energetic position, and thus, changes their energy difference that can be observed by the shift of the positions of the absorption or emission bands of the dye. Also, changes in the intensity absorption and emission bands, and changes in the lifetime of the generated luminescence can be observed. Poly(N-isopropylacrylamide) (PNIPAM) with benzoxadiazole dye is an example of a luminescent thermometer based on this effect [98]. The decrease of the inter- and intradistances between the polymer backbone and the sidechains during the phase transition can be used to induce FRET with the dye. FRET here consists of a non-radiative energy transfer from an excited donor to a fluorophore acceptor. This fluorophore acceptor then emits this energy in the form of light. This non-radiative energy transfer is only effective at short distances and, thus, depends substantially on the distance between the two moieties. In this case, when the distances between the polymer backbone and the sidechains change due to the phase transition experienced by the polymer, FRET can be activated. A phenanthrene end-functionalized poly(N-decylacrylamide-b-N,N-diethylacrylamide) block co-polymer functionalized with anthracene is a luminescent thermometer that works under this principle by exciting phenanthrene and measuring the emission of anthracene [80]. From another side, the reduction of the mobility of the sidechains after the phase transition is another effect that can be used to enhance the luminescence changes of the dye and use it as a luminescent thermometer. It has been used in PNIPAM functionalized with tetraphenylethene dye. In this system, the precipitation of the polymer restricts the rotation of the phenyl rings around the central double bond, and this increases the intensity of the emission of the dye [96]. The formation of excimers, i.e., dimers formed that are already in an excited state, can also be used to generate a luminescent thermometer with a polymer. As an example, the temperature dependence on the intensity ratio of the luminescence of the monomer and the excimer in the ladder-like 1,4-phenylene-bridged polyvinylsiloxane (LPPVS) has been used for this purpose [112]. The ratio between the intensities of the excimer and the monomer of LPPVS increases linearly with the inverse of the temperature, and this change can be used as luminescent thermometer. However, the main disadvantage of the luminescent thermometers based on polymers is that they can only be used in a narrow range of temperatures, typically 10–20 K around their phase transitions. Despite these limitations, polymer-based luminescent thermometers have been used to measure the internal temperature in living cells [37]. Figure 1.27 shows a schematic representation of the structural change induced by temperature in a N-isopropylacrylamide (NIPAM) polymer containing a benzodiazol luminescent organic dye. Here it can be seen the change between the swollen and the shrunken states of the polymer, which induces a change in the luminescent properties of the organic dye, generating a strong luminescence in the shrunken state. At low

56

J. J. Carvajal and M. C. Pujol

temperatures, the polymer swells by absorbing water molecules into its interior. This quenches the emission of the water-sensitive organic dye. At higher temperatures, the structure of the polymer collapses with the release of water molecules, allowing the emission of the organic dye. The change of the intensity of the emission band of the organic dye, located at 560 nm, is highly dependent on the temperature, as can be also see in the figure, while this response is independent of the pH. Nevertheless, these kind of luminescent thermometers have also been used to determine simultaneously temperature and pH [78], or to sense at the same time temperature and metallic ions for pollution environment analysis [79].

Fig. 1.27 Example of a luminescent thermometer based on a thermoresponsive polymer associated to an organic dye. a Schematic diagram and chemical structures of the components at low and high temperatures. b Intensity of the emission band of the organic dye as a function of temperature, concentration of KCl, and c pH. Reproduced with permission from Ref. [37]

Introduction to Luminescence Thermometry

57

6.6 Transition Metal-Doped Nanoparticles The luminescence generated by transition metal emitting centers is highly temperature dependent. In fact, it is more severely quenched by an increase of temperature than the emissions generated by lanthanide ions. Transition metal-based luminescent thermometers were developed before than lanthanide-based luminescent thermometers. Early applications of this kind of thermometers are the cryobaric diamondanvil cells that used ruby (Cr3+ -doped Al2 O3 [108] or thermographic phosphors that used Cr3+ in YAl3 (BO3 )4 [10]. Cr3+ ions possess the R lines that are two emission lines in the red region of the visible spectrum generated by the electronic transitions between the 2 E and 1 E levels. The intensity ratio between these two lines has been used for luminescence thermometry purposes. Similarly, Mn4+ in Mg4 FGeO6 , that also exhibits two emission bands in the red region of the visible spectrum, has been used to determine the temperature by luminescence [74]. Other Cr3+ -doped materials, like Cr3+ :Bi2 Ga4 O9 mullite powders that generate emissions in the nearinfrared part of the electromagnetic spectrum have also been used as luminescent thermometers [4]. These transition metal ions form 3d transition ionic centers when incorporated in solids that can generate several emission lines originating from the electronic transitions between the multiple energy levels created by the incompletely filled d shells. 3d electrons are highly influenced by the surrounding ions and their spectroscopic behavior is a consequence of the dn electronic configuration and the crystal field potential created by the surrounding ligands [27]. However, transition metal based luminescent thermometers suffer from providing low values of relative thermal sensitivity. In some cases, it is also difficult to discriminate the different R lines generated by these materials, especially when they form broad emission bands [27]. To better resolve the emission lines arising from transition metal ions, and at the same time increase the relative thermal sensitivity of these luminescent thermometers, materials with diverse emitting centers have been developed. For instance, Mn2+ :Zn2 SiO4 particles containing defects in the host acting as traps, is one example of these materials. From one side they have the green emission generated by Mn2+ that is highly sensitive to temperature. From another side, they exhibit a blue emission generated by the artificially inserted traps in the host that is almost insensitive to temperature [60]. They can also be combined with Ln3+ ions, so that again two different emitting centers are present in the luminescent thermometer, like the highly temperature-sensitive emissions of Mn4+ or Cr3+ and those originating from Eu3+ , Tb3+ or Dy3+ that are almost insensitive to temperature changes [19]. Figure 1.28 shows the emission spectra of Mn4+ , Eu3+ :YAG powders at different temperatures. It can be seen how the intensity of the emission bands of Mn4+ decrease fast as the temperature increases, while the emission bands of Eu3+ are almost temperature insensitive, which give to these luminescent thermometers relative thermal sensitivities as high as 4.81% K−1 . Transition metal ions can also be used as sensitizers, i.e., ions that absorb the energy of the excitation source and transfer it to the emitting ions, to lanthanide

58

J. J. Carvajal and M. C. Pujol

Fig. 1.28 (left) Temperature-dependent emission spectra of Mn4+ , Eu3+ :YAG powders. (right) Variation of the intensity of the emissions of Mn4+ and Eu3+ in YAG as the temperature increases. Adapted with permission from Ref. [19]

ions for instance. More details about this particular mechanism of luminescent thermometers containing transition metal ions can be found in Chap. 2.

6.7 Metal Nanoparticles Metal nanoparticles, especially gold nanoparticles that exhibit a good biocompatibility and are relatively easy to conjugate with biomolecules, have interesting luminescent properties associated to the surface plasmon resonance. This made of these metal nanoparticles preferred fluorescent probes for in vivo and in vitro imaging [62]. Also, since they present a high photothermal conversion efficiency, i.e., the possibility of converting part of the energy of the light absorbed into heat, they have been the class of materials more intensively investigated as potential photothermal converters to be used in photothermal therapy [35]. Furthermore, since their luminescence properties are influenced by temperature because of effective nonradiative recombination of electrons and holes at higher temperatures, they can be used as luminescent nanothermometers [9]. As an example, luminescent gold nanorods with diameters of 1–2 nm with a luminescence quantum yield of 16.6% at room temperature could be used to determine temperature, since their quantum yield increased up to 28.6% when the temperature was decreased to −7 °C. In fact, as shown in Fig. 1.29, the photoluminescence decay curves of these gold nanorods show that their lifetimes decay faster at higher temperatures than at lower temperatures, which could be used as the thermometric parameter to develop luminescent thermometers from these materials. However, the luminescence properties of metal nanoparticles are also sensible to other local environmental parameters, such as oxygen content and pH, which might induce errors in the determination of temperature. The combination of metal nanoparticles with other luminescent systems can alleviate this problem [106].

Introduction to Luminescence Thermometry

59

Fig. 1.29 (left) Photography of a 0.5 mg/mL of Au nanorods solution in chloroform illuminated with UV light (366 nm). The surface of the Au nanorods was functionalized with 1tedradecylmercaptane. (left) Photoluminescence decay curves of the Au nanorods functionalized with 1-octadecylmercaptane at different temperatures. Adapted with permission from Ref. [9]

6.8 Carbon-Based Materials: Nanodiamonds and Carbon Dots Carbon-based materials, such as nanodiamonds and carbon dots, have also been used as luminescent nanothermometers. Nanodiamonds present a temperature-dependent lattice strain that produces important changes in the spin properties of nitrogen vacancies, that conform color centers in these structures [53]. This, in turn, generates variations in the emission properties of these color centers, which allowed to determine temperature variations of the order of 1.8 mK, the smallest ever reported for a luminescent thermometer. Nanodiamonds are chemically inert and have excellent thermal conductivity. These properties make, from one side all the nitrogen vacancies within a nanodiamond to be in thermal equilibrium with the environment, and allow, from the other side, to use these luminescent thermometers in a wide temperature range, extending from 200 to 600 K. However, the set up used to excite the emission of the nitrogen vacancies in a controlled way is very complex, including microwave pulses together with laser excitation. Nevertheless, an example of how these materials can be used as luminescent thermometers can be seen in Fig. 1.15. Another type of carbon-based materials that can be used as luminescent thermometers are carbon quantum dots (CQDs). These CQDs are characterized by having broad-band absorption, strong luminescence, good resistance to photobleaching, high chemical stability, low toxicity and good biocompatibility. All these properties convert CQDs into a very attractive choice for optical and photoacoustic imaging, photothermal and photodynamic therapies, drug delivery and biosensing [51]. These properties also convert CQDs in good candidates to develop luminescent thermometers operating under the intensity based and the intensity ratio luminescence thermometry techniques [18, 107, 113, 114]. Despite the mechanism of generation of the luminescence emission properties of these CQDs is not well understood, it is

60

J. J. Carvajal and M. C. Pujol

Fig. 1.30 Examples of the different thermometric parameters that can be used in carbon quantum dots to develop luminescent thermometers with them. a Variation of the intensity of the luminescence spectra of SiC dots as a function of the temperature. The inset shows the linearity of the thermal quenching of luminescence in an Arrhenius plot. Adapted with permission from Ref. [18]. b Photoluminescence decay time of CQDs doped with N and S recorded at different temperatures inside HeLa cells. Adapted with permission from Ref. [51]

believed that they originate from π-π* transitions in the aromatic sp2 domains and the trapping of excited-state energy by surface states [22]. Also, the luminescence lifetime of these CQDs is heavily influenced by temperature and could be used as the thermometric parameter [51]. Figure 1.30 shows the potentiality of CQDs as luminescent thermometers.

6.9 Biomaterials Green fluorescent protein (GFP), a molecule that can be found in several living organisms like jellyfish, corals, sea anemones, zoanithids, copepods and lancelets, is a biomarker used in cell imaging due to its unique optical properties. This protein has also been used as luminescent thermometer by measuring its luminescence polarization anisotropy, that depends on temperature, as shown previously in Fig. 1.10 [23]. This luminescent thermometer allowed to discriminate subdegree temperature changes on HeLa cells and on the neurons of a Caenorhabditis elegans in a fast way [24]. However, to internalize GFP on such biological systems requires cellular transfection, a procedure that is difficult to achieve in certain primary cellular types. Also, the emission lifetime of these fluorescent proteins depends strongly on the pH, which might induce non accurate measurements of the temperature [2]. The DNA oligomer decorated with organic dyes presented in Fig. 1.25 can also be considered as a luminescent thermometer based on a biomaterial. Another DNAbased ratiometric luminescent thermometer was demonstrated using the fluorescence resonance energy transfer (FRET) mechanism between two organic dyes (6, carboxyfluorescein, FAM, and tetramethyl rhodamine, TAMRA) [109]. This system

Introduction to Luminescence Thermometry

61

Fig. 1.31 Illustration of the working principle of a DNA-based ratiometric luminescent thermometer consistent on a rigid DNA tetrahedron to which a thermally sensitive molecular beacon decorated with two organic dyes (FAM and TAMRA) was attached in one of the six edges of the DNA tetrahedron. Relaying on the temperature-responsive hairpin structure and the FRET mechanism, the ON and OFF states were reversibly achieved, corresponding to temperatures below and above the melting temperature. Reproduced with permission from Ref. [109]

was composed of a rigid DNA tetrahedron containing a thermally sensitive molecular beacon in one edge. The two organic dyes were attached at the two ends of the oligonucleotide strand. By increasing the temperature, the distance between the two organic dyes is modified, which induced an acceptor-to-donor luminescence intensity change. Using the luminescence intensity ratio between the acceptor (TAMRA) and the donor (FAM), a thermal resolution smaller than 0.5 K was achieved. Figure 1.31 shows the working principle of this luminescent thermometer.

7 Conclusions In this chapter, we reviewed the fundamentals of luminescence thermometry to help the reader easily understand the following chapters of this book. The different mechanisms in which luminescence thermometry is based, and the main indicators of the performance of a particular luminescent thermometer have been explained. Some of these performance indicators are normally found in the different publications in which new luminescent thermometers are presented, like absolute and relative thermal resolution and temperature resolution. However, in some cases, the definition of these parameters is taken inversely, and what is defined here as relative thermal sensitivity in some cases is labelled as the absolute thermal sensitivity. Thus, by giving these clear definitions, we hope that the readers can easily identify each

62

J. J. Carvajal and M. C. Pujol

performance parameter. Other performance parameters are scarcely reported in the literature. However, for a full comparison of the potentiality of the different luminescent thermometers, it is recommended to include such performance parameters in future publications. Finally, a review of the different materials used in luminescence thermometry is given. Some of these materials will appear in the following chapters, especially lanthanide-doped materials and quantum dots, since they are the most used classes of luminescent thermometers. For others, however, that have been less explored, or that will not appear in a great extend in the following chapters, some key examples of their use as luminescent thermometers are given, so that, the reader can have a clear image of the importance of the different materials to develop this class of temperature sensors.

References 1. L. Aiguoy, G. Tessier, M. Mortier, B. Charlot, Scanning thermal imaging of microelectronic circuits with a fluorescent nanoprobe. Appl. Phys. Lett. 87, 184105 (2005). https://doi.org/ 10.1063/1.2123384 2. A.E. Albers, E.M. Chan, P.M. McBride, C.M. Ajo-Franklin, B.E. Cohen, B.A. Helms, Dualemitting quantum dot/quantum rod-based nanothermometers with enhanced response and sensitivity in live cells. J. Am. Chem. Soc. 134, 9565–9568 (2012). https://doi.org/10.1021/ ja302290e 3. S.W. Allison, G.T. Gillies, Remote thermometry with thermographic phosphors: instrumentation and applications. Rev. Sci. Instrum. 68, 2615–5260 (1997). https://doi.org/10.1063/1. 1148174 4. M. Back, E. Trave, J. Ueda, S. Tanabe, Ratiometric optical thermometer based on dual nearinfrared emission in Cr3+ -doped bismuth-based gallate host. Chem. Mater. 28, 8347–8356 (2016). https://doi.org/10.1021/acs.chemmater.6b03625 5. L. Banyai, S.W. Koch, Semiconductor Quantum Dots (World Scientific Publishing, Singapore, 1993) 6. A. Benayas, E. Escuder, D. Jaque, High-resolution confocal fluorescence thermal imaging of tightly pumped microchip Nd:YAG laser ceramics. Appl. Phys. B 107, 697–701 (2012). https://doi.org/10.1007/s00340-012-4879-y 7. M. Bettinelli, L.D. Carlos, X. Liu, Lanthanide-doped upconversion nanoparticles. Phys. Today 68, 38–44 (2015). https://doi.org/10.1063/PT.3.2913 8. S.J. Blundell, K.M. Blundell, Concepts in Thermal Physics (Oxford University Press, UK, 2006) 9. J. Bomm, C. Gunter, J. Stumpe, Synthesis and optical characterization of thermosensitive, luminescent gold nanodots. J. Phys. Chem. C. 116, 81–85 (2012). https://doi.org/10.1021/jp2 06260r 10. S.M. Borisov, K. Gatterer, B. Bitschnau, I. Klimant, Preparation and characterization of Chromium (III)-activated Yttrium Aluminum Borate: a new thermographic phosphor for optical sensing and imaging at ambient temperatures. J. Phys. Chem. C 114, 9118–9124 (2010). https://doi.org/10.1021/jp1016467 11. L.C. Bradley, A temperature-sensitive phosphor used to measure surface temperatures in aerodynamics. Rev. Sci. Instrum. 24, 219–220 (1953). https://doi.org/10.1063/1.1770668 12. Britannica, The Editors of Encyclopaedia. “Temperature”. Encyclopedia Britannica, https:// www.britannica.com/science/temperature. Accessed 10 Dec 2022 13. C.D.S. Brites, P.P. Lima, N.J.O. Silva et al., Thermometry at the nanoscale. Nanoscale 4, 4799–4829 (2012). https://doi.org/10.1039/C2NR30663H

Introduction to Luminescence Thermometry

63

14. C.D.S. Brites, A. Millan, L.D. Carlos, Lanthanides in luminescent thermometry, ed. by J.C.G. Bunzli, V.K. Pecharsky, in Handbook on the Physics and Chemistry of Rare Earths, vol. 49 (Elsevier, 2016), pp. 339–427 15. C.D.S. Brites, S. Balabhadra, L.D. Carlos, Lanthanide-based thermometers: at the cuttingedge of luminescence thermometry. Adv. Opt. Mater. 7, 1801239 (2019). https://doi.org/10. 1002/adom.201801239 16. L.D. Carlos, R.A.S. Ferreira, V. de Zea Bermudez, S.J.L. Ribeiro, Lanthanide-containing light-emitting organic-inorganic hybrids: a bet on the future. Adv. Mater. 21, 509–534 (2009). https://doi.org/10.1002/adma.200801635 17. W.T. Carnall, H. Crosswhite, H.M. Crosswhite, in Energy Level Structure and Transition Probabilities in the Spectra of Trivalent Lanthanides in LaF 3 . Argonne Nat. Lab. 78-XX-95 Rep. (1977). https://doi.org/10.2172/6417825 18. P.C. Chen, Y.N. Chen, P.C. Hsu, C.C., Shih, H.T. Chang, Photoluminescent organo-silane functionalized carbon dots as temperature probes. Chem. Commun. 49, 1639–1641 (2013). https://doi.org/10.1039/C3CC38486A 19. D. Chen, S. Liu, Y. Zhou, Z. Wan, P. Huang, Z. Ji, Dual-activator luminescence of RE/TM: Y3 Al5 O12 (RE = Eu3+ , Tb3+ , Dy3+ ; TM = Mn4+ , Cr3+ ) phosphors for self-referencing optical thermometry. J. Mater. Chem. C 4, 9044–9051 (2016). https://doi.org/10.1039/C6TC02934E 20. A. Ciric, T. Gavrilovic, M.D. Dramicanin, Luminescence intensity ratio thermometry with Er3+ : performance overview. Crystals 11, 189 (2021). https://doi.org/10.3390/cryst11020189 21. S.F. Collins, G.W. Baxter, S.A. Wade, Comparison of fluorescence-based temperature sensor schemes: theoretical analysis and experimental validation. J. Appl. Phys. 84, 4649 (1998). https://doi.org/10.1063/1.368705 22. Y. Dong, H. Pang, H.B. Yang, Carbon-based dots codoped with Nitrogen and Sulfur for high quantum yield and excitation-independent emission. Angew. Chem. Int. Ed. 52, 7800–7804 (2013). https://doi.org/10.1002/anie.201301114 23. J.S. Donner, S.A. Thompson, M.P. Kreuzer, G. Baffou, R. Quidant, Mapping intracellular temperature using green fluorescent protein. Nano Lett. 12, 2107–2111 (2012). https://doi. org/10.1021/nl300389y 24. J.S. Donner, S.A. Thompson, C. Alonso-Ortega et al., Imaging of plasmonic heating in a living organism. ACS Nano 7, 8666–8672 (2013). https://doi.org/10.1021/nn403659n 25. P. Dorenbos, Charge transfer bands in optical materials and related defect level location. Opt. Mater. 69, 8–22 (2017). https://doi.org/10.1016/j.optmat.2017.03.061 26. F.J. Duarte, Tunable organic dye lasers: physics and technology of high-performance liquid and solid-state narrow-linewidth oscillators. Prog. Quant. Electron. 36, 29–50 (2012). https:// doi.org/10.1016/j.pquantelec.2012.03.002 27. M.D. Dramicanin, Luminescence Thermometry (Woodhead Publishing, Cambridge, 2018) 28. S. Ebert, K. Travis, B. Lincoln, J. Guck, Fluorescence ratio thermometry in a microfluidic dual-beam laser trap. Opt. Express 15, 15493–15499 (2007). https://doi.org/10.1364/OE.15. 015493 29. C.E. Estrada-Perez, Y.A. Hassan, S. Tan, Experimental characterization of temperature sensitive dyes for lased induced fluorescence thermometry. Rev. Sci. Instrum. 82, 074901 (2011). https://doi.org/10.1063/1.3590929 30. D. Farrusseng, Metal-Organic Frameworks: Applications from Catalysis to Gas Storage (Wiley-VCH, Weinheim, 2011) 31. S. Feryforgues, J.P. Fayet, A. Lopez, Drastic changes in the fluorescence properties of NBD probes with the polarity of the medium. Involvement of a TICT state? J. Photochem. Photobiol. A 70, 229–243 (1993). https://doi.org/10.1016/1010-6030(93)85048-D 32. H. Fonger, C.W. Struck, Eu3+ 5 D quenching to the charge-transfer states in Y2 O2 S, La2 O2 S and LaOCl. J. Chem. Phys. 52, 6364–6366 (1970). https://doi.org/10.1063/1.1672952 33. T. Förster, Zwischenmolekulare Energiewanderung und Fluoreszenz. Ann. Phy. 437, 55–75 (1948). https://doi.org/10.1002/andp.19484370105 34. T. Glawdel, Z. Almutairi, S. Wang, C.L. Ren, Photobleaching absorbed Rhodamine B to improve temperature measurements in PDMS microchannels. Lab Chip 9, 171–174 (2009). https://doi.org/10.1039/B805172K

64

J. J. Carvajal and M. C. Pujol

35. G.P. Goodrich, L. Bao, K. Gill-Sharp, K.L. Sang, J. Wang, J.D. Payne, Photothermal therapy in a murine colon cancer model using near-infrared absorbing gold nanorods. J. Biomed. Opt. 15, 1–8 (2010). https://doi.org/10.1117/1.3290817 36. C. Görller-Walrand, K. Binnemans, in Handbook on the Physics and Chemistry of Rare Earths, vol. 25, ed. by K.A. Gschneidner Jr., L. Eyring, (North-Holland Publishers, Amsterdam 1998), p. 101 37. C. Gota, K. Okabe, T. Funatsu, Y. Harada, S. Uchiyama, Hydrophilic fluorescent nanogel thermometer for intracellular thermometry. J. Am. Chem. Soc. 131, 2766–2767 (2009). https:// doi.org/10.1021/ja807714j 38. M. Gu, Advanced Optical Imaging Theory (Springer, Berlin, Heidelberg, 2000) 39. B. Han, W.L. Hanson, K. Bensalah, A. Tuncel, J.M. Stern, J.A. Cadeddu, Development of quantum dot-mediated fluorescence thermometry for thermal therapies. Ann. Biomed. Eng. 37, 1230–1239 (2009). https://doi.org/10.1007/s10439-009-9681-6 40. P. Haro-Gonzalez, L. Martinez-Maestro, I.R. Martin, J. Garcia-Sole, D. Jaque, Highsensitivity fluorescence lifetime thermal sensing based on CdTe quantum dots. Small 8, 2652–2658 (2012). https://doi.org/10.1002/smll.201102736 41. E. Hemmer, A. Benayas, F. Légaré, F. Vetrone, Exploiting the biological windows: corrent perspectives on fluorescent bioprobes emitting above 1000 nm. Nanoscale Horiz. 1, 168–184 (2016). https://doi.org/10.1039/C5NH00073D 42. B. Henderson, G.F. Imbush, Optical Spectroscopy of Inorganic Solids (Oxford Science Publications, London, 2006) 43. M.A. Hernandez-Rodriguez, A.D. Lozano-Gorrin, I.R. Martin, U.R. Rodriguez-Mendoza, V. Lavin, Comparison of the sensitivity as optical temperature sensor of nano-perovskite doped with Nd3+ ions in the first and second biological windows. Sens. Actuators B 255, 970–976 (2018). https://doi.org/10.1016/j.snb.2017.08.140 44. C.H. Hsia, A. Wutting, H. Yang, An accessible approach to preparing water-soluble Mn2+ doped (CdSSe) ZnS (core) shell nanocrystals for ratiometric temperature sensing. ACS Nano 5, 9511–9522 (2011). https://doi.org/10.1021/nn2025622 45. N. Ishiwada, S. Fujioka, T. Ueda, T. Yokomori, Co-doped Y2 O3 :Tb3+ /Tm3+ multicolor emitting phosphors for thermometry. Opt. Lett. 36, 760 (2011). https://doi.org/10.1364/OL.36. 000760 46. D. Jaque, J. García Solé, in Thermometry at the Nanoscale: Techniques and Selected Applications, ed. by L.D. Carlos, F. Palacio (Royal Society of Chemistry, Oxfordshire, 2015), p. 85 47. S.M. Jeon, J. Turner, S. Granick, Noncontact temperature measurement in microliter-sized volumes using fluorescent-labeled DNA oligomers. J. Am. Chem. Soc. 125, 9908–9909 (2003). https://doi.org/10.1021/ja036431+ 48. A. Jorio, M. S. Dresselhaus, in Nanostructured Materials: Metrology. Encyclopedia of Materials: Science and Technology (2010), pp. 1–7. ISBN: 978-0-0804-3152-9 49. B.R. Judd, Optical absorption intensities of rare-earth ions. Phys. Rev. 127, 750–761 (1962). https://doi.org/10.1103/PhysRev.127.750 50. W. Jung, Y.W. Kim, D. Yim, J.Y. Yoo, Microscale surface thermometry using SU8/Rhodamine-B thin layer. Sensor Actuat. 171, 228 (2011). https://doi.org/10.1016/j.sna. 2011.06.025 51. S. Kalytchuk, K. Polakova, Y. Wang, J.P. Froning, K. Cepe, A.L. Rogach, R. Zboril, Carbon dot nanothermometry: Intracellular photoluminescence lifetime thermal sensing. ACS Nano 11, 1432–1442 (2017). https://doi.org/10.1021/acsnano.6b06670 52. F.H. Ko, L.Y. Weng, C.J. Ko, T.C. Chu, Characterizagtion of imprinting polymeric temperature variation with fluorescent Rhodamine B molecule. Microelectron Eng. 83, 864–868 (2006). https://doi.org/10.1016/j.mee.2006.01.009 53. G. Kucsko, P.C. Maurer, N.Y. Yao et al., Nanometrescale thermometry in a living cell. Nature 500, 54–58 (2013). https://doi.org/10.1038/nature12373 54. M.G. Lahoud, R.C.G. Frem, D.A. Galico, G. Bannach, M.M. Nolasco, R.C.G. Ferreira, L.D. Carlos, Intriguing light-emission features of ketoprofen-based Eu(III) adduct due to a strong

Introduction to Luminescence Thermometry

55. 56. 57.

58. 59.

60.

61.

62. 63.

64. 65.

66.

67.

68. 69.

70.

71.

72. 73.

65

electron-phonon coupling. J. Lumin. 170, 357–363 (2016). https://doi.org/10.1016/j.jlumin. 2015.08.050 R.N. Lawson, L.L. Alt, Skin temperature recording with phosphors: a new technique. Canad. Med. Ass. J. 92, 255–260 (1965) J. Lee, N.A. Kotov, Thermometer design at the nanoscale. Nano Today 2, 48–51 (2007). https://doi.org/10.1016/S1748-0132(07)70019-1 S.F. León-Luis, U.R. Rodríguez-Mendoza, P. Haro-González, I.R. Martín, V. Lavín, Role of the host matrix on the thermal sensitivity of Er3+ luminescence in optical temperature sensors. Sens. Actuators B 174, 176–186 (2012). https://doi.org/10.1016/j.snb.2012.08.019 S. Li, K. Zhang, J.M. Yang, L. Lin, H. Yang, Single quantum dots as local temperature markers. Nano Lett. 7, 3102–3105 (2007). https://doi.org/10.1021/nl071606p R. Lisiecki, P. Solarz, G. Dominiak-Dzik, W. Ryba-Romanowski, M. Sobchyk, P. Cerny, J. Sulc, Comparative optical study of thulium-doped YVO4 , GdVO4 and LuVO4 single crystals. Phys. Rev. B: Condens. Matter Mater. Phys. 74, 035103 (2006). https://doi.org/10.1103/Phy sRevB.74.035103 V. Lojpur, M.G. Nikolic, D. Jovanovic, M. Medic, Z. Antic, M.D. Dramicanin, Luminescence thermometry with Zn2+ SiO4 : Mn powder. Appl. Phys. Lett. 103, 141912 (2013). https://doi. org/10.1063/1.4824208 V. Lojpur, Z. Antic, M.D. Dramicanin, Temperature sensing from the emission rise times of Eu3+ in SrY2 O4 . Phys. Chem. Chem. Phys. 16, 25636–25641 (2014). https://doi.org/10.1039/ C4CP04141K N. Lomenie, D. Racoceanu, A. Gouaillard, Advances in Bioimaging: From Physics to Signal Understanding Issues (Springer, Berlin, 2001) P. Löw, B. Kim, N. Takama, C. Bergaud, High-spatial-resolution surface-temperature mapping using fluorescent thermometry. Small 4, 908–914 (2008). https://doi.org/10.1002/smll.200 700581 Q.C. Lum, Über Fluorescenz und Phosphorescenz, I. Abhandlung. Ann. Phys. 34, 446–463 (1888) L.M. Maestro, E.M. Rodriguez, M.C.I. Cruz et al., CdSe quantum dots for two-photon fluorescence thermal imaging. Nano Lett. 10, 5109–5115 (2010). https://doi.org/10.1021/nl1 036098 L. Martínez Maestro, C. Jacinto, U. Rocha, M.C. Iglesias-de la Cruz, F. Sanz-Rodrigez, A. Juarranz, J. García Solé, D. Jaque, Optimum quantum dot size for highly efficient fluorescent bioimaging. J. Appl. Phys. 111, 023513 (2012). https://doi.org/10.1063/1.3676251 L.M. Maestro, P. Haro-Gonzalez, A. Sanchez-Iglesias, L.M. Liz-Marzan, J.G. Sole, D. Jaque, Quantum dot thermometry evaluation of geometry dependent heating efficiency in gold nanoparticles. Langmuir 30, 1650–1658 (2014). https://doi.org/10.1021/la403435v E.J. McLaurin, L.R. Bradshaw, D.R. Gamelin, Dual-emitting nanoscale temperature sensors. Chem. Mater. 25, 1283–1292 (2013). https://doi.org/10.1021/cm304034s V.K. Natrajan, K.T. Christensen, Two-color laser-induced fluorescent thermometry for microfluidic systems. Meas. Sci. Technol. 20, 015401 (2009). https://doi.org/10.1088/09570233/20/1/015401 A. Nexha, M.C. Pujol, F. Díaz, M. Aguiló, J.J. Carvajal, Luminescence nanothermometry using selfassembled Er3+ , Yb3+ doped Y2 O3 nanodiscs: might the upconversion mechanism condition their use as primary thermometers?. Opt. Mater. 134, 113216 (2022). https://doi. org/10.1016/j.optmat2022.113216 D.J. Norris, M.G. Bawendi, Measurement and assignment of the size-dependent optical spectrum in CdSe quantum dots. Phys. Rev. B: Condens. Matter Mater. Phys. 53, 16338–16346 (1996). https://doi.org/10.1103/PhysRevB.53.16338 G.S. Ofelt, Intensities of crystal spectra of rare-earth ions. J. Chem. Phys. 37, 511–520 (1962). https://doi.org/10.1063/1.1701366 T.C. O’Haver, in A Pragmatic Introduction to Signal Processing: With Applications in Scientific Measurement, 1st edn. (CreateSpace Independent Publishing Platform, 2016)

66

J. J. Carvajal and M. C. Pujol

74. A. Omrane, G. Juhlin, F. Ossler, M. Alden, Temperature measurements of single droplets by use of laser induced phosphorescence. Appl. Opt. 43, 3523–3529 (2004) 75. Y. Park, C. Ko, H.Y. Chen, A. Han, D.H. Son, Ratiometric temperature imaging using environment sensitive luminescence of Mn-doped core-shell nanocrystals. Nanoscale 5, 4944–4950 (2013). https://doi.org/10.1039/C3NR00290J 76. C. Paviolo, A.H. Clayton, S.L. McArthur, P.R. Stoddart, Temperature measurement in the microscopic regime: a comparison between fluorescence lifetime- and intensity-based methods. J. Microsc. 250, 179–188 (2013). https://doi.org/10.1111/jmi.12033 77. E.J.G. Peterman, F. Gittes, C.F. Schmidt, Laser-induced heating in optical traps. Biophys. J. 84, 1308–1316 (2003). https://doi.org/10.1016/S0006-3495(03)74946-7 78. C. Pietsch, R. Hoogenboom, U.S. Schubert, Soluble polymeric dual sensor for temperature and pH value. Angew. Chem. Int. Ed. 48, 5653–5656 (2009). https://doi.org/10.1002/anie.200 901071 79. C. Pietsch, A. Vollrath, R. Hoogenboom, U.S. Schubert, A fluorescent thermometer based on a pyrene-labeled thermoresponsive polymer. Sensors 10, 7979–7990 (2010). https://doi.org/ 10.3390/s100907979 80. T.J.V. Prazeres, M. Beija, M.T. Charreyre, J.P.S. Farinha, J.M.G. Martinho, RAFT polymerization and self-assembly of thermoresponsive poly(Ndecylacrylamide-b-N, Ndiethylacrylamide) block copolímers bearing a phenanthrene fluorescent α-end group. Polymer 51, 355–367 (2010). https://doi.org/10.1016/j.polymer.2009.11.055 81. M. Quintanilla, A. Benayas, R. Naccache, F. Vetrone, in Thermometry at the Nanoscale: Techniques and Selected Applications, ed. by L.D. Carlos, F. Palacio (Royal Society of Chemistry, Oxfordshire, 2015) p. 124 82. U. Rocha, C.J. da Silva, W.F. Silva et al., Subtissue thermal sensing based on neodymiumdoped LaF3 nanoparticles. ACS Nano 7, 1188–1199 (2013). https://doi.org/10.1021/nn3 04373q 83. D. Ross, M. Gaitan, L.E. Locascio, Temperature measurement in microfluidic systems using a temperature-dependent fluorescent dye. Anal. Chem. 73, 4117–4123 (2001). https://doi.org/ 10.1021/ac010370l 84. D. Roy, W.L.A. Brooks, B.S. Sumerlin, New directions in thermoresponsive polymers. Chem. Soc. Rev. 42, 7214–7243 (2013). https://doi.org/10.1039/C3CS35499G 85. E. Saïdi, N. Babinet, L. Lalouat, J. Lesueur, L. Aigouy, S. Volz, J. Labéguerie-Egéa, M. Mortier, Tuning temperature and size of hot spots and hot-spot arrays. Small 7, 259–264 (2011). https://doi.org/10.1002/smll.201001476 86. K.E. Sapsford, W.R. Algar, L. Berti et al., Functionalizing nanoparticles with biological molecules: developing chemistries that facilitate nanotechnology. Chem. Rev. 113, 1904–2074 (2013). https://doi.org/10.1021/cr300143v 87. O.A. Savchuk, J.J. Carvajal, C. Cascales, J. Massons, M. Aguilo, F. Diaz, Thermochromic upconversion nanoparticles for visual temperature sensors with high thermal, spatial and temporal resolution. J. Mater. Chem. C 4, 6602–6613 (2016). https://doi.org/10.1039/C6T C01841F 88. O.A. Savchuk, J.J. Carvajal, C.D.S. Brites, L.D. Carlos, M. Aguilo, F. Diaz, Upconversion thermometry: a new tool to measure the thermal resistance of nanoparticles. Nanoscale 10, 6602–6610 (2018). https://doi.org/10.1039/C7NR08758F 89. O.A. Savchuk, J.J. Carvajal, P. Haro-Gonzalez, M. Aguiló, F. Díaz, Luminescence nanothermometry using short-wavelength infrared light. J. Alloys Compd. 746, 710–719 (2018). https://doi.org/10.1016/j.jallcom.2018.03.002 90. O. Savchuk, J.J. Carvajal, C. Cascales, P. Haro-Gonzalez, F. Sanz-Rodríguez, M. Aguilo, F. Diaz, Bifunctional Tm3+ , Yb3+ :GdVO4 @SiO2 core-shell nanoparticles in HeLa cells: upconversion luminescence nanothermometry in the first biological window and biolabelling in the visible. Nanomater 10, 993 (2020). Error! Hyperlink reference not valid. 91. J.J. Shah, M. Gaitan, J. Geist, Generalized temperature measurement equations for Rhodamine B dye solution and its application to microfluidics. Anal. Chem. 81, 8260–8263 (2009). https:// doi.org/10.1021/ac901644w

Introduction to Luminescence Thermometry

67

92. L. Shang, F. Stockmar, N. Azadfar, G.U. Nienhaus, Intracellular thermometry by using fluorescent gold nanoclusters. Angew. Chem. Int. Ed. 52, 11154–11157 (2013). https://doi.org/ 10.1002/anie.201306366 93. R.S. Sirohi, Introduction to Optical Metrology (CRC Press/Taylor & Francis Group, New York, 2016) 94. J.G. Sole, L.E. Bausa, D. Jaque, An Introduction to the Optical Spectroscopy of Inorganic Solids (Wiley & Sons, West Sussex, 2005) 95. M.I.J. Stich, L.H. Fischer, O.S. Wolfbeis, Multiple fluorescent chemical sensing and imaging. Chem. Soc. Rev. 39, 3102–3114 (2010). https://doi.org/10.1039/B909635N 96. L. Tang, J.K. Jin, A.J. Qin et al., A fluorescent thermometer operating in aggregation-induced emission mechanism: probing thermal transitions of PNIPAM in water. Chem. Commun. 4974–4976 (2009). https://doi.org/10.1039/B907382E 97. S. Uchiyama, Y. Matsumura, A.P. de Silva, K. Iwai, Fluorescent molecular thermometers based on polymers showing temperature-induced phase transitions and labelled with polarityresponsive benzofurazans. Anal. Chem. 75, 5926–5935 (2003). https://doi.org/10.1021/ac0 346914 98. S. Uchiyama, K. Kimura, C. Gota et al., Environment-sensitive fluorophores with benzothiadiazole and benzoselenadiazole structures as candidate components of a fluorescent polymeric thermometer. Chem. Eur. J. 18, 9552–9563 (2012). https://doi.org/10.1002/chem.201200597 99. G. Vancoillie, Q. Zhang, R. Hoogenboom, in Thermometry At The Nanoscale: Techniques and Selected Applications, ed. by L.D. Carlos, F. Palacio (Royal Society of Chemistry, Oxfordshire, 2015) p. 3 100. T. van Swieten, A. Meijerink, F.T. Rabouw, Impact of noise and background on measurement uncertainties in luminescence thermometry. ACS Photonics 9, 1366–1374 (2022). https://doi. org/10.1021/acsphotonics.2c00039 101. L. Van Uitert, L. Johnson, Energy transfer between rare-earth ions. J. Chem. Phys. 44, 3514 (1966). https://doi.org/10.1063/1.1727258 102. F. Vetrone, R. Naccache, A. Zamarrón et al., Temperature sensing using fluorescent nanothermometers. ACS Nano 4, 3254–3258 (2010). https://doi.org/10.1021/nn100244a 103. V.A. Vlaskin, N. Janssen, J. van Rijssel, R. Beaulac, D.R. Gamelin, Tunable dual emission in doped semiconductor nanocrystals. Nano Lett. 10, 3670–3674 (2010). https://doi.org/10. 1021/nl102135k 104. S.A. Wade, S.F. Collins, G.W. Baxter, Fluorescence intensity ratio technique for optical fiber point temperature sensing. J. Appl. Phys. 94, 4743–4756 (2003). https://doi.org/10.1063/1. 1606526 105. G.W. Walker, V.C. Sundar, C.M. Rudzinski, A.W. Wun, M.G. Bawendi, D.G. Nocera, Quantum-dot optical temperature probes. Appl. Phys. Lett. 83, 3555–3557 (2003). https:// doi.org/10.1063/1.1620686 106. C. Wang, Y. Huang, H. Lin et al., Gold nanoclusters based dual-emission hollow TiO2 microspheres for ratiometric optical thermometry. RSC Adv. 5, 61586–61592 (2015). https://doi. org/10.1039/C5RA13475G 107. C. Wang, H. Lin, Z. Xu et al., Tunable carbon-dotbased dual-emission fluorescent nanohybrids for ratiometric optical thermometry in living cells. ACS Appl. Mater. Interfaces 8, 6621–6628 (2016). https://doi.org/10.1021/acsami.5b11317 108. B.A. Weinstein, Ruby thermometer for cryobaric diamond-anvil cell. Rev. Sci. Instrum. 57, 910 (1986). https://doi.org/10.1063/1.1138833 109. N. Xie, J. Huang, X. Yang et al., Scallop-inspired DNA nanomachine: a ratiometric nanothermometer for intracellular temperature sensing. Anal. Chem. 89, 12115–12122 (2017). https:// doi.org/10.1021/acs.analchem.7b02709 110. E.C. Ximendes, U. Rocha, T.O. Sales, N. Fernández, F. Sanz-Rodríguez, I.R. Martín, C. Jacinto, D. Jaque, In vivo subcutaneous thermal video recording by supersensitive infrared nanothermometers. Adv. Funct. Mater. 27, 1702249 (2017). https://doi.org/10.1002/adfm.201 702249

68

J. J. Carvajal and M. C. Pujol

111. L. Xing, Y. Xu, R. Wang, W. Xu, Z. Zhang, Highly sensitive Optical thermometry based on upconversion emissions in Tm3+ /Yb3+ codoped LiNbO3 single crystal. Opt. Lett. 39, 454–457 (2014). https://doi.org/10.1364/OL.39.000454 112. J. Yang, F. Bai, Temperature dependence of the fluorescence spectra of ladderlike polyphenyl silsesquioxane and ladderlike 1,4-phenylenen-bridged polyvinylsiloxane. Chin. J. Polym. Sci. 20, 15–23 (2002) 113. Y. Yang, W. Kong, H. Li et al., Fluorescent N-doped carbon dots as in vitro and in vivo nanothermometer. ACS Appl. Mater. Interfaces 7, 27324–27330 (2015). https://doi.org/10. 1021/acsami.5b08782 114. G. Yang, X. Liu, J. Feng, S. Li, Y. Li, in Thermometry At The Nanoscale: Techniques and Selected Applications, ed. by L.D. Carlos, F. Palacio (Royal Society of Chemistry, Oxfordshire, 2015) p. 85 115. S.V. Yap, R.M. Ranson, W.M. Cranton, D.C. Koutsogeorgis, G.B. Hix, Temperature dependent characteristics of La2 O2 S: Ln (Ln = Eu, Tb) with various Ln concentrations over 5–60 º. J. Lumin. 129, 416–422 (2009). https://doi.org/10.1016/j.jlumin.2008.10.019 116. A.R. Zanatta, D. Socca, F. Alvarez, A suitable (wide-range + linear) temperature sensor based on Tm3+ ions. Sci. Rep. 7, 14113 (2017). https://doi.org/10.1038/s41598-017-14535-1 117. J. Zhou, B. del Rosal, D. Jaque, S. Uchiyama, D. Jin, Advances and challenges for fluorescence nanothermometry. Nat. Methods 17, 967–980 (2020). https://doi.org/ https://doi.org/10.1038/ s41592-020-0957-y 118. H. Zollinger, Color Chemistry: Syntheses, Properties and Applications of Organic Dyes and Pigments (Wiley-VCH, Weinheim, 2003)

New Strategies to Improve Thermal Sensitivity and Temperature Resolution in Lanthanide-Doped Luminescent Thermometers L. Marciniak, W. M. Piotrowski, M. Szymczak, M. Pieprz, and K. Trejgis

Abstarct One of the fundamental and most important findings in luminescent thermometry is that it is impossible to develop a universal optical temperature sensor with high sensitivity over a very wide temperature range. The measurement requirements of a particular type of application impose certain strict limitations to which the luminescent thermometer should be adjusted. For this reason, new solutions that meet those requirements are constantly being sought. The purpose of this chapter is to review the five most promising approaches in lanthanide-based luminescence thermometry introduced recently: (i) sensitization of lanthanide-based luminescence thermometers by the transition metal ions, (ii) thermometers with negative thermal expansion coefficient (iii) thermometers involving thermally induced first order phase transition; (iv) single band ratiometric approach and (v) thermometry based on multilevel thermal coupling. The chapter includes a critical discussion of the advantages and disadvantages of the aforementioned solutions and outlines possible paths for the development of these techniques. Keywords Luminescent thermometry · Relative sensitivity · Lanthanide ions · Ln3+

L. Marciniak (B) · W. M. Piotrowski · M. Szymczak · M. Pieprz · K. Trejgis Institute of Low Temperature and Structure Research, Polish Academy of Sciences, Okólna 2, 50-422 Wroclaw, Poland e-mail: [email protected] W. M. Piotrowski e-mail: [email protected] M. Szymczak e-mail: [email protected] M. Pieprz e-mail: [email protected] K. Trejgis e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_2

69

70

L. Marciniak et al.

1 Introduction The growing scientific interest in luminescent thermometry, manifested by the exponential increase in the number of publications on this subject observed in recent years, is mainly due to the possibilities and advantages offered by this technique. Among the numerous and often emphasized advantages of luminescent thermometry are the remote readout mode, the readout in electrically passive mode and the high sensitivity. In addition, the fact that in order to read the temperature the variation of one of the spectroscopic parameters of the phosphor is analyzed, luminescence thermometry allows not only a spot readout, but above all allows the imaging of the temperature of objects in their working environment. Among the many different materials used for remote temperature measurement, such as organic and carbon–metal materials, quantum dots, dyes and polymeric materials, inorganic phosphors doped with luminescent ions are the most widely used. This is due to their numerous benefits, which include high chemical, physical and thermal durability, even at very high temperatures, high sensitivity and emission efficiency and the absence of photobleaching. These advantages have been confirmed and described in numerous review papers. Nevertheless, large-scale research races include both the impact of optimizing matrix chemistry, dopant ion type and concentration and measurement parameters. It is indisputable that as luminescent dopants, lanthanide ions Ln3+ are the most widely used. They are characterized by narrow emission bands originating from the nature of spin-forbidden 4f-4f electronic transitions, which are significantly simpler to separate spectrally for use in remote temperature readout. In addition to the arrangement of the composition of the luminophore material, it is also extremely important to understand the processes responsible for thermal changes in spectroscopic parameters, which in recent years have been extremely and meticulously studied, correctly identified and theoretically described, so that today luminescence thermometry has entered an advanced level of development. This advancement of the technique has been made possible by the emergence of many works aimed at increasing the accuracy and precision of temperature imaging using phosphors doped with Ln3+ ions. According to equation Eq. 1, the temperature resolution is inversely proportional to the relative sensitivity of the thermometer and directly proportional to the uncertainty in the readout of the spectroscopic parameters: δT =

1 δΩ SR Ω

(1)

where Ω is the thermometric parameter while δΩ/Ω is the uncertainty of the Ω determination, while SR is the relative sensitivity defined as follows: SR =

1 ∂Ω 100% Ω ∂T

δΩ is the change of the Ω corresponding to the ΔT change of temperature.

(2)

New Strategies to Improve Thermal Sensitivity and Temperature …

71

Fig. 1 Schematics illustrating the processes of sensitization using transition metal ions (a), negative expansion coefficient (b), phase transition (c), excited state absorption (d) and thermal coupling between multiple levels (e) used in different approaches to increase the precision of temperature determination in luminescence thermometry

The latter parameter affecting the temperature determination uncertainty is related to the emission brightness, hence intensive work is being carried out to increase the sensitivity of Ln3+ based thermometers and/or to increase their emission brightness. The aim of this chapter is to describe the most important strategies developed in the last 5 years to increase the precision of temperature determination. These include, as illustrated in Fig. 1: (i) the use of transition metal ions to enhance the emission intensity and sensitivity of Ln3+ -based thermometers; (ii) the use of a multilevel thermally coupled level system; (iii) the use of materials that exhibit a thermally induced first-order structural phase transition; (iv) the use of an absorption process from the excited state in single band ratiometric based thermometers; and (v) the use of materials with a negative thermal expansion coefficient.

2 Multilevel Thermal Coupling A limitation of luminescence thermometers based on the luminescence intensity ratio from two thermally coupled levels, often raised in the literature, is the low maximal achievable value of relative sensitivity. When the LIR is described by the Boltzmann relation, the sensitivity of the luminescence thermometer is directly proportional

72

L. Marciniak et al.

to the energy gap between these levels. However, when the energy gap exceeds 2000 cm−1 , very high thermal energy is required to populate the higher lying state. Therefore, even at high temperatures, the population of this level is low enough to prevent the observation of efficient luminescence for this multiplet. Consequently, the maximum sensitivity of a luminescent thermometer based on a thermally coupled levels usually does not exceed SR = 1.5%/K. In order to circumvent this limitation, a solution using thermal coupling of three (or even higher number) energy levels has been proposed. If a system of three energy levels, labeled N1 , N2 and N3 , (Fig. 2a) separated from each other by ΔE21 and ΔE23 , respectively, is considered, the ratio of the emission intensities from these levels may be written with the following equations: ) ( N2 ΔE21 = A1 exp − N1 kT ) ( N3 ΔE32 ∼ = A2 exp − N2 kT

LIR21 ∼

(3)

LIR32

(4)

and hence: S21 =

ΔE21 kT 2

(5)

S32 =

ΔE32 kT 2

(6)

Simple transformations lead to the result that the ratio of the luminescence intensities of the N3 and N1 levels is the product of the Boltzmann terms of Eqs. 3 and 4: ) ( ΔE31 N3 (7) = A3 exp − LIR31 ∼ N1 kT Hence, the relative sensitivity of such a system is the sum of the relative sensitivities of the thermometers based on the individual level pairs (Fig. 2b): S31 =

ΔE32 ΔE31 ΔE21 + = S21 + S32 = 2 2 kT kT kT 2

(8)

The introduction of an additional intermediate energy level between the N3 and N1 levels facilitates an increase in the population probability of the N3 level through a ladder-like thermalization process. Consequently, the theoretical maximum ΔE31 value in such a system can reach about 4000 cm−1 , which results in a significantly increased sensitivity of such a luminescent thermometer. This concept was most strongly developed for systems doped with Er3+ or Nd3+ ions due to the relatively dense ladder of energy levels occurring in the case of these ions. The work carried out on GdVO4 :Er3+ , Yb3+ has clearly shown the advantage of the thermometric parameters of the thermometer using the LIR of 4 F7/2 → 4 I15/2 and

New Strategies to Improve Thermal Sensitivity and Temperature …

73

Fig. 2 Energy level diagram of Dy3+ ions in YAG and observed emissions with marked corresponding energy differences (a) [6]; and the thermal dependence of relative sensitivities of luminescent thermometers involving different pairs of energy levels (b) [6]; thermal dependence of relative sensitivities of luminescent thermometers involving different pairs of energy levels in YF3 :Er3+ (c) [4]; plot of the thermal population probabilities per microstate, pm/gm, of an excited three-level system in thermodynamic equilibrium determining the response of a luminescent thermometer in YAl3 (BO3 )4 :Gd3+ , Pr3+ (d); and the relative sensitivities of the different temperature measures and their improvement upon thermodynamically guided change of the respective energy gap of the excited levels of Gd3+ used for temperature sensing (e); overall theoretical relative temperature uncertainty of the different energy gaps assuming a constant integrated intensity measure of I10 = 107 counts of the selected lowest energetic emission. The shaded areas indicate the statistical fluctuations in the relative temperature uncertainty (f) [8]; temperature-dependent normalized PL spectra of LaPO4 : Yb3+ /Nd3+ nanourchins along with absolute and relative sensitivity based on three TCLs of Nd3+ (λexc = 980 nm); (g) [12]; thermal dependences of SR for different pairs of energy levels in CaWO4 : Nd3+ , Yb3+ , 7 mol% Li+ phosphors; t. (h) [11]; normalized (at 362.4 nm) temperaturedependent luminescence spectra of the SrB4 O7 :Eu2+ material, measured in the 11–600 K range upon excitation at 300 nm (i) [13]

S3/2 → 4 I15/2 over the thermometer most commonly reported in the literature using the LIR of 2 H11/2 → 4 I15/2 and 4 S3/2 → 4 I15/2 . In the former case, a more rapid rise of the LIR with increasing temperature was obtained which reflected into a more than doubling of the maximum SR from SR = 0.9%/K to SR = 2.3%/K at 333K [1]. Li et al. demonstrated that using this strategy, the limitation for luminescent thermometers according to which the thermal dependence of SR should not exceed 2000/T2 can

4

74

L. Marciniak et al.

be surpassed. Whereas in the CaWO4 :Er3+ , Yb3+ system, the authors observed an increase in maximum sensitivity to 2820/T2 bypassing this limitation [2]. The potential importance of this approach and its influence on the thermometric performance of the luminescent thermometer has been confirmed by theoretical calculations made by Ciric et al. [3]. According to theoretical predictions, the maximum sensitivity in this approach determined taking into account the value of the energy gap between the 4 S3/2 and 4 F7/2 levels is expected to be SR = 3.8%/K for Sb2 O3 -35P2 O5 -5MgO-AgCl glass material and SR = 3.68%/K for NaY(MoO4 )2 crystals. These values are well above those predicted for a conventional two-level thermometer on Er3+ ions. Exactly a twofold increase in sensitivity was reported by Ciric et al. for an analogous system of dopant ions in YF3 :Er3+ , Yb3+ (SR = 2.03%/K (4 F7/2 → 4 I15/2 / 4 S3/2 → 4 I15/2 ) versus SR = 1.06%/K obtained for a system based on intensity analysis of two adjacent thermally coupled levels (2 H11/2 → 4 I15/2 /4 S3/2 → 4 I15/2 ) at 293 K (Fig. 2c) [4]. However, as the authors reasonably noted, despite the very high sensitivity obtained using the LIR of the 4 F7/2 → 4 I15/2 /4 S3/2 → 4 I15/2 levels, the low intensity of luminescence from the 4 F7/2 level in the temperature range 313–413 K resulted in a relatively large uncertainty in the determination of the LIR value. The authors indicated that the relative uncertainty of the LIR (4 F7/2 → 4 I15/2 / 4 S3/2 → 4 I15/2 ) was ~1.3% versus 0.31% obtained for LIR associated with the 2 H11/2 → 4 I15/2 / 4 S3/2 → 4 I15/2 electronic transitions. This resulted in a significantly higher temperature resolution in the first case (δT = 1.8 K) relative to the second approach (δT = 0.3 K). In addition to the doubling of the maximum sensitivity obtained for the three-level thermally coupled system, Li et al. [5] showed that, in contrast to the two-level system, for which an increase in temperature results in a decrease in relative sensitivity, for the three-level system an increase in sensitivity with increasing temperature is observed. Therefore, this type of thermometers reveals higher thermometric performance in the high temperature range (>400 K). This fact results not only from the higher SR , but also from the higher emission intensity from the upper energy level. Depending on the energy level scheme of the luminescent ion, this approach can be extrapolated to a multilevel array of energy states (Fig. 2a). An extension of this approach was the use of a 5-level system of Dy3+ for temperature readout, where multilevel cascade LIR (McLIR) was used as a thermometric parameter [6]. It has been demonstrated that, depending on the energy levels used in the analysis, an increase in sensitivity was achieved from SR = 0.35%/K, through SR = 0.84%/K, SR = 1.47%/K, up to SR = 1.73%/K, using the ratios of the emission intensities of the bands associated with electronic transition from 4 I15/2 ; 4 G11/2 ; 4 I13/2 , 4 M21/2 , 4 K17/2 , 4 F7/2 and 4 F7/2 to 6 H15/2 ground state, respectively, to the intensity of the 4 F9/2 → 6 H15/2 emission band. The increase in sensitivity is naturally caused by an increase in the energy distance between the levels considered from 900 cm−1 , through 2172 cm−1 , 4536 cm−1 up to 4643 cm−1 , respectively. Moreover, the analytical calculations have predicted the optimum operating temperature range of individual luminescent thermometers. According to expectations, the larger the energy separation, the higher the operating temperature. Therefore, using a multilevel thermal coupling approach, it is possible, by exploiting a single phosphor, to measure the

New Strategies to Improve Thermal Sensitivity and Temperature …

75

temperature over a very wide temperature range. This can be accomplished by optimizing the selection of energy levels from which the emission will be analyzed for a given thermal interval. Similar to Ciric et al. [4] for a thermometer based on Er3+ ion emission, Li et al. [7] proved that by using emission from a higher energy level in the case of Dy3+ ions, despite an increase in sensitivity, the uncertainty of the temperature determination increased from 3.9 K (for LIR = 4 I15/2 → 6 H15/2 /4 F9/2 → 6 H15/2 ) to 6.8 K (for LIR = 4 G11/2 → 6 H15/2 /4 F9/2 → 6 H15/2 ). As already mentioned, luminescence thermometry based on multi-level thermal coupling allows not only to increase the relative sensitivity, but also to extend the useful temperature range over which Boltzmann thermometers can be applied. Yu et al. [8] proposed the use of a three-level system of energy levels of Gd3+ ions in YAl3 (BO3 )4 :Pr3+ , Gd3+ , thus developing a thermometer operating in the 30–800 K range (Fig. 2d–e). The main motivation for this work was that for a thermometer operating on a two-level system in the high temperature range, the high probability of nonradiative processes leads to a reduction in its relative sensitivity. As shown in Fig. 2d the calculated thermal population of the m = 3 level clearly increases in the range of much higher temperatures comparing to that for two-level system counterpart. Thus, by an appropriate selection of the energy levels from which the luminescence is used to measure temperature, relative sensitivities above SR > 0.5%/K can be provided over the entire 30–800 K temperature range. As stated by the authors for this purpose, it is preferred to use the thermal coupling between the Stark components of the 6 P7/2 state in the temperature range below 200 K, while in the ranges 220–388 K and 480–800 K it is suggested to use the emission from pairs of 6 P5/2 and 6 P7/2 and 6 P3/2 and 6 P7/2 levels, respectively. Such optimization of the thermal ranges performed by selecting appropriate energy levels allows to read the temperature over the entire analyzed range with an uncertainty of the temperature determination below 0.1 K. The multilevel thermal coupling approach has also been intensively studied for materials doped with Nd3+ ions, in which case the sensitivity of the thermometer using the LIR from levels 4 F7/2 , 4 S3/2 to 4 F3/2 significantly exceeded that obtained for the 4 F5/2 , 2 H11/2 and 4 F3/2 levels. Song et al. showed that the thermal sensitivity relation in the former case follows the 3044/T2 relation compared to 1406/T2 for the 4 F5/2 , 2 H11/2 / 4 F3/2 levels pair [9]. Importantly, in this case the emission intensity of the 4 F7/2 , 4 S3/2 → 4 I9/2 band is comparable to that of the 4 F3/2 → 4 I9/2 band. This is very important because the uncertainty of the temperature determination in this case should not deviate as significantly from the two-level system as was observed for Er3+ ions. However, this is not the rule and for NaYF4 :Nd3+ , Yb3+ the 4 F7/2 , 4 S3/2 → 4 I9/2 emission band takes up less than 10% of the 4 F3/2 → 4 I9/2 band intensity [10]. Nevertheless, the intensity of the 4 F7/2 , 4 S3/2 → 4 I9/2 band increases significantly with increasing temperature, as observed, for instance, for CaWO4 :Nd3+ , Yb3+ , in which the intensity of this band was observed to significantly exceed that of the 4 F3/2 → 4 I9/2 band at 873 K (Fig. 2h) [11]. The rapid increase in the intensity of the 4 F7/2 , 4 S3/2 → 4 I9/2 band with increasing temperature can result in very high sensitivities such as SR = 3.51% captured for LaPO4 :Nd3+ , Yb3+ [12] (Fig. 2g). Furthermore, as shown by Suo et al. this approach can be successfully applied in sub-tissue thermal imaging.

76

L. Marciniak et al.

Interestingly, a similar approach has been used in a luminescent thermometer based on Eu2+ ion emission in SrB4 O7 , for which, depending on the selection of the components of the 6 P7/2 → 8 S7/2 electronic transition, it was possible to achieve relative sensitivity as high as SR = 22.6%/K at 16.5 K (Fig. 2i) [13]. An important advantage of the multilevel thermal coupling approach is that the emission bands analyzed for temperature determination are, due to the larger energy distance between the analyzed levels, spectrally better separated from each other. This facilitates separation of the analyzed signals during temperature readout. Thus, in thermometers based on Er3+ ion luminescence, emission bands with maxima at wavelengths from 490 nm (4 F7/2 → 4 I15/2 ) to 550 nm (4 S3/2 → 4 I15/2 ) are analyzed compared to bands at 530 nm (2 H11/2 → 4 I15/2 ) and 550 nm (4 S3/2 → 4 I15/2 ) in the case of a two-level system. For Dy3+ ions this is an increase in spectral band separation from 460 nm (4 I12/2 → 6 H15/2 ) versus 490 nm (4 F9/2 → 6 H15/2 ) to 390 nm (4 I15/2 ; 4 G11/2 ; 4 I13/2 , 4 M21/2 , 4 K17/2 , 4 F7/2 and 4 F7/2 → 6 H15/2 ) relative to that at 490 nm (4 F9/2 → 6 H15/2 ). In contrast, for Nd3+ ions, this approach allows the analysis of signals with maxima at 890 nm (4 F3/2 → 4 I9/2 ) and 750 nm (4 F7/2 , 4 S3/2 → 4 I9/2 ) instead of 890 nm (4F3/2 → 4 I9/2 ) and 808 nm (4 F5/2 , 2 H11/2 → 4 I9/2 ). Although, as shown in this subchapter, the proposed approach allows an enhancement of the maximum sensitivity, it should be kept in mind that the higher the energy separation from the metastable level, the lower the emission intensity from this level. This carries the risk of low accuracy of the temperature readout.

3 Excited State Absorption Based Luminescent Thermometers Apart from the numerous advantages of luminescence thermometers based on two neighboring thermally coupled energy levels of Ln3+ ions, this solution has some constraints limiting its application possibilities. Among them, there is the limited number of available pairs of thermally coupled levels, which narrows the spectral ranges over which such thermometry can be realized. In addition, the relative sensitivity of this type of luminescent thermometer, which is proportional to the energy gap between these levels, is constrained by the maximum value of the energy gap that allows the recording of a signal from a higher-lying energy level (, |2 F5/2 , 6 A1 (6 S)> and |2 F7/2 , 4 T1 (4 G)> representing the ground state, first excited state and second excited state of the dimer, respectively, which are formed by the interaction between 4f electrons of Yb3+ and 3d electrons of Mn2+ . As a result, the |2 F7/2 , 4 T1 (4 G)> level participates both in the energy transfer to the 4 F9/2 level of Er3+ ions, resulting in an enhancement of the red emission intensity, but also to the 2 H11/2 , 4 S3/2 level (Er3+ ions), resulting in an enhancement of the signal intensity in the green range (520, 540 nm). Additionally, Mn2+ doping leads to lattice distortion, that increases the asymmetry of the crystal field around Er3+ ions thus reducing the probability of nonradiative transition. This structural modification does not affect the SR , which was maintained at ~SRmax = 1.277%/K and 1.223%/K at 298.

New Strategies to Improve Thermal Sensitivity and Temperature …

83

Fig. 4 The upconversion spectra of NaScF4 : 20% Yb3+ , 2% Er3+ , x% Mn2+ nanoparticles, the inset shows the luminescent photograph of NaScF4: 20% Yb3+ , 2% Er3+ , 10% Mn2+ nanoparticles (a) [40]; the excitation spectra of NaScF4: 20% Yb3+ , 2% Er3+ and NaScF4 : 20% Yb3+ , 2% Er3+ , 10% Mn2+ nanoparticles (b) [40]; simplified energy level diagrams of Er3+ , Yb3+ , and Mn2+ ions and proposed energy transfer mechanism in NaGdF4 (c) [41]; the absolute sensitivity of NaLuF4 :Yb3+ , Er3+ and NaLuF4 :Yb3+ , Er3+ , Mn2+ phosphors as a function of temperature based on 2 H11/2 /4 S3/2 and 4 F9/2 /4 S3/2 levels (d) [42]; the upconversion luminescence spectra of NaGdF4 :20% Yb3+ , 2% Er3+ , x% Cr3+ nanoparticles with different Cr3+ concentration, inset shows ratio of red (R) (635– 680 nm) and green (G) (515–565 nm) emitting level integrated intensities (e) [44]; the absolute sensitivity curves for CaWO4 : Er3+ , Yb3+ co-doped with different concentrations of Cr3+ ions (f) [45]; simplified energy diagram of TM, Tb3+ and Eu3+ ions (g) [47]; thermal evolution of LIR based on 5 D0 → 7 F1 transition of Eu3+ ion and 5 D4 → 7 F5 electronic transition of Tb3+ ion (h) [47] and SR (LIR) for different Mn4+ concentration in YAG:Tb3+ , Eu3+ , x% Mn4+ powder (i) [47]; SR (LIR) for different Cr3+ concentration in YAG:Nd3+ , Er3+ , x% Cr3+ powder (j) [48]; excitation spectra measured for Nd3+ emission (λem = 1064 nm) in YAG:1% Nd3+ , 1% Er3+ and YAG:1% Nd3+ , 1% Er3+ , 20% Cr3+/4+ powders with the emission spectrum of the used white LED (k) [48]; absolute emission intensity at 1064 nm for white light illumination (930 mW cm−2 ) and derived enhancement factor YAG:1% Nd3+ , 1% Er3+ , x% Cr3+/4+ powders (l) [48]

84

L. Marciniak et al.

Similar approach has been evaluated for other TM ions as well. As presented by Mikalauskaite et al., a similar role is successfully played by Cr3+ ions in NaGdF4 :20% Yb3+ , 2% Er3+ [44]. As depicted in Fig. 4e, after normalization of the emission spectra performed for different Cr3+ ion concentrations to the maximum of the 4 S3/2 → 4 I15/2 band, the ratio of the 4 F9/2 → 4 I15/2 band to the other Er3+ bands clearly increases with the Cr3+ ion concentration, reaching about fivefold enhancement in the emission intensity for 50% of Cr3+ relative to the counterpart without Cr3+ . In this case, the authors suggest that as a result of energy transfer from Yb3+ to Cr3+ , the 2 T2 level of Cr3+ is occupied in a two-photon process, followed by relaxation to the 4 T2 leading to a Cr3+ → Er3+ energy transfer to the 4 F9/2 excited state. In this work, the authors also set out to investigate the change in the ratio of emission from thermally coupled 4 S3/2 and 2 H11/2 levels as a function of temperature. However, the presence of Cr3+ resulted in a decrease of SR and SA from SR = 1.12% K−1 and SA = 3.05 × 10–3 K−1 for NaGdF4 :20% Yb3+ , 2% Er3+ to SR = 0.93% K−1 at 298 K and SA = 2.84 × 10–3 K−1 for NaGdF4 :20% Yb3+ , 2% Er3+ , 15%Cr3+ . Xu et al. [45] presented that a 400% increase in emission intensity was observed in CaWO4 :Yb3+ , Er3+ , Cr3+ when the chromium ions concentration increased up to 3% Cr3+ . Further increase in Cr3+ concentrations led to a reduction in Er3+ emission intensity. Additionally, Cr3+ co-doping brought about an increase in SA from 1.36 K−1 for CaWO4 :Yb3+ , Er3+ up to 1.91 × 10–2 K−1 at around 490 K for CaWO4 :Yb3+ , Er3+ , 9%Cr3+ (Fig. 4f). When doped with Mn ions at concentrations up to 0.2%, significant enhancement of these bands was observed, namely about tenfold and about threefold increase in signal intensity for 4 F3/2 → 4 I9/2 and 4 F3/2 → 4 I11/2 , respectively, compared to the undoped sample. The presence of TM affects the thermometric properties not only up-converting luminescent thermometers but also of classical systems, an example of which isLaSrGaO4 :Nd3+ , in which an enhancement of the Nd3+ emission intensity was observed after co-doping with Mn2+ [46]. When doped with Mn2+ ions at concentrations up to 0.2%, significant enhancement of emission bands was observed, namely about tenfold and about threefold more intense signal for 4 F3/2 → 4 I9/2 and 4 F3/2 → 4 I11/2 , respectively, compared to the non-doped sample. According to the interpretation of the authors, the energy transfer originating from Mn2+ is responsible for the enhancement, although some indications, such as those reported in this work for LaSrGaO4 :Mn, an emission band with a maximum at 722 nm suitable for 2 E → 4 A2 electronic transition and two absorption bands around 420 and 570 nm resembling the typical 4 A2 → 4 T1 and 4 A2 → 4 T2 , may suggest the presence of Mn4+ ions. Comparison of the thermometric properties for LIR = 882 nm/913 nm corresponding to the Stark splitting sublevels of the 4 F3/2 → 4 I9/2 transition revealed that introducing manganese ions improved the relative sensitivity from SR = 0.19% K−1 for LaSrGaO4 :4% Nd3+ to SR = 0.41% K−1 at 300 K for LaSrGaO4 :4% Nd3+ , 0.2% Mn. A slightly different approach has been presented in works focusing on improving the relative sensitivity of thermometers based on the emission of two Ln3+ ions in materials co-doped with TM ions as well. In a such system co-doped with TM and two Ln3+ ions, which will be characterised by different energy mismatch between

New Strategies to Improve Thermal Sensitivity and Temperature …

85

the excited states of TM and Ln3+ ions, there are a number of co-existing nonradiative processes that are strongly temperature dependent (as presented in Fig. 4g). These include: TM → Ln1 3+ , TM → Ln2 3+ , Ln1 3+ → Ln2 3+ and inverse processes. Therefore, an enhancement in the thermal variation of the LIR of Ln3+ would be expected after the introduction of TM ions. To verify the correctness of this hypothesis, Piotrowski et al. [47] used a Y3 Al5 O12 :Tb3+ , Eu3+ phosphor in which the LIR of the Tb3+ to Eu3+ emission intensity was almost thermally independent and investigated this material with co-doped TM (Mn4+ , Cr3+ and Ti3+/4+ ) ions by verifying the influence of the TM dopants on thermal dependence of the LIR. The low thermal susceptibility of LIR was associated with the low probability of the nonradiative depopulation of Ln3+ excited states due to the high energy separation between the 5 D0 and 7 F6 states of Eu3+ ions (around 12,200 cm−1 ) and between the 5 D4 and 7 F0 in the case of Tb3+ ions (around 14,700 cm−1 ). The relative sensitivity based on the ratio of the intensities of the emission bands associated with the 5 D0 → 7 F1 electronic transition of Eu3+ ions (at 592 nm) and 5 D4 → 7 F5 of Tb3+ ions (at 545 nm) reached a maximum of SR = 0.09% K−1 at 190 K. It is also worth mentioning that for the YAG:Tb3+ , Eu3+ sample, the simultaneous emission of both Ln3+ ions was achieved for 266 nm excitation, corresponding to the O2− → Eu3+ charge transfer transition. In contrast, for the sample co-doped with Mn4+ ions, due to the higher absorption cross-section of Mn4+ with respect to the Ln3+ ions, the used λexc = 266 nm was mainly absorbed by the 4 A2 → 4 T1 transition of Mn4+ ions. Comparing the thermal evolution of integral band intensities of Eu3+ and Tb3+ ions for a wide range of Mn4+ ion concentrations (from 0.1 to 20%), one could see that both emission bands extinguished faster compared to the Mn4+ -undoped counterpart. However, only by calculating the LIR of the Eu3+ and Tb3+ signals, it was clear that already for the sample co-doped with 0.5% Mn4+ there was an improvement in the LIR variation as a function of temperature (Fig. 4h). This trend continued in accordance with increasing Mn4+ ion concentration, reaching the best result for 20% of Mn4+ ions. Hence, the SRmax = 0.28% K−1 was found at 250 K for the sample co-doped with 20% of Mn4+ ions, which meant that the presence of Mn4+ as a sensitizer resulted in an enhancement of the SR by over 300% compared to the unco-doped counterpart (Fig. 4i). However, it should be mentioned here that the temperature resolution may be insufficient, which is expected with the still relatively low relative sensitivity, namely the obtained temperature determination uncertainty was δT = 4 K. The maximal SR values for other TM ions analysed were SRmax = 0.16% K−1 at 370 K for 5% Cr3+ ions and SRmax = 0.30% K−1 at 555 K for 10%Ti3+/4+ . However, it is worth noting that for samples co-doped with Cr3+ ions it was possible to use λexc = 445 nm excitation (using the 4 A2 → 4 T1 band) compared to 266 nm for the Mn4+ counterpart, which also positively demonstrates the possibility to choose from a larger pool of excitation sources. A certain limitation of the presented previously results was the fact of spectral overlap between the Ln3+ emission bands and the TM absorption bands, especially evident for Cr3+ ions. Therefore, a similar approach was evaluated for Ln3+ ions emitting in the NIR range, namely Nd3+ and Er3+ [48] in the same host material

86

L. Marciniak et al.

(YAG). Interestingly, it was found that in this host material at high dopant concentration, both Cr3+ and Cr4+ oxidation states of chromium ions were observed, that significantly affects the thermometric properties of the luminescent thermometer. It was found that at λexc = 445 nm, there was an evident increase in the sensitivity of LIR based on emission of Nd3+ and Er3+ ions from SR = 0.25% K−1 at 468 K for the Cr3+/4+ undoped to SR = 1.67% K−1 at 300 K for powders co-doped with 20%Cr3+/4+ (Fig. 4j). As stated by the authors, this significant enhancement was a result of two effects. The first was the Cr3+ → Ln3+ energy transfer, as discussed above. Second, in the case of this structure, where it was possible to incorporate Cr4+ into the partially tetrahedral position of Al3+ , these ions undoubtedly served beneficially as an energy bridge for Nd3+ → Cr4+ → Er3+ energy transfer. This resulted in efficient pumping of Er3+ from three different sources (Nd3+ , Cr3+ , Cr4+ → Er3+ ) and contributed to the observed increase in the intensity of the 4 I13/2 → 4 I15/2 band at 1540 nm with temperature. Moreover, the authors quantitatively confirmed the enhancement of the Ln3+ brightness after introduction of Cr3+ /Cr4+ ions. For this purpose, excitation spectra taken for the 4 F3/2 → 4 I11/2 transition of Nd3+ and normalised to the most intense absorption band of 4 I9/2 → 4 F5/2 , 2 H9/2 of Nd3+ ions were compared for samples with 0% and 20% of Cr3+ ions (Fig. 4k). As can be seen, the maximal intensity of the 4 A2 → 4 T1 band of Cr3+ in the 20% Cr3+ co-doped sample is more than 2 times higher than that of the narrow 4 I9/2 → 4 F5/2 , 2 H9/2 transitions of Nd3+ . To qualitatively compare the Nd3+ emission at 1064 nm, white LED excitation was chosen to efficiently excite both Nd3+ /Er3+ and Cr3+ bands. Quantitative spectroscopic measurements show that for an excitation power density of 930 mW/cm2 , the absolute emission brightness at 1064 nm increased from 5 × 1012 photons/s/mg for a sample without Cr3+ to 1.6 × 1014 photons/s/mg for the phosphor doped with 20% of Cr3+/4+ (Fig. 4l). This meant that for this phosphor, a 30 times brighter emission was obtained for the λem = 1064 nm emission wavelength. This clearly confirmed that both the efficient Cr3+ → Nd3+ , Er3+ energy transfer and the larger absorption cross-section of Cr3+ ions than that of Nd3+ /Er3+ ions favourably influence the thermometric properties of this type of luminescent thermometers. In summary, the co-doping of Ln3+ based luminescent thermometers with TM ions may beneficially influence the thermometric properties of this type of temperature sensors by enhancing the emission brightness and relative sensitivity. Additionally, such a strategy extends spectral range of optical excitation. However, in order to match the appropriate TM ions, it is necessary to determine in which spectral region the absorption bands of TM ions are located, since they have a high susceptibility to the crystal field of the host material. In this way, the energy mismatch between the energy levels of the TM and Ln3+ ions can be determined to predict whether energy transfer will be efficient. Ultimately, it is necessary to optimise the concentration of dopant ions for which the TM → Ln3+ energy transfer processes are most favourable and dominant over the parasitic Ln3+ → TM processes. Moreover, TM ions have special requirements for the crystallographic position they can occupy. Therefore, not all host materials are appropriate for TM ions.

New Strategies to Improve Thermal Sensitivity and Temperature …

87

5 Negative Expansion Coefficient It is well known that the distance between Ln3+ dopant ions in the host material can influence the interionic energy transfer probability and thus the relative sensitivity of the luminescent thermometer. Therefore, it is necessary to adjust the dopant concentration to ensure an optimum distance between the luminescent centres and thus make the thermometric performance of the thermometer more efficient. However, even for a given dopant concentration, the interionic distance is not temperature independent. For the vast majority of materials, due to the thermal expansion of the solid states, the interionic distance begins to increase at elevated temperatures. In such cases, the so-called thermal expansion parameter reaches the positive values (α > 0) [positive thermal expansion (PTE)]. However, there is a special group of host materials in which this distance is shortened as the temperature increases. In such a case, the negative value of the thermal expansion coefficient (α < 0) is observed. The effect of negative thermal expansion (NTE) provides great opportunities for fine-tuning the thermal behaviour of functional materials (Fig. 5a). Host lattice shrinkage at high temperature may induce a decrease in the distance between the sensitizer and the activator, enhancing the energy transfer efficiency, which is especially favourable in the case of up-converting phosphors. Most of such examples in the literature concern molybdates and tungstates doped with Ln3+ ions. Among them, compounds of the general chemical formula RE2 3+ (M6+ O4 )3 or Zr4+ (M6+ O4 )2 (where RE refers to rare earth ions and M = Mo6+ or W6+ ) have been particularly intensively investigated for applications in luminescent thermometry. Interestingly, all of these host materials crystallise in orthorhombic symmetry and their structures consist of corner-sharing polyhedra (Fig. 5b). It was found that the NTE phenomenon in these structures is attributed to transverse vibrations of the crystal lattice in the direction perpendicular to the RE/Zr-O-M bonds [49]. Consequently, as the temperature increases, the angle between the polyhedra changes, resulting in a reduction of the unit cell size. This is shown in the visualisation of the thermal contraction of Sc2 (MoO4 )3 crystallized in the Pnab space group presented by Liao et al. [50] (Fig. 5c). It is worth noting that Sc2 (MoO4 )3 contains the smallest of all the metallic ions of the RE3+ family, which in this structure are located in the ScO6 octahedra, which are corner-sharing oxygen with the MoO4 tetrahedra. However, as can be seen, this is a special case of an NTE crystal, as Sc2 (MoO4 )3 is characterised by a unique two-dimensional NTE coefficient with αa = −8.62 × 10–6 K−1 , αb = 4.25 × 10–6 K−1 and αc = −6.35 × 10–6 K−1 . It indicates that both the a-axis and c-axis contract while the b-axis expands at elevated temperature, which results in reciprocal rotation at higher temperature and thus in the lattice shrinkage. This was confirmed by Rietveld refinement calculations performed as a function of temperature (Fig. 5d). This feature directly affects the distances between the Yb3+ and Er3+ dopant ions, which will approach each other with increasing temperature. Moreover, it should be noticed that the presence of dopant ions enhances the volumetric shrinkage coefficient (αv = −10.70 × 10–6 K−1 ) of Sc2 (MoO4 )3 :Yb3+ , Er3+ comparing to the un-doped counterpart (αv = −6.30 × 10–6 K−1 ) [50], due to the substitution of Sc3+ by Ln3+

88

L. Marciniak et al.

dopants (i.e., Yb3+ and Er3+ ) with a larger radius, which further implies that doping with Ln3+ ions can lead to the control of cell shrinkage efficiency. This is consistent with example values from D.J. Fisher’s book, where e.g. in the case of tungstates, for Sc2 (WO4 )3 αa = −6.50 × 10–6 K−1 , αb = 5.63 × 10–6 K−1 and αc = −5.74 × 10–6 K−1 were obtained, while for Yb2 (WO4 )3 it is αa = −10.20 × 10–6 K−1 , αb = − 2.65 × 10–6 K−1 and αc = −6.41 × 10–6 K−1 [51]. As could be seen both in the aforementioned extracts and later in this section, in luminescent thermometers based on NTE materials, the Yb3+ -Er3+ ion pair has been most extensively applied as dopants. In Sc2 (MoO4 )3 :Yb3+ , Er3+ , the overall intensity of the green emission of 2 H11/2 → 4 I15/2 and 4 S3/2 → 4 I15/2 increases 26-fold times up to 473 K, while in the range from 298 to 773 K there is a 45-fold increase [50]. However, in the same material and temperature range (298–773 K), a 450-fold increase of the integral luminescence intensity in the NIR spectral range at 1540 nm originating from the 4 I13/2 → 4 I15/2 electronic transition was also reported (Fig. 5f). This effect was explained in terms of a reduction of the distance between Yb3+ and Er3+ , which results in an enhancement of the Yb3+ -Er3+ energy transfer efficiency from 29 to 55% in the temperature range from 298 to 673 K. The same process was also responsible for elongation of the average lifetimes τavr of the 4 I11/2 level of Er3+ with temperature from 61.2 μs at 298 K to 2004 μs at 473 K, and even to 7789 μs at 698 K. Hence, the lifetime-based thermometer reached a maximum sensitivity SRmax of 13.4%/K at about 300 K. As shown by Lv et al. [52], the beneficial influence of the NTE host material on the thermometric properties of luminescent thermometers was also observed for other pairs of Ln3+ ions, like Yb3+ -Tm3+ or Yb3+ -Ho3+ . The authors proved that in Y2 (MoO4 )3 :Yb3+ , Tm3+ the largest variation of LIR values as a function of temperature was obtained for emissions arising from 3 F2 , 3 F3 → 3 H6 to 1 G4 → 3 H6 transitions. In the 303–503 K temperature range a maximal value of SA = 0.0317 K−1 at 503 K was obtained, while SRmax was 3.27%/K at 303 K (Fig. 5i). This is so far the highest sensitivity obtained for LIR-based luminescence thermometers in NTE materials. Moreover, a δT value in this thermal range was as low as 0.01–0.07 K. In Sc2 (MoO4 )3 :2% Ho3+ , 18% Yb3+ [53] as the temperature increased, two processes occurred simultaneously. Due to cell contraction and shortening of the Yb3+ to Ho3+ distance, the energy transfer efficiency intensified, however, additionally, by thermally enhanced probability of multiphonon relaxation in Ho3+ ions, an increase in the population of 5 F5 levels occurred. As a result, there was a ~fivefold enhancement in the intensity of the red emission band at 660 nm (5 F5 → 5 I8 transitions) in the 303–573 K temperature range, while a 50% decrease in the intensity for the green emission band (5 F4 , 5 S2 → 5 I8 transitions) was simultaneously observed. Consequently, the absolute sensitivity of LIR of the 550 and 660 nm bands reached SAmax of 0.0275 K−1 at 543 K. However, it should be noted that the thermal contraction of the host material does not result in such a high sensitivity when the ratio of Tb3+ , Eu3+ emission intensities in Y2 (MoO4 )3 was analysed [54]. The maximal relative sensitivity achieved in this case was SR = 0.32%/K at 363 K, which is not a spectacular performance. In the case of Yb2 (WO4 )3 :6% Er3+ , in which Yb3+ ions not only form the structure but can also act as a sensitiser for Er3+ ions, upon λexc = 980 nm, the intensity of the

New Strategies to Improve Thermal Sensitivity and Temperature …

89

Fig. 5 Schematic representation of the mechanism of thermal contraction of the distance between ions (a) [50] and in unit cell (b) [60]; 3D visualization of thermal contraction of Sc2 (MoO4 )3 :Yb3+ /Er3+ unit cell (c) [50]; thermal dependence of unit cell parameters in the Sc2 (MoO4 )3 :20% Yb3+ , 1% Er3+ phosphor (d) [50]; two-dimensional topographical mapping of VIS emission (e) [50] and NIR emission of Er3+ within the temperature from 25 to 500 °C in Sc2 (MoO4 )3 :Yb3+ /Er3+ (f) [50]; energy level diagram of Yb3+ and Er3+ ions (g) [50]; thermal evolution of the integral intensities of emission bands 2 H11/2 → 4 I15/2 (H1 and H2) and 4 S3/2 → 4 I15/2 (S1 and S2) of Er3+ ions in Y2 (MoO4 )3 :Yb3+ , Er3+ (h) [60]; SA and SR values of Y2 (MoO4 )3 :1% Tm3+ , 13% Yb3+ particles based on the LIR (i) [52]; photographs of Sc2 (MoO4 )3 :Yb3+ /Er3+ phosphors (j) [50] and Sc2 (MoO4 )3 :Yb3+ /Ho3+ microcrystals at various temperatures upon 980 nm excitation (k) [53]

emission band at 523 nm (2 H11/2 → 4 I15/2 ) increases 29-fold, while only a threefold enhancement was observed for the 660 nm emission band (4 F9/2 → 4 I15/2 ) in the 303– 573 K temperature range [55]. Considering the LIR of these two bands, the maximal SA = 0.105 K−1 was found at 573 K. Another interesting example of a tungstate is Lu2 (WO4 )2.5 (MoO4 )0.5 (also written as Lu2 W2.5 Mo0.5 O12 ), for which Peng et al. [56] reported a giant NTE coefficient of α = −20.0 × 10–6 K−1 in the range of 200–800 °C (α = −18.6 × 10–6 K−1 and α = −16.9 × 10–6 K−1 for Lu2 (WO4 )3 or Lu2 (MoO4 )3 ,

90

L. Marciniak et al.

respectively). Encouraged by this result, Cui et al. [57] investigated the luminescence properties of Lu2 (WO4 )2.5 (MoO4 )0.5 :2% Er3+ , x% Yb3+ (where x = 20, 30, 40). In this case, an enhancement of the Er3+ emission upon λexc = 980 nm in the range from 300 to 700 K was also observed. The absolute sensitivity of the ratiometric luminescent thermometer based on LIR of 2 H11/2 → 4 I15/2 relative to 4 S3/2 → 4 I15/2 was found to depend on the Yb3+ concentration and reached SA = 0.00741 K−1 at 409 K, SA = 0.00744 K−1 at 405 K and SA = 0.00723 K−1 at 391 K for 20%, 30% and 40% of Yb3+ ions, respectively. However, it should be noted that in this host material, due to the NTE, intensities of both these bands increased at elevated temperature. This is the opposite situation to that usually observed in the case of such a thermometer, when the intensity of the 2 H11/2 → 4 I15/2 emission band increases at the expense of the 4 S3/2 → 4 I15/2 one. In the case of the Zr2 (WO4 )(PO4 )2 host, due to the much smaller NTE coefficient of α = −3.4 × 10–6 K−1 , such spectacular increases in band intensities were not observed [58]. Hence, the maximal SR in this case was only 1.05%/K for Zr2 (WO4 )(PO4 )2 :2% Er3+ , 2% Yb3+ , which was further enhanced slightly by the introduction of Li+ ions to a value of SR = 1.20% /K at 300 K. The role of the Li+ ions on the enhancement of the relative sensitivity and upconversion emission intensity has been discussed in terms of reduction of structural defects associated with the difference in ionic charge between Zr2+ and Ln3+ ions. A separate group of materials revealing the NTE effect that has not been mentioned previously, but which can also exhibit shrinkage with temperature, are halides. This effect has been reported for ScF3 , TiF3 and ZnF2 [51], which became the motivation for Ren et al. [59] to synthesize cube-shaped ScF3 :18% Yb3+ , 2% Er3+ @ScF3 nanoparticles and investigate their thermometric properties. The 3.7-fold enhancement in the intensity of the 4 S3/2 → 4 I15/2 band was found when temperature increased in the 168–248 K temperature range. This positive luminescence thermal coefficient is an indisputable advantage compared to classical NaYF4 :Yb3+ , Er3+ nanoparticles, where a decrease in signal intensity is found for this emission band. Then, taking advantage of the fact that 2 H11/2 and 4 S3/2 levels of Er3+ ions are thermally coupled, it was noticed that the intensity ratio of bands at 530 nm and 545 nm was strongly susceptible to temperature changes in the range of 168–308 K. Therefore, the thermal evolution of LIR for these two bands was determined and the SR was then calculated. A Maximal relative sensitivity of SR = 1.73% K−1 at 168 K was obtained. In summary, NTE materials are a very promising and still unexplored group of materials, especially interesting from the thermometric perspective. A particular advantage that distinguishes this type of material over others is that, as the temperature increases, the signal is amplified instead of quenched, preventing the high temperature uncertainty δT associated with the low noise-to-signal ratio. To take full advantage of the benefits of these materials, the dopant ions should be properly selected and currently the best identified pair of ions is Yb3+ -Ln3+ (mostly Yb3+ Er3+ ). Nevertheless, in order to obtain satisfactory thermometer sensitivities, a large variation in signal intensity is necessary, which has been the subject of many studies and which can be observed, for instance, in the photographs taken for Sc2 (MoO4 )3 : Yb3+ , Er3+ upon 980 nm laser diode excitation or under the same excitation where, with a constant Ho3+ emission signal from the green range (at 550 nm), the intensity

New Strategies to Improve Thermal Sensitivity and Temperature …

91

of the red emission (at 660 nm) increases significantly, as manifested by the effective thermochromism of the Sc2 (MoO4 )3 :Yb3+ , Ho3+ sample (Fig. 5k). In addition to the type of dopant ions, their concentration should be carefully selected, taking into account their influence on the NTE of the host material. It should be also noted that a large number of the host materials that reveal NTE are hygroscopic, as observed among others for Y2 (WO4 )3 , Yb2 (WO4 )3 and Y2 (MoO4 )3 , basing on thermogravimetric analysis (TGA) [53, 55, 60]. The problem is substantial because absorption of water molecules occurs in the crystallographic voids, leading to an easy loss of the NTE properties [61]. It happens because when the NTE crystals are hydrated, water molecules can hinder rocking motion and thus inhibit the contraction of the crystal lattice upon heating. Moreover, from the luminescence point of view, it is of great importance the well-known fact that water molecules are high-energy oscillators, which significantly quench luminescence [62]. Therefore, the presence of water is very undesirable in luminescence thermometry because it not only causes a weakening of the signal intensity, but it can also lead to an erroneous temperature readout due to the removal of OH− groups during the drying of the material upon heating.

6 Structural First Order Phase Transition Derived Luminescent Thermometers As is well known, the host material in which the Ln3+ ions are located has a fundamental influence on their spectroscopic properties and on the thermometric properties of Ln3+ ion-based luminescence thermometer. In the vast majority of host materials, material parameters such as phonon energy, crystal field strength, crystal symmetry, or local symmetry, i.e., the immediate environment in which the dopant is located, among others, are thermally unchanging, however, in some cases, broad structural polymorphism and temperature-induced first-order structural phase transitions can lead to changes in these parameters, significantly affecting the luminescence properties of Ln3+ ions. When such a transition takes place, a symmetry breaking process often occurs, which directly affects the spectroscopic properties of materials doped with Ln3+ ions, such as the number of Stark sublevels determined by the crystal field splitting, the probabilities of radiative transitions, and the strength of multiplets splitting. Such a phenomenon can be observed for LiYO2 crystals, which crystalizes in a monoclinic (P21 /c) and tetragonal (I41 /amd) symmetries (Fig. 6a). The transition between these phases occurs at around 363 K in bulk materials [63], however, as the particle size of the crystals decreases, the transition temperature is reduced to around room temperatures [64–66]. Since the Y3+ sites can be occupied by Ln3+ ions, this material has recently received considerable interest in thermometric application. As reported by Wang et al. [66] based on differential scanning calorimetry (DSC) analysis in LiYO2 :0.1% Pr3+ microcrystals, the transition from the low-temperature phase by heating occurs in the 320–330 K range, while the return by cooling occurs

92

L. Marciniak et al.

in the range of 310–320 K. In contrast, in the work of Marciniak et al. [65] the same measurement indicated maxima at 293.6 and 308.5 K during heating and 288.8 and 302.2 K during cooling for LiYO2 :Nd3+ nanocrystals. The two maxima for each measurement in this case refer to two fractions of particles size with average diameters of 30 and 400 nm. This transition is also reflected in the structural measurements (Fig. 6b). As can be seen, regardless of the direction of phase transition, there is a temperature range where both phases coexist. As can be seen in Fig. 6c, initially at 223 K the tetragonal structure was observed to be only 5% of the weight fraction, and the remaining part was attributed to the monoclinic structure, however, with increasing temperature, the contribution of the tetragonal phase gradually increased to the detriment of the monoclinic phase, and at 313 K it covered practically the entire structure and the monoclinic phase was no longer observed above this threshold temperature of 313 K. A perfect tool for in situ structural analysis is Eu3+ ions, which are often named luminescent structural probe. This is due to the fact that the emission bands related to the magnetic dipole transition (5 D0 → 7 F1 ) are independent of structural changes in contrast to the electric dipole transitions (5 D0 → 7 F2 ) [67]. It was found that even small variations in temperature result in significant changes in the shape of the emission spectra associated with the change in the local point symmetry of the Y3+ site substituted by the Eu3+ ions. Although the C 2 symmetry was found in the low temperature monoclinic structure, at elevated temperatures an increase in the symmetry to D2d was observed. This significantly changes the number of Stark components into which the 7 FJ states were split (three components in the case of C 2 symmetry comparing to two components for D2d ). On the other hand, in the case of LiYO2 doped with Nd3+ ions, slightly different changes in the emission spectra were observed [65] (Fig. 6d). The emission spectra of Nd3+ ions consist of three bands, which are typically located at about 880 nm (4 F3/2 → 4 I9/2 ), 1060 nm (4 F3/2 → 4 I11/2 ) and 1350 nm (4 F3/2 → 4 I13/2 ). However, for Nd3+ ions in lower than cubic symmetry, the number of Stark components is constant for all states, namely two Stark components for 4 F3/2 and five, six and seven components for 4 I9/2 , 4 I11/2 , and 4 I13/2 , respectively. In this case, the strength of the crystal field splitting plays a crucial role in determining the energy that will cause the splitting of sublevels originating from a single multiplet. Since this strength depends on the point symmetry in which the Nd3+ ions are located, the emission spectra performed for monoclinic and tetragonal symmetry differ significantly. A similar result was obtained in LaGaO3 :3% Nd3+ , where a transition from orthorhombic (Pnma) to rhombohedral (R-3c) symmetry was observed at 500 K, resulting in a change in the local symmetry of the Nd3+ ions from C s to D3 , and thus a change in the shape of the emission spectrum [68]. All these thermally induced changes in the luminescent properties of Ln3+ can be successfully utilized in luminescence thermometry. For Pr3+ ions in LiYO2 , upon λexc = 280 nm excitation, four emission bands at around 500 nm, 560 nm, 630 nm and 665 nm were observed, that can be assigned to the 3 P0 → 3 H4 , 3 P0 → 3 H5 , 3 P0 → 3 H6 transitions superimposed spectrally with 1 D2 → 3 H4 and 3 P0 → 3 F2 transitions. In this case, similar to the Eu3+ and Nd3+ doped counterparts, a change in the shape of the emission band around 320 K can be

New Strategies to Improve Thermal Sensitivity and Temperature …

93

Fig. 6 Visualization of monoclinic (M-LiYO2 ) and tetragonal (T-LiYO2 ) structures (a) [65]; partial XRPD data with the rise and decline in temperature for LiYO2 :Pr3+ (b) [66]; the contribution of the M-LiYO2 and T-LiYO2 in the LiYO2 :1% Nd3+ phosphor as a function of temperature (c) [65]; the comparison of the emission spectra of LiYO2 :1%Nd3+ nanocrystals measured at 223 K (blue line) and 323 K (red line) (d) [65]; temperature-dependent emission spectra of the LiYO2 :0.25% Pr3+ sample (λexc = 280 nm) (e) [66]; the thermal dependence of the selected emission intensities in the spectral ranges of the 4 F3/2 → 4 I9/2 transition in LiYO2 :1%Nd3+ (f) [65]; temperature dependence of the emission spectra of the BaTiO3 :1% Ho3+ , 1% Yb3+ sample (g) [69]; relative sensitivity SR versus temperature of LiYO2 :0.25% Pr3+ (h) [66]; the δT determined based on LIR for LiYO2 :1%Nd3+ (i) [65]; tunability of the phase-transition temperature calculated from LIR value in two different ways as a function of the average A-site cation average radii in LaGaO3 :1% Nd3+ , x% Gd3+ (j) [68]; LIR versus temperature profiles as a function of Gd3+ content in LaGaO3 :1% Nd3+ , x% Gd3+ (x = 0, 1, 2, 4) (k) [68]; thermal dependence of the SR (LIR) associated with the M-LiYO2 :Nd3+ and T-LiYO2 :Nd3+ for different concentrations of Nd3+ ions (l) [65]

94

L. Marciniak et al.

observed, but due to the spectrally overlapping emission lines, it is experimentally difficult to unambiguously assign them to specific Stark sublevels. On the other hand, even more impressively, above 325 K, a sharp decrease in the intensity of the emission deriving from the 3 P0 → 3 H4 transition in respect to the 1 D2 → 3 H4 one was observed (Fig. 6e), which can be explained in terms of activation of phonon-assisted {3 P0 , 3 H4 } → {1 D2 , 3 H6 } and {3 P0 , 3 H4 } → {1 G4 , 1 G4 } cross-relaxations. This observation indicates that the LIR of these two bands [3 P0 → 3 H4 (470–525 nm) and 1 D2 → 3 H4 (565–690 nm)] can be used as a thermometric parameter. The maximal sensitivity in this case reached SRmax = 23.04%/K at 329 K (Fig. 6h). Such high sensitivity in the approximately physiological temperature range provides high application potential for this material. A similar abrupt increase in sensitivity was determined in LiYO2 :Eu3+ for LIR based on the Stark lines of the 5 D0 → 7 F1 emission band. By reducing the number of emission bands (from 3 to 2) associated with the Stark sublevels, a maximum relative sensitivity of 11.8% K−1 at 299 K was achieved. The inverse monotonicity of emission lines associated with the electronic transitions between Stark components was also observed in LiYO2 :Nd3+ . Since such changes were observed in all three spectral regions characteristic of Nd3+ ions, this enabled the development of three ratiometric luminescent thermometers operating in the spectral regions of the I and II optical biological windows upon excitation with a wavelength from the I optical window. As it turned out, the largest changes were obtained for the bands within the 4 F3/2 → 4 I9/2 transition (Fig. 6f), The maximal values of SR were found to be SR = 5.06%/K (4 F3/2 → 4 I9/2 ), SR2 = 5.85%/K (4 F3/2 → 4 I11/2 ) and SR3 = 4.05%/K (4 F3/2 → 4 I13/2 ) at 290 K. Although these values are relatively high, the practical application of such a thermometer may be demanding due to the close spectral separation between considered emission lines. Therefore, among the analysed bands, those characterized by extremely opposite monotonicity (spectral range between 942 and 955 nm of 4 F3/2 → 4 I9/2 and 1066 and 1073 nm of 4 F3/2 → 4 I11/2 ) were selected for which enhancement of the sensitivity up to SRmax = 6.5%/K at 290 K was achieved. Another important parameter evaluating the application potential of a luminescent thermometer for temperature determination is the temperature resolution δT (Fig. 6i). Due to the high SR and high emission brightness, the δT < 1 K was maintained below 400 K, and the lowest temperature uncertainty, which reached δT = 0.03 K, was observed at temperatures of the phase transition of ~290 K that reached. In the case of LaGaO3 :Nd3+ , although a structural phase transition was observed over the temperature range analysed, the LIRs of thermally coupled levels (4 S3/2 , 4 F7/2 ; 4 F5/2 , 2 H9/2 ; 4 F3/2 ) were analysed as thermometric parameters, yielding rather typical values of relative sensitivity of 1.8% K−1 at 300 K (4 S3/2 , 4 F7/2 → 4 I9/2 to 4 F3/2 → 4 I9/2 ) and 1.45% K−1 (4 F5/2 , 2 H9/2 → 4 I9/2 to 4 F3/2 → 4 I9/2 ) at 440 K. Despite the high values of relative sensitivities obtained, attention should be drawn to the narrow temperature range in which such phase transition based luminescent thermometer can be applied. It is strongly dictated by the phase transition temperature of the material in which the luminescence center is located. In order to address this issue, several possibilities for tuning the range over which the thermometer achieves such high relative sensitivity have been proposed. Back et al. [68] noted that in a

New Strategies to Improve Thermal Sensitivity and Temperature …

95

LaGaO3 sample doped with 3% of Nd3+ , the phase transition occurs at about 500 K, which differs from the value reported in the literature, which describes LaGaO3 as a material characterized by a first-order phase transition at about 418 K. Therefore, it can be concluded that the type and concentration of dopant affect the temperature of phase transition. In this particular case, doping with Nd3+ ions, which are essentially smaller than La3+ ions (110.9 pm and 116 pm for eightfold coordinated Nd3+ and La3+ , respectively), contributed to an increase in the transition temperature. This became their motivation to perform measurements with the LaGaO3 :1% Nd3+ sample co-doped with 1%, 2% and 4% of Gd3+ ions (Fig. 6j). The findings of these experiments revealed that for smaller Gd3+ ions (effective ionic radii of eightfold coordinated Gd3+ is 105.3 pm), the transition temperature increased even more rapidly in an approximately linear sequence, namely 490, 550, 670 K for LaGaO3 :1% Nd3+ co-doped with 1%, 2% and 4% of Gd3+ in respect to 450 K for the singly Nd3+ ion-doped counterpart. Moreover, when the concentration of Gd3+ ions increased, the temperature at which the maximum value of LIR was reached shifted (Fig. 6k), whereas simultaneously the LIR value decreased. This clearly affects the maximum obtained relative sensitivity of such a thermometer. A similar effect of increasing the Nd3+ dopant ion concentration on increasing the temperature at which maximal SR was achieved, but also a simultaneous decrease in the value if relative sensitivity was observed in LiYO2 :Nd3+ (Fig. 6l). Thus, for LiYO2 :0.1%Nd3+ powder the value of SRmax = 7.9%/K at 291 K was recorded, while for LiYO2 :5%Nd3+ powder SRmax = 4.5%/K at ~312 K was found. Nevertheless, it can be concluded that the range in which SR was sufficiently high (SR > 1%/K) expanded with increasing dopant concentration. Therefore, it can be said that the optimization of the Ln3+ ion concentration in materials exhibiting a phase transition has a completely different and unusual purpose in respect to other approaches. In this case, the focus is not on improving energy transfer efficiency while maintaining high brightness, but on the effect of concentration on the relative sensitivity of the thermometer and the temperature at which the transition takes place. An interesting approach to phase-transition driven luminescent thermometers based on intensity ratio in BaTiO3 :1% Ho3+ , 1% Yb3+ , which undergoes a phase transition from a tetragonal (P4mm) to a cubic (Pm-3 m) structure at about 423 K was proposed by Zheng et al. [69]. Under 976 nm excitation, the authors observed an up-conversion signal centered at 550 and 650 nm, associated with the 5 S2 , 5 F4 → 5 I8 and 5 F5 → 5 I8 transitions of Ho3+ ions, and a sharp band associated with the second harmonic generation (SHG) process, centered at 488 nm (Fig. 6g). In this case, as a result of the phase transition, there is a change in the strength of the crystal field splitting, resulting in a change of the shape of the emission bands of Ho3+ ions, however, an additional effect is the transition from broken inversion symmetry to inversion symmetry, which strongly affects the intensity of SHG signal. Due to the fact that the SHG process is forbidden in the centrosymmetric cubic phase of BaTiO3 , a rapid decrease of SHG intensity occurs around the phase transition temperature (~423 K). At this time, the overall intensity of the bands associated with 5 S2 , 5 F4 → 5 I8 and 5 F4 → 5 I8 was monotonically quenched. This gives a luminescence thermometer based

96

L. Marciniak et al.

on the LIR of SHG and Ho3+ emission with a relative maximum sensitivity SRmax of 2.78%/K at 419 K. Finally, it must be mentioned that, despite the impressive sensitivities and uncertainties, a hysteresis loop occurring for the LIR values within heating and cooling cycles for phase transition-driven luminescent thermometers is nevertheless an important issue to pay attention to. It results from the difference between the endothermic and exothermic reactions occurring during heating and cooling routes, respectively, which has also been confirmed by DSC measurements. Therefore, it is still possible to use phase transition materials for high-sensitivity temperature imaging, but taking into account that the monotonic change of temperature in the system (only if temperature rises or only if temperature falls) must be maintained. In summary, luminescent thermometers based on phase transitions are characterized by very high relative sensitivity and low temperature determination uncertainty. It is particularly valuable that thermometers operating in the physiological temperature range have already been obtained. The main parameter defining the application of this type of material is transition temperature, which can be controlled not only by changing the host material itself, but also by the crystal size determined by the synthesis and appropriate doping. The choice of the spectral range in which the thermometer would operate is dictated by the selection of suitable Ln3+ ions, but is not in any way limited by the concept itself. It is even possible to use two Stark sublevel bands related to single electronic transition to create an LIR with high variability, but then one must bear in mind the need to select appropriate narrowband filters. When using the single dopant ion band approach, it can be advantageous in terms of scaling up the synthesis process to a larger one used in industry when there is a risk that the two dopant ions will not be distributed equally in the material and will form clusters. However, this approach has two major limitations. The first is the narrow temperature range, limited to the thermal vicinity of the phase transition temperature. Therefore, such a luminescent thermometer needs to be optimized according to the requirements of particular application. The second limitation in the hysteresis loop of thermal changes of LIR. This limits the applicability of such a thermometer to application in which single monotonicity of temperature changes is expected. An example of such an application is the analysis of heat diffusion in solids [64].

7 Conclusions and Perspectives Although work devoted to the use of Ln3+ ions in luminescence thermometry has been carried out for decades, as shown in this chapter, recent years continue to abound with new developments and strategies to improve their thermometric properties. The use of new structural effects induced by temperature change (host materials with negative thermal coefficient and phase transition-based thermometry) and processes whose probability changes significantly as a function of temperature (excited state absorption, sensitization, multilevel thermalization) allows for an increase in relative sensitivity and/or an increase in emission brightness, thereby reducing the inaccuracy

New Strategies to Improve Thermal Sensitivity and Temperature …

97

of temperature determination. Each of these solutions has advantages and disadvantages, which are presented schematically in Fig. 7, together with the determination of the maximum relative sensitivities obtained for each of these solutions. In the case of sensitization, as shown in the present work, the use of TM ions as sensitizers of Ln3+ ions luminescence leads not only to a tens of times enhancement of the emission intensity, a broadening of the spectral range of optical excitation, but also to an increase of the relative sensitivity. However, in this case, the concentration of TM ions should be optimized in order to avoid back energy transfer, which will negatively affect the thermometric performance. Additionally, TM ions can be successfully introduced into the host material only if certain requirements are met regarding the specific size of the crystallographic positions of host and their coordination. Hence, this may be a limitation of the number of available host materials used in this strategy. The maximum relative sensitivities achieved in this case are not very

Fig. 7 Schematic representation of the advantages and disadvantages of all described in this chapter approaches in Ln3+ -based luminescent thermometers with the maximal relative sensitivities reported for each of them

98

L. Marciniak et al.

high, but significantly higher than the counterparts undoped with TM ions. It can be speculated that by applying this solution to high sensitivity thermometers, the gain obtained when TM ion are inserted can result in very high values of SR . Thus, this is a direction worth exploring in depth in the future. Luminescent thermometers based on materials exhibiting negative thermal expansion coefficient are a small group of thermometers due to the limited number of materials that exhibit this effect. A second obstacle, besides the limited number of available materials manifesting this effect, is their hygroscopicity. Nevertheless, this is undoubtedly an attractive research direction which has not been thoroughly investigated so far. As shown, its application results in high sensitivities of ratiometric thermometers, but the highest sensitivity was reported for a thermometer based on the kinetics of luminescence. Surprisingly, this type of research is marginal to the work published so far, although the first results are very promising. In the case of thermometers based on excited state absorption, the main limitation is that the density of optical excitation reaching the thermometer has a significant effect on the temperature readout. The fact that such a solution exploits two excitation wavelengths, for which the light extinction coefficient in many media can be significantly different, can make the temperature readout unreliable, e.g., in biological systems. This aspect should be kept in mind when using this type of thermometers. However, as published studies have shown, in media such as air, or using ESA in conventional ratiometric thermometers, very high accuracy can be obtained not only for sensing but also for thermal imaging. The significant influence of the host material parameters on the probability of ESA processes enables optimization of the thermometric parameters of a luminescent thermometer over a fairly large extent. Importantly, the luminescence obtained with ESA excitation exhibits a positive thermal coefficient, i.e., an increase in the emission intensity with increasing temperature is observed, which is an important advantage to obtain a high signal-to-noise ratio even at high temperatures. The use of a multilevel system of thermally coupled states not only increases the relative sensitivity, but also extends the useful temperature range over which the thermometer can be used. Importantly, this solution does not require modification of the phosphor, but only a change in the emission bands used to determine the temperature. However, a significant limitation of this approach affecting, as confirmed in several studies, the deterioration of the uncertainty of temperature determination is the low intensity of emission from the highest thermally coupled level. The materials in which temperature-induced phase transitions have been observed represent a group of thermometers in which very high relative sensitivity values have been reported (the highest among all those discussed in this chapter), which is a significant advantage of this approach. In addition, this is a single-ion approach, which facilitates the scaling up of mass production of the thermometer, eliminating the risk of inhomogeneous ions distribution in the host material which can affect the thermometric properties of the co-doped thermometer. A limitation of this solution, however, is that the temperature range over which the temperature can be read by using this solution is usually very narrow. It is possible to change the phase transition temperature of the phosphor, and thus the useful range of the thermometer, by

New Strategies to Improve Thermal Sensitivity and Temperature …

99

changing the chemical composition of the host. This gives the possibility to design highly sensitive thermometers dedicated to specific application. However, the fact that there is a hysteresis in the LIR parameter implies that such thermometers can only be used for monotonic temperature variation. The increase in the number of new developments in the field of Ln3+ -based luminescent thermometry observed in recent years allows to assume that the coming years will bring solutions to at least some of the above-mentioned limitations, allowing for increased reliability and accuracy of temperature readouts using this technique. This leads us to believe that luminescent thermometry has a bright future and the potential to become one of the most widely used techniques for temperature readout and imaging. Acknowledgements This work was supported by the National Science Center (NCN) Poland under project no. 2018/31/G/ST5/03258.

References 1. W. Zhang, L. Zheng, W. Xu, L. Li, Z. Zhang, Boosting the sensitivity of Er3+ -based luminescent ratiometric thermometry. Opt. Mater. (Amst) 106, 110002 (2020) 2. L. Li, F. Qin, Y. Zheng, Z. Zhang, Strategy for highly sensitive optical ratiometric temperature measurement. Opt. Mater. Express 9(8), 3260 (2019) ´ c, T. Gavrilovi´c, M.D. Drami´canin, Luminescence intensity ratio thermometry with 3. A. Ciri´ Er3+ : performance overview. Crystals, pp. 1–19 (2021) - c, M. Medi´c, J. Periša, I. Zekovi´c, M. ´ c, J. Aleksi´c, T. Barudžija, Ž. Anti´c, V. Ðordevi´ 4. A. Ciri´ Mitri´c, M.D. Drami´canin, Comparison of three ratiometric temperature readings from the Er3+ upconversion emission. Nanomaterials, pp 1–10 5. L. Li, F. Qin, Y. Zhou, Y. Zheng, H. Zhao, Z. Zhang, Temperature sensing based on the 4 F / 4 S −4 I 3+ 3+ nanocrys7/2 3/2 15/2 upconversion luminescence intensity ratio in NaYF4 :Er /Yb tals. J. Lumin. 206, 335–341 (2019) ´ c, J. Periša, I. Zekovi´c, Ž Anti´c, M.D. Drami´canin, Multilevel-cascade intensity ratio 6. A. Ciri´ temperature read-out of Dy3+ luminescence thermometers. J. Lumin. 245, 118795 (2022) 7. L. Li, F. Qin, Y. Zhou, Y. Zheng, J. Miao, Z. Zhang, Three-energy-level-cascaded strategy for a more sensitive luminescence ratiometric thermometry. Sens. Actuators, A Phys. 304, 111864 (2020) 8. D. Yu, H. Li, D. Zhang, Q. Zhang, A. Meijerink, M. Suta, One ion to catch them all: targeted high-precision Boltzmann thermometry over a wide temperature range with Gd3+ . Light Sci. Appl. 10(1), 236 (2021) 9. H. Song, Q. Han, X. Tang, X. Zhao, K. Ren, T. Liu, Nd3+ /Yb3+ codoped SrWO4 for highly sensitive optical thermometry based on the near infrared emission. Opt. Mater. 84, 263–267 (2018) 10. W. Xu, H. Qi, L. Zheng, Z. Zhang, W. Cao, Multifunctional nanoparticles based on the Nd3+ /Yb3+ codoped NaYF4 . Opt. Lett. 40(23), 5678 (2015) 11. W. Xu, Y. Hu, L. Zheng, Z. Zhang, W. Cao, H. Liu, X. Wu, Enhanced NIR-NIR luminescence from CaWO4 : Nd3+ /Yb3+ phosphors by Li+ codoping for thermometry and optical heating. J. Lumin. 208, 415–423 (2019) 12. H. Suo, X. Zhao, Z. Zhang, C. Guo, Ultra-sensitive optical nano-thermometer LaPO4 : Yb3+ /Nd3+ based on thermo-enhanced NIR-to-NIR emissions. Chem. Eng. J. 389, 124506 (2020)

100

L. Marciniak et al.

13. M. Sójka, M. Runowski, T. Zheng, A. Shyichuk, D. Kulesza, E. Zych, S. Lis, Eu2+ emission from thermally coupled levels–new frontiers for ultrasensitive luminescence thermometry. J. Mater.Chem. C 10(4), 1220–1227 (2022) 14. L. Labrador-Páez, M. Pedroni, A. Speghini, J. García-Solé, P. Haro-González, D. Jaque, Reliability of rare-earth-doped infrared luminescent nanothermometers. Nanoscale 10(47), 22319–22328 (2018) 15. A.S. Souza, L.A.O. Nunes, I.G.N. Silva, F.A.M. Oliveira, L.L. Da Luz, H.F. Brito, M.C.F.C. Felinto, R.A.S. Ferreira, S.A. Júnior, L.D. Carlos, O.L. Malta, Highly-sensitive Eu3+ ratiometric thermometers based on excited state absorption with predictable calibration. Nanoscale 8(9), 5327–5333 (2016) 16. K. Trejgis, A. Bednarkiewicz, L. Marciniak, Engineering excited state absorption based nanothermometry for temperature sensing and imaging. Nanoscale 12(7), 4667–4675 (2020) 17. K. Trejgis, F. Tian, J. Li, A. Bednarkiewicz, L. Marciniak, The role of surface related quenching in the single band ratiometric approach based on excited state absorption processes in Nd3+ doped phosphors. Mater. Res. Bull. 139, 111288 (2021) 18. K. Trejgis, K. Ledwa, L. Li, L. Marciniak, Effect of the nanoparticle size on thermometric properties of a single-band ratiometric luminescent thermometer in NaYF4 :Nd3+ . J. Mater. Chem. C 10(8), 3006–3014 (2022) 19. J. Stefanska, K. Maciejewska, L. Marciniak, Blue-emitting single band ratiometric luminescent thermometry based on LaF3 :Pr3+ . New J. Chem. 45(27), 11898–11904 (2021) 20. J. Stefanska, L. Marciniak, M. Chrunik, Sensitivity enhancement of the Tb3+ -based single band ratiometric luminescent thermometry by the metal-to-metal charge transfer process. J. Phys. Chem. C 125(9), 5226–5232 (2021) 21. J. Stefa´nska, A. Bednarkiewicz, L. Marciniak, Advancements in excited state absorption-based luminescence thermometry. J. Mater. Chem. C 10(15), 5744–5782 (2022) 22. K. Trejgis, K. Ledwa, A. Bednarkiewicz, L. Marciniak, Impact of host composition and dopant ion concentration on the thermometric properties of a Eu3+ activated fluoride-based single-band ratiometric luminescent thermometer. J. Alloys Compd. 898, 162839 (2022) 23. D. Duan, Y. Wang, S. Jiang, L. Li, G. Xiang, X. Tang, Y. Li, X. Zhou, An optical thermometry based on abnormal negative thermal quenching of the charge transfer band edge. J. Lumin. 215, 116636 (2019) 24. L.J. Borrero-González, S. Acosta, C. Bittencourt, M. Garvas, P. Umek, L.A. Nunes, Eu3+ −doped titanium oxide nanoparticles for optical thermometry in the first biological window. Opt. Mater. 101, 109770 (2020) ´ c, I. Zekovi´c, M. Medi´c, Ž Anti´c, M.D. Drami´canin, Judd-Ofelt modelling of the dual25. A. Ciri´ excited single band ratiometric luminescence thermometry. J. Lumin. 225, 117369 (2020) 26. I.E. Kolesnikov, M.A. Kurochkin, E.V. Golyeva, D.V. Mamonova, A.A. Kalinichev, E.Y. Kolesnikov, E. Lähderanta, Multimode high-sensitivity optical YVO4 :Ln3+ nanothermometers (Ln3+ = Eu3+ , Dy3+ , Sm3+ ) using charge transfer band features. Phys. Chem. Chem. Phys. 22(48), 28183–28190 (2020) 27. V.S. Zanuto, O.A. Capeloto, R.F. Muniz, M. Sandrini, J.H. Rohling, M.L. Baesso, L.A.O. Nunes, Two ratiometric thermometry methods based on the interplay between Eu2+ and Eu3+ and single Eu3+ emissions on OH−-free low-silica calcium aluminosilicate glass. Mater. Res. Bull. 135, 111115 (2021) 28. K. Trejgis, K. Maciejewska, A. Bednarkiewicz, L. Marciniak, Near-infrared-to-near-infrared excited-state absorption in LaPO4 :Nd3+ nanoparticles for luminescent nanothermometry. ACS Appl. Nano Mater. 3(5), 4818–4825 (2020) 29. K. Trejgis, K. Ledwa, A. Bednarkiewicz, L. Marciniak, A single-band ratiometric luminescent thermometer based on tetrafluorides operating entirely in the infrared region. Nanoscale Adv. 4(2), 437–446 (2022) 30. L. Li, F. Qin, L. Li, H. Gao, Q. Deng, Z. Zhang, Thermal enhancement of the green luminescence of Tb3+ ion generated by ground state thermal population. J. Lumin. 211, 258–263 (2019) 31. J. Drabik, R. Kowalski, L. Marciniak, Enhancement of the sensitivity of single band ratiometric luminescent nanothermometers based on Tb3+ ions through activation of the cross relaxation process. Sci. Rep. 10(1), 11190 (2020)

New Strategies to Improve Thermal Sensitivity and Temperature …

101

32. J. Drabik, L. Marciniak, KLaP4 O12 :Tb3+ nanocrystals for luminescent thermometry in a singleband-ratiometric approach. ACS Appl. Nano Mater. 3(4), 3798–3806 (2020) 33. J. Stefanska, L. Marciniak, Single-band ratiometric luminescent thermometry using Pr3+ ions emitting in yellow and red spectral ranges. Adv. Photonics Res. 2(7), 2170021 (2021) 34. S. Zhou, C. Duan, S. Han, A novel strategy for thermometry based on the temperature-induced red shift of the charge transfer band edge. Dalt. Trans. 47(5), 1599–1603 (2018) 35. C. D. S. Brites, A. Millán, L.D. Carlos, in Lanthanides in luminescent thermometry, ed. by B. Jean-Claude, K.P. Vitalij. Including Actinides, vol. 49 (Elsevier, 2016), pp. 339–427 36. J. Drabik, R. Lisiecki, L. Marciniak, Optimization of the thermometric performance of single band ratiometric luminescent thermometer based on Tb3+ luminescence by the enhancement of thermal quenching of GSA-excited luminescence in TZPN glass. J. Alloys Compd. 858, 157690 (2021) ´ c, Ł Marciniak, M.D. Drami´canin, Luminescence intensity ratio squared—a new 37. A. Ciri´ luminescence thermometry method for enhanced sensitivity. J. Appl. Phys. 131(11), 114501 (2022) 38. W. Piotrowski, L. Dalipi, K. Elzbieciak-Piecka, A. Bednarkiewicz, B. Fond, L. Marciniak, Self-referenced temperature imaging with dual light emitting diode excitation and single-band emission of AVO4 :Eu3+ (A=Y, La, Lu, Gd) nanophosphors. Adv Photonics Res 3(6), 2100139 (2021) 39. L. Li, F. Qin, L. Li, H. Gao, Z. Zhang, New strategy for circumventing the limitation of thermally linked states and boosting the relative thermal sensitivity of luminescence ratiometric thermometry. J. Phys. Chem. C 123(10), 6176–6181 (2019) 40. G. Xiang, X. Liu, Q. Xia, X. Liu, S. Xu, S. Jiang, X. Zhou, L. Li, D. Wu, L. Ma, X. Wang, J. Zhang, Design of a bi-functional NaScF4 : Yb3+ /Er3+ nanoparticles for deep-tissue bioimaging and optical thermometry through Mn2+ doping. Talanta 224, 121832 (2021) 41. Q. Qiang, Y. Wang, Effect of Mn2+ on upconversion emission, thermal sensing and optical heater behavior of Yb3+ - Er3+ codoped NaGdF4 nanophosphors. Front. Chem. (2019) 42. H. Lu, H. Hao, H. Zhu, G. Shi, Q. Fan, Y. Song, Y. Wang, X. Zhang, Enhancing temperature sensing behavior of NaLuF4 :Yb3+ /Er3+ via incorporation of Mn2+ ions resulting in a closed energy transfer. J. Alloys Compd. 728, 971–975 (2017) 43. Y. Guo, J. Xie, M. Yu, W. Huang, H. Yang, X. Li, L. Wang, Q. Zhang, The enhanced upconversion green by Yb-Mn dimer in NaBiF4 :Yb3+ /Er3+ /Mn2+ for optical fiber temperature sensor. J. Alloys Compd. 888, 161497 (2021) 44. I. Mikalauskaite, G. Pleckaityte, M. Skapas, A. Zarkov, A. Katelnikovas, A. Beganskiene, Emission spectra tuning of upconverting NaGdF4 :20% Yb, 2% Er nanoparticles by Cr3+ Codoping for optical temperature sensing. J. Lumin. 213, 210–217 (2019) 45. W. Xu, Y. Hu, L. Zheng, Z. Zhang, W. Cao, Improved green upconversion emissions from CaWO4 : Er3+ -Yb3+ by Cr3+ codoping for optical thermometry. J. Lumin. 215, 116617 (2019) 46. Z. Xing, P. Li, S. Wu, C. Liu, D. Dai, X. Li, L. Zhang, D. Wang, Z. Yang, Z. Wang, A perovskitelike LaSrGaO4 :Mn2+ , Nd3+ , Yb3+ NIR luminescent material for fluorescent temperature sensor. J. Lumin. 225, 117352 (2020) 47. W. Piotrowski, K. Kniec, L. Marciniak, Enhancement of the Ln3+ ratiometric nanothermometers by sensitization with transition metal ions. J. Alloys Compd. 870, 159386 (2021) 48. W. Piotrowski, L. Dalipi, R. Szukiewicz, B. Fond, M. Dramicanin, L. Marciniak, The role of Cr3+ and Cr4+ in emission brightness enhancement and sensitivity improvement of NIRemitting Nd3+ /Er3+ ratiometric luminescent thermometers. J. Mater. Chem. C 9(37), 12671– 12680 (2021) 49. J. Chen, L. Hu, J. Deng, X. Xing, Negative thermal expansion in functional materials: controllable thermal expansion by chemical modifications. Chem. Soc. Rev. 44(11), 3522–3567 (2015) 50. J. Liao, M. Wang, F. Lin, Z. Han, B. Fu, D. Tu, X. Chen, B. Qiu, H.-R. Wen, Thermally boosted upconversion and downshifting luminescence in Sc2 (MoO4 )3 :Yb/Er with two-dimensional negative thermal expansion. Nat. Commun. 13(1), 2090 (2022)

102

L. Marciniak et al.

51. J.S.O. Evans, in Negative Thermal Expansion Materials (Materials Research Forum LLC, Millersville, PA, 2018, 1999) 52. H. Lv, P. Du, W. Li, L. Luo, Tailoring of upconversion emission in Tm3+ /Yb3+ -codoped Y2 Mo3 O12 submicron particles via thermal stimulation engineering for non-invasive thermometry. ACS Sustain. Chem. Eng. 10(7), 2450–2460 (2022) 53. H. Zou, B. Chen, Y. Hu, Q. Zhang, X. Wang, F. Wang, Simultaneous enhancement and modulation of upconversion by thermal stimulation in Sc2 Mo3 O12 crystals. J. Phys. Chem. Lett. 11(8), 3020–3024 (2020) 54. P. Du, L. Luo, Y. Hou, W. Li, Energy transfer-triggered multicolor emissions in Tb3+ /Eu3+ coactivated Y2 Mo3 O12 negative thermal expansion microparticles for dual-channel tunable luminescent thermometers. Mater. Adv. 2(14), 4824–4831 (2021) 55. H. Zou, X. Yang, B. Chen, Y. Du, B. Ren, X. Sun, X. Qiao, Q. Zhang, F. Wang, Thermal enhancement of upconversion by negative lattice expansion in orthorhombic Yb2 W3 O12 . Angew. Chemie Int. Ed. 58(48), 17255–17259 (2019) 56. J. Peng, X.Z. Liu, F.L. Guo, S.B. Han, Y.T. Liu, D.F. Chen, Z.B. Hu, Crystal structure and negative thermal expansion of solid solution Lu2 W3-X Mox O12 . Int. J. Miner. Metall. Mater. 17(6), 786–790 (2010) 57. H. Cui, Y. Cao, Y. Zhang, T. Peng, L. Cao, S. Ran, Y. Wang, D. Wu, X. Li, X. Zhang, B. Chen, Thermal enhancement of up-conversion luminescence in Lu2 W2.5 Mo0.5 O12 : Er3+ , Yb3+ phosphors. Ceram. Int. 47(15), 21271–21275 (2021) 58. T. Peng, Y. Cao, H. Cui, Y. Zhang, Y. Wang, X. Li, X. Zhang, B. Chen, Enhancement of green upconversion luminescence and temperature sensitivity of Zr2 (WO4 )(PO4 )2 :Er3+ , Yb3 + phosphors by co-doping Li + ions. J. Alloys Compd. 893, 162345 (2022) 59. B. Ren, B. Chen, J. Zhao, Y. Guo, X. Zhang, X. Chen, Y. Du, Z. Deng, G. Zhu, F. Wang, Synthesis of core-shell ScF3 nanoparticles for thermal enhancement of upconversion. Chem. Mater. 33(1), 158–163 (2021) 60. Y. Yang, L. Lin, P. Lu, Z. Feng, Z. Li, J. Cai, Z. Mei, Y. Huang, W. Guo, Z. Wang, Z. Zheng, A linear calibrated high temperature sensor based on up-conversion fluorescence of Y2 Mo3 O12 :Er3+ , Yb3 + enhanced by negative thermal expansion. J. Lumin. 240, 118410 (2021) 61. M.Y. Wu, L. Wang, Y. Jia, Z.X. Guo, Q. Sun, Theoretical study of hydration in Y2 Mo3 O12 : effects on structure and negative thermal expansion. AIP Adv. 5(2), 27126 (2015) 62. Z. Wang, J. Christiansen, D. Wezendonk, X. Xie, M.A. Van Huis, A. Meijerink, Thermal enhancement and quenching of upconversion emission in nanocrystals. Nanoscale 11(25), 12188–12197 (2019) 63. M.D. Faucher, P. Sciau, J.-M.M. Kiat, M.-G.G. Alves, F. Bouree, P. Sciau, J.-M.M. Kiat, F. Bouree, Refinement of the monoclinic and tetragonal structures of Eu3+ -doped LiYO2 by neutron diffraction at 77 and 383 K differential scanning calorimetry, and crystal field analysis. J. Solid State Chem. 137(2), 242–248 (1998) 64. L. Marciniak, W. Piotrowski, M. Szalkowski, V. Kinzhybalo, M. Drozd, M. Dramicanin, A. Bednarkiewicz, Highly sensitive luminescence nanothermometry and thermal imaging facilitated by phase transition. Chem. Eng. J. 427, 131941 (2022) 65. L. Marciniak, W.M. Piotrowski, M. Drozd, V. Kinzhybalo, A. Bednarkiewicz, M. Dramicanin, Phase transition-driven highly sensitive, NIR–NIR band-shape luminescent thermometer based on LiYO2 : Nd3+ . Adv. Opt. Mater. 10(9), 2102856 (2022) 66. S. Wang, J. Zhang, Z. Ye, H. Yu, H. Zhang, Exploiting novel optical thermometry near room temperature with a combination of phase-change host and luminescent Pr3+ ion. Chem. Eng. J. 414, 128884 (2021) 67. R. Reisfeld, E. Zigansky, M. Gaft, Europium probe for estimation of site symmetry in glass films, glasses and crystals. Mol. Phys. 102(11–12), 1319–1330 (2004) 68. M. Back, J. Ueda, J. Xu, D. Murata, M.G. Brik, S. Tanabe, Ratiometric luminescent thermometers with a customized phase-transition-driven fingerprint in perovskite oxides. ACS Appl. Mater. Interfaces 11(42), 38937–38945 (2019)

New Strategies to Improve Thermal Sensitivity and Temperature …

103

69. T. Zheng, M. Runowski, I.R. Martín, S. Lis, M. Vega, J. Llanos, Nonlinear optical thermometry—a novel temperature sensing strategy via second harmonic generation (SHG) and upconversion luminescence in BaTiO3 :Ho3+ , Yb3 + perovskite. Adv. Opt. Mater. 9(12), 2100386 (2021)

An Overview of Luminescent Primary Thermometers Joana C. Martins, Carlos D. S. Brites, Albano N. Carneiro Neto, Rute A. S. Ferreira, and Luís D. Carlos

Abstract Luminescence thermometry is a spectroscopic technique for remote detection of temperature that becomed very popular since 2010. Up to now, the majority of the reported luminescent thermometers require calibration against a reference. Recurrent calibrations are, thus, mandatory, particularly when the thermometers are used in different media, and this is one limitation of the present technology. The determination of the temperature based on well-grounded physical principles by primary thermometers is the only way to overcome this limitation, contributing, therefore, to the implementation of the technique. This chapter is the first revision of the few examples of luminescent primary thermometers reported so far. Emphasis will be given to the comparison of their thermometric performance with those of non-luminescent examples, stressing the principles behind their operation and the importance of the concept for the desired consistent growth of this exciting research field.

1 Introduction Accurate and precise temperature measurements are crucial across a broad spectrum of areas, including scientific research, metrology, medicine, healthcare, and industry [1–3]. Presently, temperature sensors account for ≈80% of the worldwide sensor market, which was valued at USD 6412.5 million in 2020 and is projected to be worth USD 10,028.5 million by 2026, according to Mordor Intelligence Inc. [4]. J. C. Martins · C. D. S. Brites (B) · A. N. C. Neto · R. A. S. Ferreira · L. D. Carlos (B) Phantom-G, CICECO—Aveiro Institute of Materials, Department of Physics, University of Aveiro, 3810-193 Aveiro, Portugal e-mail: [email protected] L. D. Carlos e-mail: [email protected] A. N. C. Neto e-mail: [email protected] R. A. S. Ferreira e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_3

105

106

J. C. Martins et al.

From the statistical physics perspective, the thermodynamic temperature is an absolute measure of the average total internal energy of an object, essentially its kinetic energy [5, 6]. Quantitatively, the temperature (T ) can be defined from the second law of thermodynamics as the rate of change of entropy (S) with internal energy (U) at constant volume (V ), for a constant number of particles (N p ): 1 T = ( ∂S )

.

(1)

∂U V,N P

This quantitative definition of temperature is basilar for some applications, particularly those where temperatures are low or rapidly changing [1]. For most practical applications, however, temperature denotes “hotness”, or correctly, the amount of thermal energy, whereas temperature difference between two systems reflects the potential for the occurrence of heat transfer between them. Thermometers are devices that measure temperature or temperature changes. All thermometers have two key components: a sensor, i.e., a thermometric material whose physical change is used to sign temperature variations, and a reading procedure converting the physical change of the material into a numerical scale. For example, in a conventional liquid thermometer, the physical temperature-dependent quantity is the volumic mass of the liquid, and the procedure allowing the temperature determination is the comparison of the height of the liquid column (in a low-diameter tube) with a printed scale (previously calibrated) [7]. The thermometers evolved in parallel to the technological demands of industry and society [7, 8]. The initial thermometers were developed during the XVI and XVII centuries in Europe, essentially by Galileo Galilei, Santorio, Amontons, Fahrenheit, and Celsius (Fig. 1). Since the earliest days of medicine, physicians have recognized that the human body can exhibit an abnormal rise in temperature, usually defined as fever, as an obvious symptom of illnesses. Amontons developed an air-pressure thermometer (1702), that is very similar to our days’ air thermometers measuring a change in temperature in terms of a proportional change in pressure of a constant mass and volume of air [9]. This method eventually led to the concept of absolute zero temperature in the XIX century. The technology has then improved providing novel and disruptive approaches, e.g., thermal imaging, a technique still growing in medicine [10], and highly accurate devices. In the industry, the temperature is a critical parameter controlling most of the processes requiring, therefore, highly accurate measurements. Indeed, some industries use almost exclusively a specific kind of thermometer. For example, the electronic industry uses mainly thermocouples whereas in semiconductor and steel & coal industries infrared thermometers are more often employed. The so-called optical thermometers are gaining market share recently, particularly in industrial contexts of strong electromagnetic fields and moving objects [4]. Figure 2 provides a clear idea about the evolution of the thermal optical sensors field based on a literature survey of more than 10,100 publications within the last ten years [11]. Two main groups were identified as sensors based on radiation guidance, essentially involving

An Overview of Luminescent Primary Thermometers

107

Fig. 1 Schemes of the early thermoscopes of a Galileo and b, c Santoro. The only difference between a and b is the graduated scale. The first medical thermoscope is pictured in (c). d Illustration of an early air thermometer inserted in a graduated frame developed by Amontons. Adapted with permission from reference [9]

the thermal expansion of optical fibers [12], and the group associated with nonguided technology and based on the temperature-dependent properties of lightemitting materials (Fig. 2b, c). If this research is analyzed at a temporal level, we observe that despite fiber optic sensors dominating the emergence of this field, more recently has been observed an increase in the contribution of the non-guided group of sensors (Fig. 2d), in which luminescent materials displaying temperature-dependent properties (so-called thermographic phosphors) stand out. Luminescence thermometry (also called phosphor thermometry) is a spectroscopic technique for remote detection of temperature based on the luminescence temperature dependence of phosphors (e.g., peak energy and intensity, band shape, and excited states lifetimes and risetimes) [13, 14], with numerous applications ranging from biosciences [15–17] to engineering [18, 19]. This is an effervescent research field with dozens of papers coming out annually (namely since 2010). After an initial step dominated essentially by the demonstration of luminescence temperature dependence in numerous materials, the field is now moving to a more mature level requiring the comprehension of the mechanisms behind the thermal sensitivity, a complete description of the thermometer performance, and the development of specific applications [14]. As the technological evolution pushed the systems towards a downscaling process, thermometers evolved exactly in that direction, becoming smaller in size. Luminescence nanothermometry is a research field aiming at measuring the temperature at micro- and nanoscale where conventional methods are unpractical [20–22]. At these

108

J. C. Martins et al.

Fig. 2 a Map generated using data (title and abstract keywords) in 10,118 publications from Web of Science in the period 2000–2021. The red and green clusters, expanded in b and c, aggregate the indexing terms in b waveguided optical signal temperature sensors (Cluster I) and c nonguided optical signal temperature sensors (Cluster II). d Chronological evolution of the map in a. Reproduced with permission from [11]

submicrometric scales, the technique has been regarded as fundamental to developing underlying laws of heat transfer [23–26], designing and improving new circuits [27, 28], comprehending quantitatively heat exchange within cells and tissues [29–31], and harmlessly applying hyperthermia therapy [32–34]. According to the level of knowledge about the physical basis of the underlying thermodynamic rules and quantities that are used to determine the temperature, thermometers are categorized into primary and secondary. In primary thermometers, the temperature is predicted by a well-known equation of state, with all parameters or constants determined a priori. Temperature is, thus, calculated without any previous calibration, meaning that it can be applied as a reference (standard) thermal readout. Well-known examples include thermometers based on the speed of the sound in a gas,

An Overview of Luminescent Primary Thermometers

109

the thermal noise, voltage, or current in electrical resistors, the angular anisotropy of gamma radiation, and the black body radiation. Secondary thermometers, although the most widely utilized, require calibration against a reference (many times a primary thermometer) at a fixed temperature at least once, if not multiple times, namely if they are used in different media. This is a tedious and time-consuming task that is not always possible to be executed, as, for instance, in living cells and in operating electronic devices. As an alternative, secondary thermometers can, in some cases, be calibrated against fixed temperature values such as the triple point of water at 273.16 K and 611.73 Pa. The importance of primary thermometers was evidenced in 2018, a milestone in metrology science because all the seven base units (e.g., length, time, amount of substance, electric current, luminous intensity, mass and temperature) began to be defined by fixing their values to that of atomic or fundamental constants. The temperature (redefined using the Boltzmann constant k B ) was exclusively measured using primary thermometers and the results were applied to redefine the international temperature scales (ITS-90 and PLTS-2000) [35]. Despite the abovementioned developments in luminescence thermometry, only a few (ca. 16) primary luminescent thermometers have been developed up to now based on the emission and excitation features of thermographic phosphors, essentially Ln3+ -doped materials. Examples include the temperature dependence of the (i) peak energy of semiconductor nanoparticles (NPs) [36], (ii) intensity emissions of two Ln3+ excited states in thermal equilibrium [26, 37–48], and (iii) Tb3+ 5 D4 lifetime [49]. Concerning the excitation features, there is only one primary luminescent thermometer in which the thermometric parameter is defined based on the excitation spectra of a Eu3+ complex [50]. The present chapter, divided into five Sections, focuses, therefore, on these examples, being the first revision of luminescent primary thermometers. After a brief overview of selected examples of non-luminescent primary thermometers (Sect. 2)— presented here to demonstrate the sophistication of the experimental setups and their exceptionally low temperature uncertainties—we summarize in Sect. 3 the parameters employed to quantify the performance of luminescent thermometers (primary and secondary). Section 4, the core of the chapter, detailed discusses the abovementioned 16 examples of primary luminescent thermometers according to the law or mechanism governing their thermometric features. Finally, Sect. 5 summarizes the chapter and provides a perspective view.

2 Non-luminescent Primary Thermometers 2.1 Optical The so-called spectral-band radiometric thermometry is based on the Planck law, which gives the spectral radiance, L b,λ (the subscript λ indicates that the value is expressed per wavelength unit), of an ideal blackbody that depends on the

110

J. C. Martins et al.

temperature: L b,λ (λ, T ) =

1 2hc2 ) ( 5 hc/λ λ exp −1

(2)

kB T

where h is the Planck constant, c the speed of light in vacuo, and λ is wavelength in vacuo. Spectral radiance is the power emitted per unit area per unit solid angle per unit wavelength and is often expressed with the units W·m−2 sr−1 nm−1 (Fig. 3a). The main absolute primary radiometric thermometry requires an accurate determination of all parameters affecting the spectral radiance, such as the emitted power of the defined spectral band and registered on a known solid angle by an isothermal cavity of known emissivity (see an example in Fig. 3b). For determining the emitted optical power, a radiometer, comprising a detector with a well-known absolute spectral responsivity and spectral filters, is employed. The optical system typically includes two co-axial circular apertures (pinholes) separated by a known distance defining the solid angle, including eventually optical lenses or mirrors. The refractive index of the medium in which the measurement is made must also be known. This system delivers thermal uncertainties down to 0.1 K at 2800.0 K, corresponding to δ T/T of about 4 ppm. In contrast, in relative primary radiometric thermometry, the absolute spectral responsivity of the radiometer and the geometry of the system (solid angle) are not required, as the optical power is measured in relation to the values of one (or more) fixed-point blackbody, each providing a reference temperature. The interpolation and extrapolation are greatly simplified with the use of parametric approximation of the integral expression of the optical power [51], which eliminates the need to

Fig. 3 a Wavelength dependence of the spectral radiance of the ideal blackbody for different temperatures. b Schematic diagram of the Physikalisch-Technische Bundesanstalt gold fixed-point blackbody. Adapted with permission from reference [28]

An Overview of Luminescent Primary Thermometers

111

iteratively solve the integral equation describing the measured optical power. Relative primary radiometric thermometry gives uncertainties that are only slightly higher than absolute primary radiometric thermometry. Recently Scigliuzzo et al. present a radiation-field thermometer for propagating fields, based on the coherent scattering of an artificial atom (a “quantum emitter”) strongly coupled to the end of a waveguide [52]. Thermal photons in the waveguide reduce the coherence of the process, leading to incoherent scattering and a detectable drop decrease in reflectance. This decrease can be converted into thermal occupation by an algebraic expression involving parameters that can be independently measured, being the system a primary thermometer in the sense that it does not need to be calibrated against another thermometer. The system is operative between 0.05 and 0.25 K. An important example of photon-based thermometry, exploiting the nuclear anisotropy of γ -ray sources, is the so-called nuclear-orientation thermometry, which is operative at very low values of temperature ( 250 K, the accuracy increases significantly, with MSD and MUD in the 0.1–0.7% and 0.3–1.4% ranges, respectively (the most accurate thermometer is 22/00 with MSD 0.03% and MUD 0.3%) [50]. The accuracy was quantified by MSD and MUD relatively to the temperature measured by the silicon diode cryogenic sensor. A graphical insight into the accuracy through violin plots is presented in the inset of Fig. 16b where the probability density (width of the violins) was normalized to the number of data points to allow a comparison between distinct methods. The 22/00, 22/01, 22/02, 22/10, and 22/11 thermometers are the most precise with SD ranging from 0.5 to 1.9 K for temperatures within 275–340 K. Another prediction from Eq. (27) that can be tested by measuring the excitation spectra of Eu3+ -based materials at different temperatures is the possibility of defining self-referencing thermometers, because for α = α , , which implies ΔE i = E α, − E α = 0, the thermometric parameter, becomes constant (or temperatureindependent), i.e., Δi = Sα→β , /Sα→β = Ai . From the abovementioned eight 7 Fα → 5 Dβ (with α, β = 0, 1, 2) transitions observed in the excitation spectra of [Eu(hfa)3 bpyO2 ], seven self-referencing thermometers were developed, e.g., 01/00, 02/01, 02/00, 11/10, 11/12, 12/10, and 20/22 (Fig. 16c for five selected examples) with negligible MSD (6

Yes

[6]

NA

1–2

Yes

[77]

Er/Tm@Yb@Nd/Yb

18 × 21 800

Yb@Nd

20

800

650

40

2

Yes

[87]

1000

NA

2

Yes

[98]

Cu9 S5 Cu2-x Se

70 × 13 980

None

26

NA

No

[127]

16

None

22

NA

No

Er/Yb@Lu@C

[128]

50 × 40 730 980

540

38

4

Yes

[130]

NaErF4

11

1530

800

75

2

Yes

[136]

Er/Yb@Y@SiO2

6600

980

540

NA

NA

Yes

[137]

Nd@Y@Nd

25

800

860

72

NA

Yes

[138]

NdOV4

2.4

800

860

72

NA

No

[139]

Au + Nd:LaF3

1200

790

1060

NA

NA

Yes

[140]

Graphene in DMF

1503000

800

None

67

NA

No

[141]

Au nanorods

44 × 13 800

None

61

NA

No

[142]

Au/AuS nanoshells

42/50

800

None

59

NA

No

[142]

Dopamine-melanin

160

800

None

40

NA

No

[143]

Biodegradable Au nanovesicles

26

800

None

37

NA

No

[144]

800

Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles

185

Fig. 18 a Light-to-heat efficiency and b luminescence intensity of NaGdF4 nanoparticles with different Nd3+ doping concentrations. Light-to-heat efficiency and relative luminescence intensity of c NaNdF4 nanoparticles with particle sizes of 9, 12, 18, and 25 nm, and d NaNdF4 @NaGdF4 core–shell nanoparticles with shell thicknesses of 1.5, 4.5, and 8 nm (Adapted with permission from Ref. [6], Copyright 2020, American Chemistry Society)

186

L. Liu and J. Zhang

References 1. M. Rodrigues, R. Piñol, G. Antorrena et al., Implementing thermometry on silicon surfaces functionalized by lanthanide-doped self-assembled polymer monolayers. Adv. Funct. Mater. 26(2), 200–209 (2016) 2. A.S. Souza, L.A.O. Nunes, I.G.N. Silva et al., Highly-sensitive Eu3+ ratiometric thermometers based on excited state absorption with predictable calibration. Nanoscale 8(9), 5327–5333 (2016) 3. J.S. Zhong, D.Q. Chen, Y.Z. Peng et al., A review on nanostructured glass ceramics for promising application in optical thermometry. J. Alloys Compd. 763, 34–48 (2018) 4. S.H. Wen, J.J. Zhou, K.Z. Zheng et al., Advances in highly doped upconversion nanoparticles. Nat. Commun. 9, 2415 (2018) 5. S.J. Zhang, J. Li, Z.W. Zhou et al., Tm3+ heavily doped NIR-III bioprobe with 1 μm Stokes shift towards deep-tissue applications. Opt. Express 29(26), 42674–42683 (2021) 6. L. Xu, J. Li, K.L. Lu et al., Sub-10 nm NaNdF4 nanoparticles as near-infrared photothermal probes with self-temperature feedback. ACS Appl. Nano Mater. 3(3), 2517–2526 (2020) 7. B. Dong, X.J. Wang, C.R. Li et al., Er3+ -Y3+ -codoped Al2 O3 for high-temperature sensing. IEEE Photon. Technol. Lett. 20(1–4), 117–119 (2008) 8. B. Dong, D.P. Liu, X.J. Wang et al., Optical thermometry through infrared excited green upconversion emissions in Er3+ -Yb3+ codoped Al2 O3 . Appl. Phys. Lett. 90(18), 181117 (2007) 9. L. Liu, Y.X. Wang, X.R. Zhang et al., Optical thermometry through green and red upconversion emissions in Er3+ /Yb3+ /Li+ :ZrO2 nanocrystals. Opt. Commun. 284(7), 1876–1879 (2011) 10. A. Nexha, J.J. Carvajal, M.C. Pujol et al., Lanthanide doped luminescence nanothermometers in the biological windows: strategies and applications. Nanoscale 13(17), 7913–7987 (2021) 11. O. Kibrisli, E. Erol, N. Vahedigharehchopogh et al., Noninvasive optical temperature sensing behavior of Ho3+ and Ho3+ /Er3+ doped tellurite glasses through up and down-converted emissions. Sens. Actuator A Phys. 315, 112321 (2020) 12. P. Du, J. Tang, W.P. Li et al., Exploiting the diverse photoluminescence behaviors of NaLuF4 :xEu3+ nanoparticles and g-C3 N4 to realize versatile applications in white lightemitting diode and optical thermometer. Chem. Eng. J. 406, 127165 (2021) 13. P.P. Li, M.C. Jia, G.F. Liu et al., Investigation on the fluorescence intensity ratio sensing thermometry based on nonthermally coupled levels. ACS Appl. Bio Mater. 2(4), 1732–1739 (2019) 14. E.M. Rodríguez, G. López-Peña, E.G. Montes Lifante et al., Persistent luminescence nanothermometers. Appl. Phys. Lett. 111, 081901 (2017) 15. J. Liu, Y. Zhao, X. Li et al., Dual-emissive CsPbBr3 @Eu-BTC composite for self-calibrating temperature sensing application. Cryst. Growth Des. 20(1), 454–459 (2020) 16. B. Wu, L. Zhao, Y. Wang et al., Er3+ /Yb3+ co-doped nanocrystals modified with 6aminocaproic acid for temperature sensing in biomedicine. RSC Adv. 9, 42228–42235 (2019) 17. M. Suta, A. Meijerink, A theoretical framework for ratiometric single ion luminescent thermometers-thermodynamic and kinetic guidelines for optimized performance. Adv. Theor. Simul. 3(12), 2000176 (2020) 18. D. Jaque, F. Vetrone, Luminescence nanothermometry. Nanoscale 4(15), 4301–4326 (2012) 19. C. Jacinto, S.L. Oliveira, L.A.O. Nunes et al., Normalized-lifetime thermal-lens method for the determination of luminescence quantum efficiency and thermo-optical coefficients: Application to Nd3+ -doped glasses. Phys. Rev. B 73(12), 125107 (2006) 20. X.J. Xie, N.Y. Gao, R.R. Deng et al., Mechanistic investigation of photon upconversion in Nd3+ -sensitized core-shell nanoparticles. J. Am. Chem. Soc. 135(34), 12608–12611 (2013) 21. M. Haase, H. Schäfer, Upconverting nanoparticles. Angew. Chem. Int. Ed. 50(26), 5808–5829 (2011)

Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles

187

22. Y.T. Zhong, G. Tian, Z.J. Gu et al., Elimination of photon quenching by a transition layer to fabricate a quenching-shield sandwich structure for 800 nm excited upconversion luminescence of Nd3+ -sensitized nanoparticles. Adv. Mater. 26(18), 2831–2837 (2014) 23. C.S. Ma, X.X. Xu, F. Wang et al., Probing the interior crystal quality in the development of more efficient and smaller upconversion nanoparticles. J. Phys. Chem. Lett. 7(16), 3252–3258 (2016) 24. F. Wang, J.A. Wang, X.G. Liu, Direct evidence of a surface quenching effect on size-dependent luminescence of upconversion nanoparticles. Angew. Chem. Int. Ed. 49(41), 7456–7460 (2010) 25. S.W. Hao, G.Y. Chen, C.H. Yang et al., Nd3+ -Sensitized multicolor upconversion luminescence from a sandwiched core/shell/shell nanostructure. Nanoscale 9(30), 10633–10638 (2017) 26. X. Yang, Z.Y. Wu, Z.N. Yang et al., Flame-made Y2 O3 :Yb3+ /Er3+ upconversion nanoparticles: mass production synthesis, multicolor tuning and thermal sensing studies. J. Alloys Compd. 854, 157078 (2021) 27. S. Bjurshagen, J.E. Hellström, V. Pasiskevicius et al., Fluorescence dynamics and rate equation analysis in Er3+ and Yb3+ doped double tungstates. Appl. Opt. 45(19), 4715–4725 (2006) 28. T.S. Yang, Y. Sun, Q. Liu et al., Cubic sub-20 nm NaLuF4 -based upconversion nanophosphors for high-contrast bioimaging in different animal species. Biomaterials 33(14), 3733–3742 (2012) 29. Y.S. Zhu, W. Xu, H.Z. Zhang et al., Inhibited local thermal effect in upconversion luminescence of YVO4 :Yb3+ , Er3+ inverse opals. Opt. Express 20(28), 29673–29678 (2012) 30. D.M. Wu, A. García-Etxarri, A. Salleo et al., Plasmon-enhanced upconversion. J. Phys. Chem. Lett. 5(22), 4020–4031 (2014) 31. N.J.J. Johnson, S. He, S. Diao et al., Direct evidence for coupled surface and concentration quenching dynamics in lanthanide-doped nanocrystals. J. Am. Chem. Soc. 139(8), 3275–3282 (2017) 32. J. Zuo, Q.Q. Li, B. Xue et al., Employing shells to eliminate concentration quenching in photonic upconversion nanostructure. Nanoscale 9(23), 7941–7946 (2017) 33. L. Liu, K.L. Lu, L. Xu et al., Highly efficient upconversion luminescence of Er heavily doped nanocrystals through 1530 nm excitation. Opt. Lett. 44(3), 711–714 (2019) 34. X.M. Liu, X.G. Kong, Y.L. Zhang et al., Breakthrough in concentration quenching threshold of upconversion luminescence via spatial separation of the emitter doping area for bioapplications. Chem. Commun. 47(43), 11957–11959 (2011) 35. F. Vetrone, R. Naccache, V. Mahalingam et al., The active-core/active-shell approach: a strategy to enhance the upconversion luminescence in lanthanide-doped nanoparticles. Adv. Funct. Mater. 19(18), 2924–2929 (2009) 36. D.M. Yang, C.X. Li, G.G. Li et al., Colloidal synthesis and remarkable enhancement of the upconversion luminescence of BaGdF5 :Yb3+ /Er3+ nanoparticles by active-shell modification. J. Mater. Chem. 21(16), 5923–5927 (2011) 37. F. Wu, X.M. Liu, X.G. Kong et al., The real role of active-shell in enhancing the luminescence of lanthanides doped nanomaterials. Appl. Phys. Lett. 102(24), 243104 (2013) 38. X.M. Li, R. Wang, F. Zhang et al., Engineering homogeneous doping in single nanoparticle to enhance upconversion efficiency. Nano Lett. 14(6), 3634–3639 (2014) 39. X.X. Xu, C. Clarke, C.S. Ma et al., Depth-profiling of Yb3+ sensitizer ions in NaYF4 upconversion nanoparticles. Nanoscale 9(23), 7719–7726 (2017) 40. J. Wang, R.R. Deng, M.A. MacDonald et al., Enhancing multiphoton upconversion through energy clustering at sublattice level. Nat. Mater. 13(2), 157–162 (2014) 41. J.B. Zhao, D.Y. Jin, E.P. Schartner et al., Single-nanocrystal sensitivity achieved by enhanced upconversion luminescence. Nat. Nanotechnol. 8(10), 729–734 (2013) 42. F. Vetrone, J.C. Boyer, J.A. Capobianco et al., Effect of Yb3+ codoping on the upconversion emission in nanocrystalline Y2 O3 :Er3+ . J. Phys. Chem. B 107(5), 1107–1112 (2003) 43. J.A. Capobianco, F. Vetrone, J.C. Boyer et al., Enhancement of red emission (4 F9/2 →4 I15/2 ) via upconversion in bulk and nanocrystalline cubic Y2 O3 :Er3+ . J. Phys. Chem. B 106(6), 1181–1187 (2002)

188

L. Liu and J. Zhang

44. L. Liu, Y.J. Chen, X.L. Zhang et al., Improved optical thermometry in Er3+ :Y2 O3 nanocrystals by re-calcination. Opt. Commun. 309, 90–94 (2013) 45. L. Liu, H.L. Jiang, Y.J. Chen et al., Power dependence of upconversion luminescence of Er3+ doped Yttria nanocrystals and their bulk counterpart. J. Lumin. 143, 423–431 (2013) 46. L. Liu, C.L. Li, X.L. Zhang et al., Suppression of energy transfer from Er3+ to OH− in Er3+ highly doped zirconia. Opt. Commun. 287, 228–233 (2013) 47. L. Xu, J.M. Zhang, D.H. Gao et al., Upconversion enhancement through a facile, ultrafast, and low-threshold laser annealing strategy. Nanotechnol. 30(43), 435703 (2019) 48. L.M. Wang, X.Y. Li, Z.Q. Li et al., A new cubic phase for a NaYF4 host matrix offering high upconversion luminescence efficiency. Adv. Mater. 27(37), 5528–5533 (2015) 49. G.S. Ofelt, Intensities of crystal spectra of rare-earth ions. J. Chem. Phys. 37(3), 511–520 (1962) 50. B.R. Judd, Optical absorption intensities of rare-earth ions. Phys. Rev. 127(3), 750–761 (1962) 51. E. Greenberg, G. Katz, R. Reisfeld et al., Radiative transition probabilities of Er3+ in yttria stabilized cubic zirconia crystals. J. Chem. Phys. 77, 4797–4803 (1982) 52. M.D. Shinn, W.A. Sibley, M.G. Drexhage et al., Optical transitions of Er3+ ions in fluorozirconate glass. Phys. Rev. B 27, 6635–6648 (1983) 53. G. Yao, C.K. Lin, Q.G. Meng et al., Calculation of Judd-Ofelt parameters for Er3+ in βNaYF4 :Yb3+ , Er3+ from emission intensity ratios and diffuse reflectance spectra. J. Lumin. 160, 276–281 (2015) 54. W.T. Carnall, P.R. Fields, B.G. Wybourne, Spectral intensities of the trivalent lanthanides and actinides in solution. I. Pr3+ , Nd3+ , Er3+ , Tm3+ , and Yb3+ . J. Chem. Phys. 42, 3797–3806 (1965) 55. L. Liu, K.L. Lu, D. Yan et al., Concentration dependent optical transition probabilities in ultra-small upconversion nanocrystals. Opt. Express 26(18), 23471–23479 (2018) 56. Q. Liu, X.B. Li, B. Zhang et al., Structure evolution and delayed quenching of the double perovskite NaLaMgWO6 :Eu3+ phosphor for white LEDs. Ceram. Int. 42(14), 15294–15300 (2016) 57. N. Yang, J.H. Li, Z.W. Zhang et al., Delayed concentration quenching of luminescence caused by Eu3+ -Induced phase transition in LaSc3 (BO3 )4 . Chem. Mater. 32(16), 6958–6967 (2020) 58. Y. Liu, Z.W. Zhou, S.J. Zhang et al., Mechanisms of upconversion luminescence of Er3+ -doped NaYF4 via 980 and 1530 nm excitation. Nanomaterials 11(10), 2767 (2021) 59. W.T. Carnall, H. Crosswhite, H.M. Crosswhite, Energy level structure and transition probabilities in the spectra of the trivalent lanthanides in LaF3 , argonne. IL (United States) (1978). https://doi.org/10.2172/6417825 60. R. Wang, X.L. Zhang, Z.L. Zhang et al., Modified FIR thermometry for surface temperature sensing by using high power laser. Opt. Express 25(2), 848–856 (2017) 61. S.A. Wade, S.F. Collins, G.W. Baxter, Fluorescence intensity ratio technique for optical fiber point temperature sensing. J. Appl. Phys. 94(8), 4743–4756 (2003) 62. H.Q. Cui, Y.Z. Cao, Y.H. Zhang, et al., Thermal enhancement of up-conversion luminescence in Lu2 W2.5 Mo0.5 O12 :Er3+ ,Yb3+ phosphors. Ceram. Int. 47(15), 21271–21275 (2021) 63. J. Gao, X.T. Ren, K. Yang et al., Detection of hydrogen concentration based on an allfiber fluorescence intensity ratio optical thermometer with Er3+ /Yb3+ codoped NaBi(WO4 )2 phosphors. Optik 242, 167280 (2021) 64. R.S. Lei, D.G. Deng, X. Liu et al., Influence of excitation power and doping concentration on the upconversion emission and optical temperature sensing behavior of Er3+ :BaGd2 (MoO4 )4 phosphors. Opt. Mater. Express 8(10), 3023–3035 (2018) 65. J. Zhou, Y.Q. Chen, R.S. Lei et al., Role of Er3+ concentration on the photoluminescence and temperature sensing properties in Er3+ doped Y2 O3 based transparent ceramics. J. Lumin. 215, 116671 (2019) 66. H.Y. Yao, H.L. Shen, Q.T. Tang, Enhanced upconversion luminescence and temperature sensitivity of NaYF4 :Er, Yb phosphors via Mn doping. Mater. Res. Express 6(12), 125017 (2019)

Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles

189

67. D. He, C.F. Guo, S. Jiang et al., Optical temperature sensing properties of Yb3+ -Er3+ co-doped NaLnTiO4 (Ln= Gd, Y) up-conversion phosphors. RSC Adv. 5(2), 1385–1390 (2015) 68. N. Wang, Z.L. Fu, Y.L. Wei et al., Investigation for the upconversion luminescence and temperature sensing mechanism based on BiPO4 :Yb3+ , Re3+ (RE3+ = Ho3+ , Er3+ and Tm3+ ). J. Alloys Compd. 772, 371–380 (2019) 69. J. Zhang, S.S. An, Y.N. Zhang, et al., Optical temperature-sensing properties based on upconversion luminescence of La9.31 Si6.24 O26 :Er3+ ,Yb3+ with different strategies. Spect. Acta A 265, 120402 (2022) 70. M. Habibi, P. Bagheri, N. Ghazyani et al., 3D printed optofluidic biosensor: NaYF4 :Yb3+ , Er3+ upconversion nano-emitters for temperature sensing. Sens. Actuator A Phys. 326, 112734 (2021) 71. T. Li, C.F. Guo, S.S. Zhou et al., Highly sensitive optical thermometry of Yb3+ -Er3+ codoped AgLa(MoO4 )2 green upconversion phosphor. J. Am. Ceram. Soc. 98(9), 2812–2816 (2015) 72. C. Hazra, A. Skripka, S.J.L. Ribeiro et al., Erbium single-band nanothermometry in the third biological imaging window: potential and limitations. Adv. Opt. Mater. 8(23), 2001178 (2020) 73. G.T. Xiang, X.T. Liu, Q. Xia et al., Deep-tissue temperature sensing realized in BaY2 O4 :Yb3+ /Er3+ with ultrahigh sensitivity and extremely intense red upconversion luminescence. Inorg. Chem. 59(15), 11054–11060 (2020) 74. W.S. Silva, A.C.A. Silva, U. Rocha et al., Nd3+ doped TiO2 nanocrystals as self-referenced optical nanothermometer operating within the biological windows. Sens. Actuat. A Phys. 317, 112445 (2021) 75. C. Wang, P. Du, W.P. Li et al., Facile synthesis and photoluminescence performance of Er3+ activated BiOF sub-micro particles for ratiometric thermometers. J. Lumin. 226, 117416 (2020) 76. X. Yang, L.X. Wang, R. Wang et al., Achieving tunable multicolor display and sensitive temperature sensing in self-sensitization of erbium-doped CaF2 nanocrystals under 808, 980 and 1532 nm irradiation. Opt. Mater. Express 11(8), 2514–2527 (2021) 77. E.C. Ximendes, U. Rocha, C. Jacinto et al., Self-monitored photothermal nanoparticles based on core-shell engineering. Nanoscale 8(5), 3057–3066 (2016) 78. L. Marciniak, K. Prorok, A. Bednarkiewicz et al., Size dependent sensitivity of Yb3+ , Er3+ upconverting luminescent nano-thermometers. J. Mater. Chem. C 5(31), 7890–7897 (2017) 79. B. Dong, B.S. Cao, Y.Y. He et al., Temperature sensing and in vivo imaging by molybdenum sensitized visibie upconversion luminescence of rare-earth oxides. Adv. Mater. 24(15), 1987– 1993 (2012) 80. B.S. Cao, Y.Y. He, Z.Q. Feng et al., Optical temperature sensing behavior of enhanced green upconversion emissions from Er-Mo:Yb2 Ti2 O7 nanophosphor. Sens. Actuat. B Chem. 159(1), 8–11 (2011) 81. S. Balabhadra, M.L. Debasu, C.D.S. Brites, Boosting the sensitivity of Nd3+ -based luminescent nanothermometers. Nanoscale 7(41), 17261–17267 (2015) 82. L.T. Liu, L.H. Cheng, S. Xu et al., Study on optical temperature sensing properties of βNaYF4 :Tm3+ /Yb3+ nanoparticles. Mater. Res. Bull. 106, 353–356 (2018) 83. H. Suo, C.F. Guo, T. Li, Broad-scope thermometry based on dual-color modulation upconversion phosphor Ba5 Gd8 Zn4 O21 :Er3+ /Yb3+ . J. Phys. Chem. C 120(5), 2914–2924 (2016) 84. R.S. Lei, X. Liu, F.F. Huang et al., Optical thermometry based on anomalous temperaturedependent 1.53 μm infrared luminescence of Er3+ in BaMoO4 :Er3+ /Yb3+ phosphor. Opt. Mater. 86, 278–285 (2018) 85. N. Rakov, G.S. Maciel, Exploring the 4 I13/2 →4 I15/2 radiative transition from Er3+ in Y2 O3 for temperature sensing. J. Lumin. 199, 293–297 (2018) 86. M. Quintanilla, Y. Zhang, L.M. Liz-Marzan, Subtissue plasmonic heating monitored with CaF2 :Nd3+ , Y3+ nanothermometers in the second biological window. Chem. Mater. 30(8), 2819–2828 (2018) 87. K.L. Lu, X.H. Sun, L. Xu et al., Self-monitored biological nanoheaters operating in the first biological window based on single-band red upconversion nanoparticles fabricated through architectural design. J. Alloys Compd. 842, 155602 (2020)

190

L. Liu and J. Zhang

88. Y.F. Shang, Q. Han, S.W. Hao et al., Dual-mode upconversion nanoprobe enables broadrange thermometry from cryogenic to room temperature. ACS Appl. Mater. Inter. 11(45), 42455–42461 (2019) 89. U. Rocha, C. Jacinto, W.F. Silva et al., Subtissue thermal sensing based on Neodymium-doped LaF3 nanoparticles. ACS Nano 7(2), 1188–1199 (2013) 90. I.E. Kolesnikov, E.V. Golyeva, A.A. Kalinichev et al., Nd3+ single doped YVO4 nanoparticles for sub-tissue heating and thermal sensing in the second biological window. Sens. Actuat. B Chem. 243, 338–345 (2017) 91. Y. Zhou, L.X. Peng, P. Tan et al., Luminescence intensity ratio thermometry based on combined ground and excited states absorptions of Tb3+ doped CaWO4 . Opt. Express 29(14), 22805– 22812 (2021) 92. L.X. Peng, Y. Zhou, F. Qin et al., High performance FIR thermometry on the basis of the redshift of CTB by dual-wavelength alternative excitation in Eu3+ :YVO4 . Opt. Lett. 46(23), 5818–5821 (2021) 93. L.X. Peng, Q.Y. Meng, W.J. Sun et al., Synthesis and optical temperature sensing performance of NaLa(MoO4 )2 :Tb3+ , Eu3+ phosphors. Ceram. Int. 45(16), 20656–20663 (2019) 94. Y. Zhu, Q.Y. Meng, W.J. Sun et al., NaLa(MoO4 )2 :Sm3+ , Tb3+ phosphor: Optical temperature sensing material with a wide change range of luminescence color. J. Lumin. 218, 116854 (2020) 95. X.S. Lian, D. Zhao, Y.J. Cui et al., A near infrared luminescent metal-organic framework for temperature sensing in physiological range. Chem. Commun. 51(100), 17676–17679 (2015) 96. L. Marciniak, A. Bednarkiewicz, M. Stefanski et al., Near infrared absorbing near infrared emitting highly-sensitive luminescent nanothermometer based on Nd3+ to Yb3+ energy transfer. Phys. Chem. Chem. Phys. 17(37), 24315–24321 (2015) 97. A. Yakovliev, T.Y. Ohulchanskyy, R. Ziniuk et al., Noninvasive temperature measurement in dental materials using Nd3+ , Yb3+ doped nanoparticles emitting in the near infrared region. Part. Part. Syst. Char. 37(2), 1900445 (2020) 98. E.C. Ximendes, U. Rocha, K.U. Kumar et al., LaF3 core/shell nanoparticles for subcutaneous heating and thermal sensing in the second biological-window. Appl. Phys. Lett. 108(25), 253103 (2016) 99. E.C. Ximendes, U. Rocha, T.O. Sales et al., In vivo subcutaneous thermal video recording by supersensitive infrared nanothermometers. Adv. Funct. Mater. 27(38), 1702249 (2017) 100. L.X. Peng, Q.Y. Meng, W.J. Sun et al., Size dependent optical temperature sensing properties of Y2 O3 :Tb3+ , Eu3+ nanophosphors. RSC Adv. 9(5), 2581–2590 (2019) 101. D.P. Zhang, Z. Yue, Z.Y. Liu, Dual-mode luminescent core-shell nanoarchitectures for highly sensitive optical nanothermometry. J. Alloys Compd. 787, 585–593 (2019) 102. Y. Pan, X.J. Xie, Q.W. Huang et al., Inherently Eu2+ /Eu3+ codoped Sc2 O3 nanoparticles as high-performance nanothermometers. Adv. Mater. 30(14), 1705256 (2018) 103. C.Y. Wang, H. Lin, X.Q. Xiang et al., CsPbBr3 /EuPO4 dual-phase devitrified glass for highsensitive self-calibrating optical thermometry. J. Mater. Chem. C 6(37), 9964–9971 (2018) 104. M. Xu, X.M. Zhou, Q.Q. Su et al., Ratiometric nanothermometer in vivo based on triplet sensitized upconversion. Nat. Commun. 9, 2698 (2018) 105. F.P. Ruan, D.G. Deng, M. Wu et al., Multichannel luminescence properties and ultrahigh-sensitive optical temperature sensing of mixed-valent Eu2+ /Eu3+ co-activated Ca8 ZrMg(PO4 )6 (SiO4 ) phosphors. J. Alloys. Compd. 784, 1153–1161 (2019) 106. M.M. Liu, Q.S. Li, Q. Ma et al., Down-conversion luminescence and temperature sensing characteristics of LaSrGaO4 :Er3+ . Lumin. 37(2), 238–246 (2021) 107. J. Chen, W.N. Zhang, S.F. Cui et al., Up-conversion luminescence properties and temperature sensing performances of Ba5 Y8 Zn4 O21 :Yb3+ , Er3+ phosphors. J. Alloys Compd. 875, 159922 (2021) 108. Z.M. Cao, X.T. Wei, L. Zhao et al., Investigation of SrB4 O7 :Sm2+ as a multimode temperature sensor with high sensitivity. ACS Appl. Mater. Inter. 8(50), 34546–34551 (2016) 109. D.Q. Chen, Z.Y. Wan, Y. Zhou, Optical spectroscopy of Cr3+ -doped transparent nano-glass ceramics for lifetime-based temperature sensing. Opt. Lett. 40, 3607–3610 (2015)

Luminescence Thermometry in Heavily Doped Lanthanide Nanoparticles

191

110. M.D. Dramicanin, Sensing temperature via downshifting emissions of lanthanide-doped metal oxides and salta. a review. Methods Appl. Fluoresc. 4, 042001 (2016) 111. A. Siai, P. Haro-Gonzalez, K.H. Naifer et al., Optical temperature sensing of Er3+ /Yb3+ doped LaGdO3 based on fluorescence intensity ratio and lifetime thermometry. Opt. Mater. 76, 34–41 (2018) 112. S.V. Yap, R.M. Ranson, W.M. Cranton et al., Decay time characteristics of La2 O2 S: Eu and La2 O2 S: Tb for use within an optical sensor for human skin temperature measurement. Appl. Opt. 47(27), 4895–4899 (2008) 113. S.W. Allison, G.T. Gillies, A.J. Rondinone et al., Nanoscale thermometry via the fluorescence of YAG: Ce phosphor particles: measurements from 7 to 77 °C. Nanotechnol. 14(8), 859–863 (2003) 114. J. Zhou, R.S. Lei, H.P. Wang et al., A new generation of dual-mode optical thermometry based on ZrO2 :Eu3+ nanocrystals. Nanophot. 8(12), 2347–2358 (2019) 115. M.L. Tan, F. Li, N. Cao et al., Accurate in vivo nanothermometry through NIR-II lanthanide luminescence lifetime. Small 16(48), 2004118 (2020) 116. R. Wang, X.L. Zhang, F. Liu et al., Concentration effects on the FIR technique for temperature sensing. Opt. Mater. 43, 18–24 (2015) 117. C.D.S. Brites, P.P. Lima, N.J.O. Silva et al., Thermometry at the nanoscale. Nanoscale 4(16), 4799–4829 (2012) 118. M.A.R.C. Alencar, G.S. Maciel, C.B. de Araújo et al., Er3+ -doped BaTiO3 nanocrystals for thermometry: Influence of nanoenvironment on the sensitivity of a fluorescence based temperature sensor. Appl. Phys. Lett. 84(23), 4753–4755 (2004) 119. Z. Chouahda, J.P. Jouart, T. Duvaut et al., The use of the green emission in Er3+ -doped CaF2 crystals for thermometry application. J. Phys. Condens. Matt. 21(24), 245504 (2009) 120. L. Liu, Z.S. Sun, C. Ma et al., Highly sensitive and accurate optical thermometer through Er doped tellurite glasses. Mater. Res. Bull. 105, 306–311 (2018) 121. V.K. Tikhomirov, K. Driesen, V.D. Rodriguez et al., Optical nanoheater based on the Yb3+ Er3+ co-doped nanoparticles. Opt. Express 17(14), 11794–11798 (2009) 122. W.L. Lu, L.H. Cheng, H.Y. Zhong et al., Dependence of upconversion emission intensity on Yb3+ concentration in Er3+ /Yb3+ co-doped flake shaped Y2 (MoO4 )3 phosphors. J. Phys. D Appl. Phys. 43(8), 085404 (2010) 123. X.J. Zhu, Q.Q. Su, W. Feng et al., Anti-stokes shift luminescent materials for bio-applications. Chem. Soc. Rev. 46(4), 1025–1039 (2017) 124. S.J. Oldenburg, R.D. Averitt, S.L. Westcott et al., Nanoengineering of optical resonances. Chem. Phys. Lett. 288(2–4), 243–247 (1998) 125. X.H. Huang, S. Neretina, M.A. El-Sayed, Gold nanorods: from synthesis and properties to biological and biomedical applications. Adv. Mater. 21(48), 4880–4910 (2009) 126. J.Y. Chen, D.L. Wang, J.F. Xi et al., Immuno gold nanocages with tailored optical properties for targeted photothermal destruction of cancer cells. Nano Lett. 7(5), 1318–1322 (2007) 127. Q.W. Tian, F.R. Jiang, R.J. Zou, et al., Hydrophilic Cu9 S5 nanocrystals: a photothermal agent with a 25.7% heat conversion efficiency for photothermal abliation of cancer cells in vivo. ACS Nano 5(12), 9761–9771 (2011) 128. C.M. Hessel, V.P. Pattani, M. Rasch et al., Copper selenide nanocrystals for photothermal therapy. Nano Lett. 11(6), 2560–2566 (2011) 129. H.K. Moon, S.H. Lee, H.C. Choi, In vivo near-infrared mediated tumor destruction by photothermal effect of carbon nanotubes. ACS Nano 3(11), 3707–3713 (2009) 130. X.J. Zhu, W. Feng, J. Chang et al., Temperature-feedback upconversion nanocomposite for accurate photothermal therapy at facile temperature. Nat. Commun. 7, 10437 (2016) 131. S.A. Stephen, W.C. Dewey, Thermal dose determination in cancer therapy. Int. J. Radiat. Oncol. 10(6), 787–800 (1984) 132. D. Jaque, L.M. Maestro, B. del Rosal et al., Nanoparticles for photothermal therapies. Nanoscale 6(16), 9494–9530 (2014) 133. F. Zhang, Photon Upconversion Nanomaterials (Springer, New York, Vol, 2015), p.416

192

L. Liu and J. Zhang

134. A. Gnach, T. Lipinski, A. Bednarkiewicz et al., Upconverting nanoparticles: assessing the toxicity. Chem. Soc. Rev. 44(6), 1561–1584 (2015) 135. O. Savchuk, J.J. Carvajal, L.G. De la Cruz et al., Luminescence thermometry and imaging in the second biological window at high penetration depth with Nd:KGd(WO4 )2 . J. Mater. Chem. C 4(31), 7397–7405 (2016) 136. J. Li, L. Xu, K.L. Lu et al., Efficient nanoheater operated in a biological window for photohyperthermia therapy. Biomed. Opt. Express 10(4), 1935–1941 (2019) 137. S. Drobczynski, K. Prorok, K. Tamarov et al., Toward controlled photothermal treatment of single cell: optically induced heating and remote temperature monitoring in vitro through double wavelength optical tweezers. ACS Photon. 4(8), 1993–2002 (2017) 138. L. Marciniak, A. Pilch, S. Arabasz et al., Heterogeneously Nd3+ doped single nanoparticles for NIR-induced heat conversion, luminescence, and thermometry. Nanoscale 9(24), 8288–8297 (2017) 139. B. del Rosal, A. Perez-Delgado, E. Carrasco et al., Neodymium-based stoichiometric ultrasmall nanoparticles for multifunctional deep-tissue photothermal therapy. Adv. Opt. Mater. 4(5), 782–789 (2016) 140. U. Rocha, J. Hu, E.M. Rodríguez et al., Subtissue imaging and thermal monitoring of gold nanorods through joined encapsulation with Nd-doped infrared-emitting nanoparticles. Small 12(39), 5394–5400 (2016) 141. O.A. Savchuk, J.J. Carvajal, J. Massons et al., Determination of photothermal conversion efficiency of graphene and graphene oxide through an integrating sphere method. Carbon 103, 134–141 (2016) 142. J.R. Cole, N.A. Mirin, M.W. Knight et al., Photothermal efficiencies of nanoshells and nanorods for clinical therapeutic applications. J. Phys. Chem. C 113(28), 12090–12094 (2009) 143. Y.L. Liu, K.L. Ai, J.H. Liu et al., Dopamine-melanin colloidal nanospheres: An efficient near-infrared photothermal therapeutic agent for in vivo cancer therapy. Adv. Mater. 25(9), 1353–1359 (2013) 144. P. Huang, J. Lin, W.W. Li et al., Biodegradable gold nanovesicles with an ultrastrong plasmonic coupling effect for photoacoustic imaging and photothermal therapy. Angew. Chem. Int. Ed. 52(52), 13958–13964 (2013)

Metal–Organic Frameworks for Luminescence Thermometry Thibault Amiaud and Hélène Serier-Brault

Abstract Metal-Organic Frameworks (MOFs) are crystalline porous materials built up from metallic nodes and organic linkers that have attracted great interest due to their wide range of possible applications. Among them, luminescence thermometry has appeared in the last decade offering new perspectives to a subject for which inorganic materials were very present. Different strategies are possible to generate luminescence in MOFs materials: the inorganic cluster, the organic ligand or the incorporation of a host-guest molecule in the porosity. In this chapter are reviewed all approaches to design a ratiometric luminescent thermometer based on MOFs. Finally, a peculiar attention has been paid on mixed Eu-Tb MOFs and the identification of keys parameters enabling the modulation of their thermometric performances. Keywords Metal-Organic Frameworks · Lanthanides · Ratiometric thermometry

1 Introduction 1.1 Metal–Organic Frameworks: History, Synthesis and Applications In 1965, Tomic reported coordination polymers constructed using carboxylic acid linkers coordinated to various metals and the group investigated the thermal stabilities of these systems [80]. Nearly 25 years later, Hoskins and Robson proposed the idea that a wide range of scaffold-like materials with infinite 3D frameworks should be accessible, tunable and potentially useful [39]. They also predicted that materials with large empty cavities should be accessible while maintaining high thermal and T. Amiaud · H. Serier-Brault (B) Nantes Université, CNRS, Institut des Matériaux de Nantes Jean Rouxel, IMN, F-44000 Nantes, France e-mail: [email protected] T. Amiaud e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_5

193

194

T. Amiaud and H. Serier-Brault

chemical stability. The term metal–organic framework (MOF) was finally imposed by Yaghi et al. who reported the use of hydrothermal synthesis to obtain a 3D crystalline and open material [95]. After that, the research groups of Yaghi [28, 49, 73], Kigatawa [25, 45] and Ferey [11, 23, 24, 37] developed many MOFs materials with well-known designations (MIL-53, UiO-66, HKUST, MOF-5…) (Fig. 1). Since then, MOFs are constructed from a variety of inorganic nodes (metal clusters or ions) and organic linkers, the number of metal–organic combinations, and therefore structural possibilities, are nearly endless. They have been synthesized with various elements from the periodic table, namely s-, p-, d-, and f-block elements. The modularity and tunability of MOFs make them very attractive for a wide variety of potential applications including gas storage and release [50, 59], chemical separations [4, 51], catalysis [48], drug delivery [38], light harvesting and energy conversion[77, 85], sensing [41, 47], conductivity [79], ion-exchange[106], removal of toxic substance from air and water[19, 40], degradation of chemical warfare agents [64]… MOFs are traditionally synthesized by hydro/solvothermal method, which basically consists on a mixture of the organic linker and the metal precursor into a solvent [78]. The most common used solvents are water, alcohols, acetone, acetonitrile, pyridine, dimethyl or diethylformamide [81]. The mixture is heated in sealed vessels such as Teflon-lined vessels in stainless steel reactors or glass tubes, generating an autogenously pressure. Crystallization processes are influenced by a large

Fig. 1 Crystal structure of well-known MOFs, a MIL-53, b UiO-66, and c MIL-103

Metal–Organic Frameworks for Luminescence Thermometry

195

number of synthesis parameters including Metal/Ligand ratio, nature of the solvent, pH of the reaction, temperature, synthesis duration among others. Heating can be also performed by microwave to enable to shorten crystallization times compared to the conventional hydro/solvothermal route. The use of microwave irradiation for the reaction synthesis instead of the conventional heating also allows to gain high reproducibility, selectivity of phase, increase of the product yields or improvement on product purities. Numerous MOFs can be obtained through this synthetic route, especially for nanoparticles synthesis [5, 25, 32, 44, 62]. Luminescence thermometry was developed to respond to some limitations of contact thermometers such as the monitoring of the temperature at the nanoscale regime especially for biological applications [27]. As a consequence, luminescent thermometer materials must be synthesized as nanoparticles. In that case, some alternatives synthetic techniques were developed to obtain nanoMOFs like reverses micelles [10], spray-drying methods [88]… If almost all studies on inorganic materials for luminescence thermometry concern nanoparticles [6, 58, 76], MOFs designed for luminescent thermometers are mainly microsized powder even when materials are sensitive in the physiological range. Consequently, in the future, efforts should be focused on the synthesis of nanoMOFs in order to render MOFs materials more suitable for practical applications and more competitive with inorganic materials well-established in this topic.

1.2 Ratiometric Thermometry in Metal–Organic Frameworks Historically, luminescence thermometry was based on three main approaches to determine the temperature: (a) the spectral shift of a given transition, (b) emission intensity measurements, and (c) lifetime measurements [9]. In MOF materials, the second approach was retained and the temperature is evaluated from the measurement of intensities from two energy levels (or two Stark components of an excited state) in thermal equilibrium. Such thermometers are self-calibrated and they are called ratiometric thermometers. In the last decades, MOF materials have been investigated as various sensors due to their multiple luminescent centers and tunable luminescence properties [47, 72]. Indeed, compared with organic dyes or inorganic luminescent materials, the luminescence of MOFs is more diverse because the metal nodes, organic linkers, ligand–metal charge transfer, and guest species within framework can potentially generate emissions (Fig. 2). Consequently, many combinations are possible to elaborate ratiometric MOF thermometers and three different behaviors can be envisioned to design dual-emitting MOFs [16]. First, the luminescence can be generated from two independent emitting centers which are electronically independent. Therefore, any energy transfer occuring between both emitters and the thermometric properties arise from the separate thermal quenching of each luminescent emitter. With this strategy, highly sensitive luminescent thermometers could be designed if emitters have very distinct thermal quenching [12]. Dual emission can be also generated from two closely related luminescent centers with one center acting

196

T. Amiaud and H. Serier-Brault

as sensitizer (or donor) via an energy transfer. In that case, emitters must be spatially close to favor the energy transfer process, which regularly occurs between the organic ligand and lanthanides (also called antenna effect) or between two lanthanides ions. The energy transfer will compete with the donor emission and can be also thermally dependent, which conducts to very different temperature-dependence of both emissions. The third behavior, which is very little exploited in MOF materials so far, is based on the emission of two thermally coupled energy levels of a single luminescent center. The electrons populations of both involved levels are governed by the Boltzmann distribution law, and the temperature changes will induce population re-distribution between energy levels, conducting by consequence to a modification in the luminescence intensity ratio. This last approach is only suitable for luminescent centers possessing thermally coupled levels, i.e. levels with an energy gap < to 2000 cm−1 . For lanthanides ions, all thermally coupled energy levels with gap energy are reported by Zaldo [100]. Consequently, metal–organic frameworks (MOFs), built up from all kinds of inorganic metal cations and/or cluster nodes with various organic ligand linkers, turn out to be rational candidates for luminescent thermometers owing to the abundance of metal cations and the infinity of organic connectors with diverse geometries and luminophore functional groups. The interest for MOF materials for luminescence thermometry dates from 2012 and the first paper related the possibility to use lanthanide-bearing MOF for ratiometric thermometry [15]. Since then, the number of interesting materials growths up but MOF material do not totally compete with

Fig. 2 Schematic representation of possible luminescent centers in MOF materials and the three different strategies to design ratiometric luminescent MOF thermometers

Metal–Organic Frameworks for Luminescence Thermometry

197

historical inorganic materials (Ag2 S, fluorides, oxides…) which are already used in more practical uses (in microfluidics, in vitro bioimaging…) [20, 94].

2 Various Strategies to Develop MOFs Luminescent Thermometers 2.1 Eu-Tb Mixed Metal–Organic Frameworks In 2012, Cui et al. [15] have proved that lanthanide-bearing MOFs, containing two distinct Ln3+ emitters, namely Eu3+ and Tb3+ ions, could present tremendous potential for applications as ratiometric thermometry. In such materials, the organic ligand has a role of structuring agent but also of sensitizer to favor the emission of lanthanide, this phenomenon being called antenna effect. Thus, the organic linker that has much high absorption cross-section than lanthanides absorbs in the UV range, and via an energy transfer from its triplet excited state energy (T1 ) to the emitting energy states of Eu3+ and Tb3+ ions to generate the characteristic lines of both emitters (Fig. 3). The efficient sensitization of Eu3+ and Tb3+ ions is ensured when the triplet excited state energy is comprised between 22,000 and 27,000 cm−1 [71]. Currently, Ln-MOFs reported to date for luminescence thermometry are based on the Eu3+ -to-Tb3+ emission ratio between the transitions 5 D4 → 7 F5 and 5 D0 → 7 F2 of the Tb3+ and Eu3+ ions, respectively (Fig. 3), the thermometric parameter Δ being equal to ITb /IEu . Materials can be sensitive in various temperature regions: cryogenic range (< 100 K), medium range (100–300 K), biological range (298– 323 K), and high-temperature ranges (> 400 K) [71]. In their chemical composition, MOFs contain organic ligands but also coordinated solvent molecules which limit

Fig. 3 a Representation of various energy transfer processes occurring between Eu3+ and Tb3+ ions with a hypothetical organic ligand in MOF materials, b representation of typically used integrated emissions in a mixed Eu-Tb MOF for thermometry

198

T. Amiaud and H. Serier-Brault

their thermal stability rendering the MOFs not suitable for very high-temperature applications (T > 500 °C). After the pioneered work of Cui et al. [15], the interest in mixed Eu-Tb MOFs for luminescence thermometry grown up and a large number of MOFs materials is already published with various organic ligands, very diverse crystal structure and chemical compositions, and distinct operating temperatures. In 2016, Rocha and Carlos wrote an interesting review where they proposed an exhaustive list of reported mixed Eu-Tb MOFs [71] to illustrate the large versatility of these materials. However, efforts must be done to rationalize the research on highly sensitive MOFs luminescent thermometers in order to be more predictive in the operating temperature range, the thermal sensitivity and other thermometric performances described in many book chapters or publications [9, 71]. Many parameters can be tuned such as the ligand nature, the synthetic parameters (solvent, temperature, duration, pH…) to synthesize MOFs materials which lead to a huge number of crystallographic structures. As studying each parameter individually would be very time consuming, we have attempted to identify structural parameters governing thermometric performances of mixed Eu-Tb MOFs through a thorough investigation of the highly sensitive materials already reported in the literature. Among numerous structural parameters that one can enumerate, the topology of the inorganic network seems to be the most relevant parameter because it governs most of energy transfers responsible for thermal-dependence of emission.

2.1.1

Effect of the Topology of the Inorganic Network

In the Table 1 are reported the most sensitive mixed Eu-Tb MOFs in the cryogenic range (T < 100 K), ordered according their maximal relative thermal sensitivity Sm (in % K−1 ) at a specific temperature, named Tm . The crystal structure of {[Tb1.87 Eu0.13 (HY)2 (H2 O)3 ]·5H2 O} [1], {[Tb0.95 Eu0.05 (btb)(H2 O)]·(H2 O)x ·(CH3 OH)x } [56], {[Tb0.914 Eu0.086 (pda)3 (H2 O)]·2H2 O} [87], and {[Tb0.9 Eu0.1 (bdc)3 (H2 O)2 ]·H2 O} [65] were thoroughly investigated (Fig. 4). In general manner, a trend is observed in the topology of sensitive materials in the cryogenic domain since these materials are mainly formed by chains of [LnOx ] polyhedra. The material {[Tb1.87 Eu0.13 (HY)2 (H2 O)3 ]·5H2 O} [56], sensitive at very low temperature (4 K), exhibits zigzag-shaped chains where [LnOx ] polyhedra are very close to each other with Ln-Ln distances of 4.21 Å for two neighboring Ln3+ ions and a relatively small inter-chain distances of 8.25 Å. In the materials {[Tb0.95 Eu0.05 (btb)(H2 O)]·(H2 O)x ·(CH3 OH)x } [1] and {[Tb0.914 Eu0.086 (pda)3 (H2 O)]·2H2 O} [87], where the chain topology is still present, the [LnOx ] polyhedra are less close with Ln-Ln distances around 4 Å and inter-chain distances ranging from 8.84 to 15.94 Å. Finally, even if the material {[Tb0.9 Eu0.1 (bdc)3 (H2 O)2 ]·H2 O} [65] exhibits [LnOx ] polyhedra arranged in dimers, the small dimer-dimer distance of 4.91 Å makes the structure similar to a chain topology, that can explain its thermal sensitivity in the cryogenic range. Consequently, a high connectivity of [LnOx ] polyhedra seems to be favorable to the design of cryogenic

Metal–Organic Frameworks for Luminescence Thermometry

199

Table 1 List of mixed Eu-Tb MOFs highly sensitive in the cryogenic range (T < 100 K) Materials

S m (% K−1 ) T m (K) Topology References

{[Tb1.87 Eu0.13 (HY)2 (H2 O)3 ]·5H2 O}

31

4

Chain

[56]

{[Tb0.914 Eu0.086 (pda)3 (H2 O)]·2H2 O}

5.96

25

Chain

[87]

{[Tb0.898 Eu0.102 (notp)(NO3 )(H2 O)]·8H2 O}

3.90

38

Chain

[70]

{[Tb0.9 Eu0.1 (bdc)3 (H2 O)2 ]·H2 O}

3.30

36

Dimer

[65]

{[Tb0.95 Eu0.05 (btb)(H2 O)]·(H2 O)x ·(CH3 OH)x }

2.85

14

Chain

[1]

luminescent thermometer. However, no obvious trend in the relative thermal sensitivities appears for the different materials. Additional materials must be designed to confirm the tendency and to go further in the identification of structural parameters governing thermometric properties, especially the relative thermal sensitivity. In the Table 2 are reported the most sensitive mixed Eu-Tb MOFs in the medium range (100 < T < 300 K), ordered according their maximal relative thermal sensitivity Sm (in % K−1 ). Then, some crystal struc[15], tures, namely {[(Eu0.0069 Tb0.9931 )2 (DMBDC)3 (H2 O)4 ·DMF·H2 O} [89], [Eu0.0878 Tb0.9122 (tpi)(DMF)2 (NO3 ] {[Tb0.99 Eu0,01 (bdc)0.5 (dstp)]·2H2 O} [90] and {[Tb0.7 Eu0.3 (cbpi)2 (COO)(H2 O)2 ]·H2 O} [105], are represented in Fig. 5 and were thoroughly investigated in order to identify a common point that could explain their thermal sensitivity in this medium range. Contrary to the cryogenic mixed Eu-Tb MOFs thermometers where [LnOx ] polyhedra are highly condensed (mainly chain topologies), medium luminescent thermometers are mostly composed by isolated [LnOx ] polyhedra or dimers with various distances between them. The distance between isolated [LnOx ] polyhedra or dimers seems to be a crucial parameter that impacts the maximal temperature Tm . Thus, the increase of distance between [LnOx ] polyhedra tends to increase the temperature where the material is the most sensitive. Actually, the material {[Tb0.7 Eu0.3 (cbpi)2 (COO)(H2 O)2 ]·H2 O} [105], where [LnOx ] polyhedra are spatially separated by a distance of around 18 Å, is the most sensitive at 300 K. By contrary, the compounds {[(Eu0.0069 Tb0.9931 )2 (DMBDC)3 (H2 O)4 ·DMF·H2 O} [15] and {[Tb0.99 Eu0,01 (bdc)0.5 (dstp)]·2H2 O} [89] are sensitive at 200 K with a distance between [LnOx ] monomers equal to around 5 Å. Finally, the intermediate case is observable for the compound [Eu0.0878 Tb0.9122 (tpi)(DMF)2 (NO3 )] [90], that is sensitive at 250 K, and which exhibits a distance between dimer of 8.80 Å. Consequently, mixed Eu-Tb MOFs luminescent thermometers exhibiting a topology of the inorganic network based on [LnOx ] dimers or monomers present a thermal sensitivity between 200 and 300 K. The distance between dimers or monomers governs the operating temperature range, a longer distance for a higher temperature. However, the topology of the inorganic network as well as the distance between the monomers or dimers are not the only parameters governing the temperature domain. Indeed, the investigation of sensitive materials in the physiological domain or in the high temperature domain shows more disparate distances between monomers (Table 3; Figs. 6 and 7). Consequently, finely predicting the operating temperature between

200

T. Amiaud and H. Serier-Brault

Fig. 4 Schematic illustration of topology effect of the inorganic network on the detection range of LnMOFs thermometers sensitive in the cryogenic Table 2 List of mixed Eu-Tb MOFs sensitive in the medium range (100 < T < 300 K) Materials

S m (% K−1 )

T m (K)

Topology

References

{[Tb0.957 Eu0.043 (H2 cpda)(Hcpda)(H2 O)]·6(H2 O)}

16

300

Monomer

[17]

[Eu0.0878 Tb0.9122 (tpi)(DMF)2 (NO3 )]

4.92

250

Dimer

[90]

{[Tb0.99 Eu0.01 (bdc)0,5 (dstp)]·2H2 O}

3.9

200

Monomer

[89]

{[(Eu0.0069 Tb0.9931 )2 (DMBDC)3 (H2 O)4 ·DMF·H2 O}

1.15

200

Monomer

[15]

{[(Tb0.9382 Eu0.0616 )(bpcd)2 (NO3 )2 ]·Cl·2H2 O}

0.34

200

Monomer

[61]

{[Tb0.7 Eu0.3 (cbpi)2 (COO)(H2 O)2 ]·H2 O}

0.17

300

Monomer

[105]

Metal–Organic Frameworks for Luminescence Thermometry

201

Fig. 5 Schematic illustration of topology effect of the inorganic network on the detection range of LnMOFs thermometers sensitive in the medium

200 and 400 K, additional structural parameters must be identified. In order to have an efficient sensitization of Eu3+ and Tb3+ ions, the organic ligand must present a triplet level with an energy between 22,000 and 27,500 cm−1 [71]. However, it is interesting to study the thermal behavior of mixed Eu-Tb MOF materials getting an organic ligand with a triplet level T1 energy around the lowest value of this domain or to the highest one.

2.1.2

Effect of the Organic Ligand

Undoubtedly, the topology of the MOF inorganic lattice governs the sensitivity range of the luminescent thermometer. It is therefore possible to anticipate and predict whether a MOF will be sensitive in the cryogenic range or at higher temperatures.

202

T. Amiaud and H. Serier-Brault

Table 3 List of mixed Eu-Tb MOFs highly sensitive in the physiological range and at higher temperature Materials

S m (% T m (K) Topology K−1 )

{[Tb0.957 Eu0.043 (H2 cpda)(Hcpda)(H2 O)]·6(H2 O)} 16 [Eu0.036 Tb0.964 (BPTC)(CH3 )2 (NH2 )]

9.42

References

300

Monomer

[17]

310

Monomer

[66]

[Eu0.01 Tb0.99 NDC]

7.32

320

Ln6 cluster [92]

[Tb0.9 Eu0.1 (PIA)(HPIA)(H2 O)2.5 ]

3.53

300

Monomer

[69]

[Tb0.999 Eu0.001 (bpdc)(ad)]

1.23

300

Monomer

[75]

[Tb0.66 Eu0.33 (AD)0,5 (phth)(H2 O)2 ]

1.21

303

Dimer

[13]

{[Tb0.8 Eu0.2 (BPDA)(NO3 )(DMF)2 ]·(DMF)}

1.19

313

Monomer

[102] [10]

[Tb0.99 Eu0.01 (BDC)1.5 (H2 O)2 ]

0.37

318

Monomer

Eu@UiO-(bpydc)

0.31

293

Ln6 cluster [107]

{[Tb0.7 Eu0.3 (cbpi)2 (COO)(H2 O)2 ]·H2 O}

0.17

300

Monomer

[105] [96]

[Eu0.0066 Tb0.9934 (cbpp)(CH3 )2 (NH2 )(H2 O)2 ]

3.76

450

Monomer

[Eu0.01 Tb0.99 BDC-NH2 ]

2.49

426

Ln6 cluster [92]

[Eu0.01 Tb0.99 BDC-OH]

2.24

425

Ln6 cluster [92]

[Tb0.9 Eu0.1 (cbpi)(NO3 )]

1.75

423

Dimer

[98]

[Tb0.99 Eu0.01 (hfa)3 (dpbp)]

0.83

450

Monomer

[63]

[Tb0.92 Eu0.08 (HPIDC)(ox)0.5 H2 O]·3H2 O

0.60

473

Monomer

[97]

[Eu0.7 Tb0.3 (D -cam)(Himdc)2 (H2 O)2 ]3

0.11

450

Monomer

[33]

At the moment, the database of Eu-Tb MOFs is not large enough to propose a more precise distinction between a sensitive material in the medium range and a sensitive material in the physiological range. The identification of the parameters governing precisely the thermometric properties is not easy because many parameters vary according the investigated material. Thus, if the topology seems to play an important role, the impact of the ligand should be not neglected, and in particular the energy positioning of its triplet T1 level. For molecular thermometers built from lanthanide complexes, there are studies highlighting the role of the ligand on the relative thermal sensitivity, and even the shape of the calibration curve Δ = f(T) [8, 67]. The effect of the ligand triplet was first demonstrated by the group of Dian Zhao [92] who worked on MOF UiO-66 with three different ligands, namely 2hydroxyterephthalic acid, 2-aminoterephthalic acid and 1, 4-naphthalendicarboxylic acid, the latter not allowing to transfer the energy on the 5 D4 level of the Tb3+ ion, leading then to another type of luminescent thermometer, based on the emission of ligand and Eu3+ ion. By only considering the mixed Eu-Tb MOFs, it turns out that calibration curves Δ = f(T) have not the same shape according the chosen ligand although the thermal sensitivities have similar values and the value of Tm is identical. However, this study lacks an interpretation of the non-radiative deactivation paths to better understand the phenomena.

Metal–Organic Frameworks for Luminescence Thermometry

203

Fig. 6 Schematic illustration of topology effect of the inorganic network on the detection range of LnMOFs thermometers sensitive in the physiological range

A more detailed study has recently been carried out on a series of MOFs for which the terminal solvent molecules have been exchanged by ligands from the imidazole family [46]. This exchange has been performed by post-synthetic modification on single-crystal which allows to exchange part of the MOFs ligands while keeping its crystal structure. Thus, the MOF Eu0.05 Tb0.95 -NBDC (NBDC = 2-amino-1,4-benzenedicarboxylate) was chosen as pristine material, and its terminal DMF molecules were exchanged by various terminal and chelating ligands

204

T. Amiaud and H. Serier-Brault

Fig. 7 Schematic illustration of topology effect of the inorganic network on the detection range of LnMOFs thermometers sensitive in the high-temperature range

Metal–Organic Frameworks for Luminescence Thermometry

3.0

Fig. 8 Relative thermal sensitivity of Eu0.05 Tb0.95 -NBDC and its exchange analogues

Sr (% K-1)

2.5

205

50 K

EuTb- NBDC/N-donor

EuTb-NBDC

2.0 1.5 1.0 0.5

100

150

200

250

300

Temperature (K)

through single-crystal-to-single-crystal coordinating solvent exchange reactions. Temperature-dependence luminescence studies reveal that all samples are highly sensitive in the medium range with a maximum relative sensitivity of 2.6% K−1 at 190 K for Eu0.05 Tb0.95 -NBDC. In addition, a shift of 50 K of the operating temperature range is evidenced for the exchanged analogues (Fig. 8). This is attributed to the occurrence of different deactivation pathways in the exchanged analogues due to the presence of N-donor terminal or N/O-donor chelating aromatic ancillary ligands in the place of DMF terminal ligands in the pristine material.

2.1.3

Effect of the Eu/Tb Molar Ratio

In the numerous investigations concerning the LnMOFs luminescent thermometers, the cations content is always accurately determined, especially by Induced Coupled Plasma technique. However, the chemical composition was never optimized to improve the thermometric performances. In 2021, a first investigation highlighted the crucial role of the Eu3+ content in the materials to optimize relative thermal sensitivity [82]. Thus, a series of Tb1-x Eux (CH3 COO)(1,3-bdc)(H2 O)2 ·0.5H2 O compounds with six different Eu3+ composition were investigated to evidence the impact of Eu/Tb molar ratio on thermometric performances. Consequently, on the series of material, a continuous decrease of the relative thermal sensitivity was observed with the increase of Eu3+ content (Fig. 9) while a small shift of the maximal temperature Tm to lower temperature was also evidenced. Furthermore, the materials that are sensitive in the physiological range exhibit the same deactivation pathways whatever the composition, confirmed by the Mott-Seitz fitting of calibration curves Δ = f(T), which enabled to extract activation energies ranging between 500 and 620 cm−1 , and attributed to a Tb3+ -to-ligand back energy transfer. Although a relatively low thermal

206

T. Amiaud and H. Serier-Brault

Fig. 9 a Temperature-dependent emission spectra in the 150–350 K range of Tb0.95 Eu0.05 (CH3 COO)(1,3-bdc)(H2 O)2 ·0.5H2 O compound (λexc = 289 nm). b Normalized integrated emission of I1 (T3+ , 5 D4 → 7 F5 ) and I2 (Eu3+ , 5 D0 → 7 F2 ). c Relative thermal sensitivity for the mixed Eu-Tb compounds. d Corresponding temperature uncertainty. The dots mark the Sm values and the minimum δT values

sensitivity, the Eu/Tb molar ratio in mixed LnMOFs appears undoubtedly a crucial parameter to improve thermal performances of such materials.

2.2 Nd-Yb Mixed Metal–Organic Frameworks Luminescence thermometry has very promising results in the field of biomedical applications which require high resolution and high-sensitivity of nanothermometry. Indeed, temperature is known to play a crucial role in determining dynamics and properties of biosystems such as cell division rates, denaturation processes… [42].

Metal–Organic Frameworks for Luminescence Thermometry

207

High-resolution thermal sensing is also required in cancer therapeutic processes as hyperthermia, which is based on externally inducing an increase of tumor’s temperature up to cytotoxic levels (43–45 °C). For these specific applications, the requirements that must satisfy the nanothermometers are completely different from those necessary for “hot-spots” detection in nano-electronics or the monitoring of the temperature during integrated photonics devices operation, for example. Indeed, the use of excitation and emission wavelengths in the visible range of the spectrum is not well adapted for in vivo experiments because the penetration of light into tissues is limited to few hundreds of micrometers [22]. That is why systems operating in the biological windows (BWs) are requested, i.e. in the near-infrared (NIR) range. The BWs correspond to the wavelength ranges in which the absorption, and scattering of biological tissues is minimum, granting a penetration of light in the order of a few centimeters. One can distinguish three BWs: BW-I between 650 and 950 nm, BW-II between 1000 and 1400 nm and BW-III between 1550 and 1870 nm [35]. Currently, the main nanothermometers operating in the different BWs are fluorides, quantum dots or oxides and some promising in vitro experiments have been already described. The first NIR luminescent metal–organic framework for temperature sensing in the physiological range was reported in 2015 by the group of Qian [53]. In that work, they identified different considerations to taken into account for the design of such systems: (a) the selection of lanthanide ions with NIR luminescence with excitation in the NIR instead of UV or visible range, (b) the good choice of the organic ligand with a much higher triplet energy level than the accepted level of lanthanides ions to limit the energy transfer or interaction between the linker and Ln3+ ions, and (c) the reduction of some high energy chemical bonds, such as C-H, N–H, and O–H which can act as oscillators and significantly quench the NIR emission [29]. Nd3+ and Yb3+ were chosen as NIR emitters for ratiometric thermometers, and the authors have used an 808 nm excitation to pump the Nd3+ ions in their 4 F5/2 energy level. The use of Nd3+ as a NIR energy sensitizer is a common approach in the field of lanthanide-based luminescent materials, owing to its large light absorption cross section. Consequently, by considering an Nd3+ -to-Yb3+ energy transfer (Fig. 10), the emission at around 980 nm for the 2 F5/2 → 2 F7/2 of Yb3+ is obtained simultaneously with the characteristics NIR emission lines of Nd3+ at around 890, 1060 and 1350 nm for the 4 F3/2 → 4 IJ (J = 13/2, 11/2, and 9/2) [57]. To limit the presence of C-H bonds, the authors used a fluorinated organic ligand, namely 1,2,5,6-tetrafluoro1,4-benzenedicarboxylate, with a triplet excited energy state of 27 465 cm−1 . The thermometric parameter Δ was defined as the intensity ratio between the 2 F5/2 → 2 F7/2 of Yb3+ and the 4 F3/2 → 4 I11/2 of Nd3+ at 1060 nm. This first NIR ratiometric thermometer based on MOF is sensitive in the physiological range with a maximal thermal sensitivity Sm equal to 0.816% K−1 at 313 K. Subsequently, the same group has reported three others NIR MOF ratiometric thermometers with different organic ligands [99, 101, 104] to modulate the thermal sensitivity by playing on the rigidity of the linker and on the spatial distance between the lanthanide ions to facilitate the Nd3+ -to-Yb3+ energy transfer. To date, the more sensitive NIR MOF is the mixed Nd0.866 Yb0.134 -BTB (BTB = 1,3,5-benzenetrisbenzoic acid), which

208

T. Amiaud and H. Serier-Brault

Transfer

890 nm

1060 nm

4

I15/2

F5/2

980 nm

5000

2

F3/2

1350 nm

4

10000 808 nm pumping

Energy (cm-1 )

4F 5/2

4I 13/2 4I 11/2

0

4

2

F7/2

I9/2

Nd3+

Yb3+

Fig. 10 Schematic representation of energy process in mixed Nd-Yb MOFs

exhibits the MIL-103 crystal structure (MIL = Matériaux Institut Lavoisier) and with a maximal thermal sensitivity of 4.75%K−1 at 335 K (62 °C). Although the research of NIR MOF for ratiometric thermometry is still at its infancy, the potentialities of this kind of platform are very promising. Indeed, by exploiting the MOF porosity, multifunctional materials can be envisioned where the temperature monitoring can be combined with therapeutic properties (drug delivery, oxygen species formation…). Thus, NIR luminescent MOFs are good candidates for theranostics [21, 36, 94] i.e. combining therapy and diagnosis.

2.3 Eu3+ -Based Metal–Organic Frameworks Ratiometric luminescent thermometers can be also designed with single lanthanide organic frameworks when the difference between the ligand excited state T1 and the Ln3+ lowest emitting levels is inferior to 1500 cm−1 , increasing the probability of ionto-ligand back energy transfer [3]. Consequently, both the ligand and Ln3+ emission may be observed and the thermometric parameter Δ was defined as the ratio of intensities between both emissions. Thus several examples of Eu-based MOFs are reported in the literature as ratiometric thermometers but with different behaviors [2, 18, 30, 52, 54, 86]. The first example was described in 2013 by D’Vries et al. [18] who reported isostructural compounds Ln7 (3,5-DSB)4 (OH)9 (H2 O)15 ]·4H2 O (Ln = Eu, Gd, and Tb) (DSB = 3,5-disulfobenzoate). A strong thermal quenching of the DSB emissions was observed while the Eu3+ emission was almost invariant with the temperature. Consequently, the thermometric parameter was defined using the ratio of intensities between the emissions of the triplet state of the disulfobenzoic ligand and the Eu3+

Metal–Organic Frameworks for Luminescence Thermometry

209

emission, i.e., Δ = IDSB /IEu , and its exponential temperature dependence leads to a thermometer covering almost the whole visible spectrum and operating in the 10– 300 K range, with a maximum relative sensitivity of 7.14% K−1 at 65 K (value calculated by [52]). Other reported materials exhibit the contrary behavior where the decrease of the temperature prevents the ion-to-ligand back energy transfer, resulting in a stronger Eu3+ emission. In that case, the thermometric parameter is defined in the opposite way that D’Vries, namely by using the ratio of the integrated areas of the 5 D0 → 7 F2 transition (IEu ) and the ligand’s emission, i.e. Δ = IEu /ILigand . To obtain highly sensitive luminescent thermometers, a judicious choice of the organic ligand is requested. For example, Li et al. [52] used the 9-fluorenone-2,7-dicarboxylic acid (H2FCD), involving a low energy difference of 553 cm−1 between the FDC2− excited triplet state and the 5 D0 emitting level, and consequently a strong energy back transfer rate. Thus, the material operates over a wide temperature range (12 K-300 K) including the physiological range with a relative thermal sensitivity up to 2.7% K−1 at 170 K and a repeatability up to 96%. The sensitivity is comparable to that of the dual center MOF-based luminescent thermometers operating in temperature range from the cryogenic to the physiological range. Doping metal–organic frameworks by small amount of Eu3+ is also an alternative to get single lanthanide organic frameworks as ratiometric thermometer. Feng et al. SPS:refid::bib30[30] introduced Eu3+ ions in a robust Zr-UiO-66 by using postfunctionalization technique where Eu3+ ions are introduced by replacing Zr4+ in the secondary building units in situ. Samples were prepared as film in PVDF polymer (Polyvinylidene fluoride), and the authors highlighted the role of the increase of the film thickness on the sensitivity (Sm = 4.26%K−1 at 337 K). The stability of films was also tested under biological conditions by submerging MOF films in a phosphate-buffered saline buffer solution for 24 h. The choice of the organic ligand being primordial to design sensitive thermometers, the same Chinese group [54, 86] has worked on MOFs built upon a rigid symmetric organic ligand, namely [1,1, :4, ,1,, terphenyl]-3,3,, ,5, ,5,, -tetracarboxylic acid, with a triplet level equal to 23,791 cm−1 , in order to obtain a good thermal sensitivity in high temperature. Thus, they prepared the Gd-MOF doped with 5 molar % of Eu3+ with a maximal relative thermal sensitivity of 4.67% K−1 at 313 K. However, the europium content was not adjusted to modulate the sensitivity. In a second investigation, they focused the attention on the same ligand with methoxyl and methyl groups to engineer the sensitization of Ln3+ ions. If the antenna effect is not efficient for Tb3+ ions, making it unusable as luminescent thermometer, the Eu-MOF built upon the methoxyl ligand exhibit a very high relative thermal sensitivity of 7.78% K−1 at 313 K confirming the impact of organic ligand on the thermometric properties of such single lanthanide systems. Finally, as the temperature readout is obtained from the emissions of the host and a single lanthanide ion, that is possible to incorporate a second lanthanide into the MOF structure in order to develop extra functionalities.

Excited State

4

I15/2

4

455 nm

Thermally coupled energy level (4 I15/2 and 4 F9/2 ) scheme of Dy3+ and emissions originating from 4I 6 15/2 - H15/2 (455 nm) and 4 F -6 H 9/2 15/2 (458 nm) transitions

T. Amiaud and H. Serier-Brault

Thermal Equilibrium

F9/2

485 nm

210

6

H15/2

2.4 Others Metals As we previously described, europium and terbium for visible range and neodymium and ytterbium for NIR emission are the most investigated lanthanide ions for luminescent thermometers in MOFs. However, some attempts were proposed with Dy3+ [91] or d10 metals (Zn2+ , Cd2+ , Ag+ ) [34, 68]. In Dy3+ -based materials, the authors used two thermally-coupled energy levels (TCELs) of Dy3+ to generate dual emission, this strategy being well-spread for Er3+ -containing materials. The temperature-induced population re-distribution between the TCELs will result in modifications in the luminescence intensity ratio, thus realizing a self-referenced luminescent thermometer. In the case of Dy3+ -based MOFs, the low energy difference (about 991 cm−1 ) between 4 I15/2 and 4 F9/2 levels (Fig. 11) seems to be advantageous to design temperature sensors in a relatively high temperature range. The Dy-based MOF, built upon the 5(4-carboxyphenoxy) isophthalic acid, exhibits nevertheless moderate relative sensitivity of 0.42% K−1 at 473 K). By contrary, for materials containing (n − 1)d10 ns0 cations, the emission is mainly originated from the organic ligand as the cation has no emitting levels, leading thus to a single emitter thermometer. However, such systems do not present a huge interest due to their very low thermometric performances. Other temperature readouts could be, nevertheless, developed to improve thermal sensitivity [60].

2.5 Host–Guest Molecule as Second Emitter Among all luminescent thermometer MOFs described previously, all emitting centers are located in the network (metal ions or organic linker). However, MOFs are very well-known for their porous character which make them very attractive for gas storage or separation, catalysis… Indeed, the content of the porosity can be modified via

Metal–Organic Frameworks for Luminescence Thermometry

211

exchange reaction and one can easily conceive the incorporation of luminescent guest entities in the porosity. Guest entities can be of diverse nature such as organic dyes [14, 84, 93, 103, 108], polyoxometalates [74, 83] or quantum dots [55]. In such systems, one emitter is contained in the framework while the second one is located in the cavities. A first dual-emitting MOF@dye acting as ratiometric temperature sensor was synthesized by Cui et al. [14] through a “one-pot” process. Actually, via an impregnation process, they succeeded to incorporate perylene molecules into a porous europium MOF. With the increase of temperature from 20 to 80 °C, the luminescence intensity at 473 nm of perylene dye in the composite material substantially decreases due to the thermal activation of nonradiative pathways, while the intensity of the 5 D0 → 7 F2 transition of Eu3+ centered at 615 nm increases. The unique energy-transfer between perylene molecules and Eu3+ cations renders the perylene@ZJU-88 very sensitive in the physiological range (Sm = 1.28% K−1 at 293 K). The same strategy was adopted to impregnate a terbium MOF with a coumarin derivative [93], the composite material exhibiting also a high thermal sensitivity in the physiological range, namely 4.48% K−1 at 300 K. To avoid the excitation in the UV domain that can be harmful for biological applications, Wan et al. [84] used two-photon luminescent dyes to afford the possibility that excitation and emission light are located in biological windows. Thus, the luminescent composite ZJU-28@DPASD (ZJU-28 = (Me2 NH2 )3[In3 (BTB)4 ]·12DMF·22H2 O, DPASD = 4-[4-(diphenylamino) styryl]1-dodecylpyridinium), which is excited by 1064 nm light in the second biological window and emits at 650 nm in the first one. In that case, the thermometric properties are based to a single emission band which has a linear dependence with the temperature between 20 and 60 °C with a relative thermal sensitivity ranging from 1.52% K−1 to 3.39%. K−1 in the 20–60 °C range. As all single emitter luminescent thermometers, the performances depend on the emitter concentration and an additional two-photon dye could be introduced within the pores to design a ratiometric luminescent thermometer. The richness of luminescent organic dyes and porous MOFs offers wide choices for realizing a great variety of MOF@dye composites, whose emission range can be elaborately designed and expanded to tune the energy transfer between the MOF and dye, and thus to optimize the properties of the dual-emitting MOF@dye composite. Inorganic entities such as polyoxometalates (POMs) or quantum dots can be also incorporated within MOFs pores [31, 55, 74, 83]. POMs can be described as anionic soluble metal oxides of early transition metals (usually W, Mo, V) [26], which can also contain lanthanide cations, the most intensively investigated being [EuW10 O39 ]9− [7]. In a recent work, [EuW10 O39 ]9− was incorporated in the cavities proposed by a Tb-MOF, and thermometric properties were investigated between 200 and 320 K [83], revealing a relative thermal sensitivity of 2.68% K−1 at 300 K. Perovskites CsPbBr3 quantum dots were also encapsulated into a Eu-MOF cavities, offering dual emitting materials with the 5 D0 → 7 F2 transition of the Eu3+ cation associated to a broad emission ranging from 500 to 580 nm that originates from CsPbBr3 [55]. The opposite variation of the temperature-dependence of each emission conducts to

212

T. Amiaud and H. Serier-Brault

Table 4 List of main guest@MOFs composites investigated as ratiometric luminescent thermometers with their temperature operating range and their maximal relative thermal sensitivity (Sm ) Material

Temperature range (K)

Sm (% K−1 )

References

CsPbBr3 @Eu-BTC

293–373

4.90

[55]

EuW10 @Tb-TATB

200–320

2.70

[83]

perylene@ZJU-88

293–353

1.28

[14]

Rh101@UiO-67

293–333

1.19

[108]

Coumarin@Tb-TATAB

100–300

4.48

[93]

C-QT@UiO-66-COOH

97–297

1.30

[31]

TbMOF@3% Eu_tfac

225–375

2.59

[43]

a very high relative thermal sensitivity with Sm = 3.9% K−1 at 373 K (and 3.1% K−1 at 300 K). Interestingly, all luminescent thermometers implying a guest entity and a lanthanide cation located in the crystal framework exhibit a good relative thermal sensitivity in the physiological domain (Table 4). Nonetheless, any investigation has been performed on these systems to identify their potentialities in bioimaging for instance, or others biological applications.

3 Conclusion and Perspectives Metal–organic frameworks are versatile materials that offer many opportunities in the field of luminescence thermometry. Indeed, the emitting centers can be the inorganic cluster, the organic ligand used to build the network or a host–guest entity located into the cavities of the MOF, enabling many combinations to generate a ratiometric luminescent thermometer. Currently, most of the investigations concern the mixed Eu-Tb MOFs, excitable in the UV region and emitting in the visible range. A thorough investigation of crystal structure of many reported MOFs evidences the role of the inorganic network topology on the operating temperature range of the thermometer. Thus, a high connectivity of [LnOx ] polyhedral like layers or chains, favors a thermal sensitivity in the cryogenic range, probably due to a low distance between Ln3+ cations. By contrary when the connectivity decreases, the thermal sensitivity of the material is more located in the medium range or physiological range. However, additional structural parameters must be identified in order to be more predictable when a new Ln-bearing MOF is designed. Among these parameters, the organic ligand is non negligible and efforts must be focused in the future on the tailoring of the ligand to design MOF luminescent thermometer with targeted performances. Furthermore, the porosity of MOFs must be more exploited to combine the measure of the temperature with another functionality such as pH measurement, O2 sensing, drug delivery, or magnetic resonance imaging… Actually, MOFs are

Metal–Organic Frameworks for Luminescence Thermometry

213

interesting platforms to design multifunctional materials, especially for biomedical applications. However, the additional functionality does not have to deteriorate the thermometric performances. Finally, the elaboration of such systems will necessitate the control of the particle size in the nano-submicron range, the surface chemistry, the stability in physiological media, and also the toxicity of the materials. Contrary to inorganic materials, MOFs have undoubtedly a certain delay, from a practical point of view, but their versatility offers a large field of research for the future. Acknowledgements This work was partially developed within the scope of the project THERMOF, ANR-18-CE09-0008-01 financed by the French National Research Agency (ANR).

Abbreviations UiO ZJU BTC Tfac TbMOF TATB TATAB HY dsb pda notp 1,3-bdc btb DMBDC hfa dpbp bpcd bpydc bpdc ad cbpi tpi qptca bdc dstp cpda BPTC NDC PIA AD

University of Oslo Zhejiang University Benzene tricarboxylic acid Trifluoroacetylacetonate [Tb2(bpydc)3(H2 O)3]·nDMF, 2,2, -bipyridine-5,5, -dicarboxylic acid 4,4, ,4,, -S-Triazine-2,4,6-triyl-tribenzoate 4,4, ,4,, -S-triazine-1,3,5-triyltri-p-aminobenzoic acid 5-Hydroxy-1,2,4-benzenetricarboxylic acid 3,5-Disulfobenzoate 1,4-Phenylenediacetic acid 1,4,7-Triazacyclononane-1,4,7-triyl-tris(methylenephosphonic) acid 1,3-Benzenedicarboxylic acid 1,3,5-Tris(4-carboxyphenyl)benzene 2,5-Diméthoxy-1,4-benzenedicarboxylic acid Hexafluoro acetylacetonate 4,4-Diphenylphosphoryl 1,4-Bis(pyridinil-4-carboxylato)-1,4-dimethylbenzene 2,2, -Bipyridine-5,5’-dicarboxylic acid Biphenyl-4,4, -dicarboxylic acid Adeninate 1,3-Bis(4-carboxyphenyl)imidazolium 5-(4-(Tetrazol-5-yl)phenyl) isophthalic acid 1,1, :4, ,1,, :4,, ,1,,, -Quaterphenyl-3,3,,, ,5,5,,, -tetracarboxylic acid 1,4-Benzenedicarboxylic acid 2,4-(2,2, :6, ,2,, -Terpyridin-4, -yl)-benzenedisulfonic acid 5-(4-Carboxyphényl)-2,6-pyridinedicarboxylic acid Biphenyl-3,3, ,5,5, -tetracarboxylic acid 1,4-Naphthalenedicarboxylic acid 5-(Pyridin-4-yl)isophthalate Adipic acid

214

phth cbpp BDC-NH2 hfa dpbp HPIDC ox D-cam Himdc

T. Amiaud and H. Serier-Brault

1,2-Benzenedicarboxylic acid 2,6-Di(2, ,4, -dicarboxylphenyl)pyridine 2-Amino-1,4-benzenedicarboxylic acid Hexafluoroacetyacetone 4,4, -Bis(diphenylphosphoryl)biphenyl 2-Pyridin-4-yl-4,5-imidazoledicarboxylic acid Oxalic acid D-camphoric acid 4,5-Imidazoledicarboxylic acid

References 1. D. Ananiass, L.D. Carlos, J. Rocha, Cryogenic nanothermometer based on the MIL-103(Tb, Eu) metal–organic framework 3, 1967–1971 (2016). https://doi.org/10.1002/ejic.201501195 2. D. Andriotou, S.A. Diamantis, A. Zacharia et al., Dual emission in a ligand and metal codoped lanthanide-organic framework: color tuning and temperature dependent luminescence. Molecules 25, 523 (2020). https://doi.org/10.3390/molecules25030523 3. L. Armelao, S. Quici, F. Barigelletti et al., Design of luminescent lanthanide complexes: from molecules to highly efficient photo-emitting materials. Coord. Chem. Rev. 254, 487–505 (2010) 4. J.E. Bachman, Z.P. Smith, T. Li et al., Enhanced ethylene separation and plasticization resistance in polymer membranes incorporating metal-organic framework nanocrystals. Nat. Mater. 15, 845–849 (2016). https://doi.org/10.1038/nmat4621 5. K. Behrens, S.S. Mondal, R. Nöske et al., Microwave-assisted synthesis of defects metalimidazolate-amide-imidate frameworks and improved CO2 capture. Inorg. Chem. 54, 10073– 10080 (2015). https://doi.org/10.1021/acs.inorgchem.5b01952 6. A. Benayas, B. Del Rosal, A. Pérez-Delgado et al., Nd:YAG near-infrared luminescent nanothermometers. Adv. Opt. Mater. 3 (2015). https://doi.org/10.1002/adom.201400484 7. G. Blasse, G.J. Dirksen, F. Zonnevijlle, The luminescence of some lanthanide decatungstates and other polytungstates. J. Inorg. Nucl. Chem. 43, 2847–2853 (1981). https://doi.org/ 10.1016/0022-1902(81)80629-X 8. C.D.S. Brites, P.P. Lima, L.D. Carlos, Tuning the sensitivity of Ln3+-based luminescent molecular thermometers through ligand design. J. Lumin. 169, 497–502 (2016). https://doi. org/10.1016/j.jlumin.2015.01.025 9. C.D.S. Brites, A. Millán, L.D. Carlos, Lanthanides in luminescent thermometry, in Handbook on the Physics and Chemistry of Rare Earths, pp 339–427 (2016b) 10. A. Cadiau, C.D.S. Brites, P.M.F.J. Costa et al., Ratiometric nanothermometer based on an emissive Ln3+-organic framework. ACS Nano 7, 7213–7218 (2013). https://doi.org/10.1021/ nn402608w 11. A.K. Cheetham, G. Férey, T. Loiseau, Open-framework inorganic materials. Angew Chemie Int. Ed. 38, 3268–3292 (1999). https://doi.org/10.1002/(SICI)1521-3773(19991115)38:22% 3c3268::AID-ANIE3268%3e3.0.CO;2-U 12. Y. Cheng, Y. Gao, H. Lin et al., Strategy design for ratiometric luminescence thermometry: circumventing the limitation of thermally coupled levels. J. Mater. Chem. C 6, 7462–7478 (2018). https://doi.org/10.1039/c8tc02401d 13. T. Chuasaard, A. Ngamjarurojana, S. Surinwong et al., Lanthanide coordination polymers of mixed phthalate/adipate for ratiometric temperature sensing in the upper-intermediate temperature range. Inorg. Chem. 57, 2620–2630 (2018). https://doi.org/10.1021/acs.inorgchem.7b0 3016

Metal–Organic Frameworks for Luminescence Thermometry

215

14. Y. Cui, R. Song, J. Yu et al., Dual-emitting MOF⊃dye composite for ratiometric temperature sensing. Adv. Mater. 27, 1420–1425 (2015). https://doi.org/10.1002/adma.201404700 15. Y. Cui, H. Xu, Y. Yue et al., A luminescent mixed-lanthanide metal-organic framework thermometer. J. Am. Chem. Soc. 134, 3979–3982 (2012). https://doi.org/10.1021/ja2108036 16. Y. Cui, F. Zhu, B. Chen, G. Qian, Metal-organic frameworks for luminescence thermometry. Chem. Commun. 51 (2015). https://doi.org/10.1039/c5cc00718f 17. Y. Cui, W. Zou, R. Song et al., A ratiometric and colorimetric luminescent thermometer over a wide temperature range based on a lanthanide coordination polymer. Chem. Commun. 50, 719–721 (2014). https://doi.org/10.1039/C3CC47225F 18. R.F. D’Vries, S. Álvarez-García, N. Snejko et al., Multimetal rare earth MOFs for lighting and thermometry: tailoring color and optimal temperature range through enhanced disulfobenzoic triplet phosphorescence. J. Mater. Chem. C 1, 6316 (2013). https://doi.org/10.1039/c3tc30 858h 19. J.B. Decoste, G.W. Peterson, Metal-organic frameworks for air purification of toxic chemicals. Chem. Rev. 114, 5695–5727 (2014) 20. B. del Rosal, D. Ruiz, I. Chaves-Coira et al., In vivo contactless brain nanothermometry. Adv. Funct. Mater. 28, 1806088 (2018). https://doi.org/10.1002/adfm.201806088 21. B. del Rosal, I. Villa, D. Jaque, F. Sanz-Rodríguez, In vivo autofluorescence in the biological windows: the role of pigmentation. J. Biophotonics 9, 1059–1067 (2016). https://doi.org/10. 1002/jbio.201500271 22. B. del Rosal, E. Ximendes, U. Rocha, D. Jaque, In vivo luminescence nanothermometry: from materials to applications. Adv. Opt. Mater. 5 (2017). https://doi.org/10.1002/adom.201 600508 23. T. Devic, O. David, M. Valls et al., An illustration of the limit of the metal organic framework’s isoreticular principle using a semirigid tritopic linker obtained by “click” chemistry. J. Am. Chem. Soc. 129, 12614–12615 (2007). https://doi.org/10.1021/ja0744091 24. T. Devic, V. Wagner, N. Guillou et al., Synthesis and characterization of a series of porous lanthanide tricarboxylates. Microporous Mesoporous Mater. 140, 25–33 (2011). https://doi. org/10.1016/j.micromeso.2010.10.022 25. S. Diring, S. Furukawa, Y. Takashima et al., Controlled multiscale synthesis of porous coordination polymer in nano/micro regimes. Chem. Mater. 22, 4531–4538 (2010). https://doi.org/ 10.1021/cm101778g 26. A. Dolbecq, E. Dumas, C.R. Mayer, P. Mialane, Hybrid organic−inorganic polyoxometalate compounds: from structural diversity to applications. Chem. Rev. 110, 6009–6048 (2010). https://doi.org/10.1021/cr1000578 27. M. Dramicanin, Luminescence Thermometry: Methods, Materials, and Applications (Woodhead Publishing, 2018) 28. M. Eddaoudi, J. Kim, N. Rosi et al., Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science 295, 469–472 (2002). https://doi.org/10.1126/science.1067208 29. S.V. Eliseeva, J.-C.G. Bünzli, Lanthanide luminescence for functional materials and biosciences. Chem. Soc. Rev. 39, 189–227 (2010). https://doi.org/10.1039/B905604C 30. J.-F. Feng, T.-F. Liu, J. Shi et al., Dual-emitting UiO-66(Zr&Eu) metal−organic framework films for ratiometric temperature sensing. ACS Appl. Mater. Interfaces 10, 48 (2018). https:// doi.org/10.1021/acsami.8b04889 31. J.F. Feng, S.Y. Gao, J. Shi et al., C-QDs@UiO-66-(COOH)2 composite film via electrophoretic deposition for temperature sensing. Inorg. Chem. 57, 2447–2454 (2018). https://doi.org/10. 1021/acs.inorgchem.7b02595 32. A.D.G. Firmino, R.F. Mendes, D. Ananias et al., Microwave synthesis of a photoluminescent metal-organic framework based on a rigid tetraphosphonate linker. Inorganica Chim Acta 455, 584–594 (2017). https://doi.org/10.1016/j.ica.2016.05.029 33. Y.-H. Han, C.-B. Tian, Q.-H. Li, S.-W. Du, Highly chemical and thermally stable luminescent EuxTb1-x MOF materials for broad-range pH and temperature sensors. J. Mater. Chem. C 2, 8065–8070 (2014). https://doi.org/10.1039/C4TC01336K

216

T. Amiaud and H. Serier-Brault

34. Y.-C. He, N. Xu, X. Zheng et al., A silver(I) coordination polymer luminescent thermometer. Dye Pigment 136, 577–582 (2017). https://doi.org/10.1016/j.dyepig.2016.09.021 35. E. Hemmer, A. Benayas, F. Légaré, F. Vetrone, Exploiting the biological windows: current perspectives on fluorescent bioprobes emitting above 1000 nm. Nanoscale Horizons 1, 168– 184 (2016) 36. E. Hemmer, N. Venkatachalam, H. Hyodo et al., Upconverting and NIR emitting rare earth based nanostructures for NIR-bioimaging. Nanoscale 5, 11339–11361 (2013) 37. P. Horcajada, R. Gref, T. Baati et al., Metal-organic frameworks in biomedicine. Chem. Rev. 112, 1232–1268 (2012). https://doi.org/10.1021/cr200256v 38. P. Horcajada, C. Serre, M. Vallet-Regí et al., Metal-organic frameworks as efficient materials for drug delivery. Angew Chemie Int Ed 45, 5974–5978 (2006). https://doi.org/10.1002/anie. 200601878 39. B.F. Hoskins, R. Robson, Infinite polymeric frameworks consisting of three dimensionally linked rod-like segments. J. Am. Chem. Soc. 111, 5962–5964 (1989). https://doi.org/10.1021/ ja00197a079 40. A.J. Howarth, Y. Liu, J.T. Hupp, O.K. Farha, Metal-organic frameworks for applications in remediation of oxyanion/cation-contaminated water. CrystEngComm 17, 7245–7253 (2015). https://doi.org/10.1039/c5ce01428j 41. Z. Hu, B.J. Deibert, J. Li, Luminescent metal-organic frameworks for chemical sensing and explosive detection. Chem. Soc. Rev. 43, 5815–5840 (2014). https://doi.org/10.1039/c4cs00 010b 42. D. Jaque, F. Vetrone, Luminescence nanothermometry. Nanoscale 4, 4301–4326 (2012). https://doi.org/10.1039/c2nr30764b 43. A.M. Kaczmarek, Y.Y. Liu, C. Wang et al., Grafting of a Eu3+-tfac complex on to a Tb3+metal organic framework for use as a ratiometric thermometer. Dalt. Trans. 46, 12717–12723 (2017). https://doi.org/10.1039/c7dt02042b 44. N.A. Khan, S.H. Jhung, Synthesis of metal-organic frameworks (MOFs) with microwave or ultrasound: rapid reaction, phase-selectivity, and size reduction. Coord. Chem. Rev. 285, 11–23 (2015) 45. M. Kondo, T. Yoshitomi, H. Matsuzaka et al., Three-dimensional framework with channeling cavities for small molecules: {[M2 (4, 4, -bpy)3 (NO3 )4 ]·xH2 O}n(M = Co, Ni, Zn). Angew Chemie Int Ed English 36, 1725–1727 (1997). https://doi.org/10.1002/anie.199717251 46. A. Kourtellaris, W. Lafargue-Dit-Hauret, F. Massuyeau, C. Latouche, A.J. Tasiopoulos, H. Serier-Brault, Tuning of thermometric performances of mixed Eu-Tb metal-organic frameworks through single-crystal coordinating solvent exchange reactions. Adv. Opt. Mater. (under revision) (2022) 47. L.E. Kreno, K. Leong, O.K. Farha et al., Metal-organic framework materials as chemical sensors. Chem. Rev. 112, 1105–1125 (2012) 48. J. Lee, O.K. Farha, J. Roberts et al., Metal-organic framework materials as catalysts. Chem. Soc. Rev. 38, 1450–1459 (2009). https://doi.org/10.1039/b807080f 49. H. Li, M. Eddaoudi, M. O’Keeffe, O.M. Yaghi, Design and synthesis of an exceptionally stable and highly porous metal-organic framework. Nature 402, 276–279 (1999). https://doi. org/10.1038/46248 50. J.R. Li, R.J. Kuppler, H.C. Zhou, Selective gas adsorption and separation in metal-organic frameworks. Chem. Soc. Rev. 38, 1477–1504 (2009). https://doi.org/10.1039/b802426j 51. J.R. Li, J. Sculley, H.C. Zhou, Metal-organic frameworks for separations. Chem. Rev. 112, 869–932 (2012) 52. L. Li, Y. Zhu, X. Zhou et al., Visible-light excited luminescent thermometer based on single lanthanide organic frameworks. Adv. Funct. Mater. 26, 8677–8684 (2016). https://doi.org/10. 1002/adfm.201603179 53. X. Lian, D. Zhao, Y. Cui et al., A near infrared luminescent metal-organic framework for temperature sensing in the physiological range. Chem. Commun. 51, 17676–17679 (2015). https://doi.org/10.1039/C5CC07532G

Metal–Organic Frameworks for Luminescence Thermometry

217

54. J. Liu, X. Han, Y. Lu et al., Isostructural single- and dual-lanthanide metal-organic frameworks based on substituent-group-modifying tetracarboxylate ligands for ratiometric temperature sensing. Inorg. Chem. 60, 4133–4143 (2021). https://doi.org/10.1021/acs.inorgchem.1c00310 55. J. Liu, Y. Zhao, X. Li et al., Dual-emissive CsPbBr 3@Eu-BTC composite for self-calibrating temperature sensing application. Cryst. Growth Des. 20, 454–459 (2020). https://doi.org/10. 1021/acs.cgd.9b01374 56. X. Liu, S. Akerboom, M. De Jong et al., Mixed-lanthanoid metal–organic framework for ratiometric cryogenic temperature sensing 24, 11323–11329 (2015). https://doi.org/10.1021/ acs.inorgchem.5b01924 57. A. Lupei, V. Lupei, A. Ikesue et al., Nd → Yb energy transfer in (Nd, Yb):Y2O3 transparent ceramics, in Optical Materials (Elsevier B.V., 2010), pp. 1333–1336 58. L. Marciniak, A. Bednarkiewicz, K. Trejgis et al., Enhancing the sensitivity of a Nd3+, Yb3+:YVO4 nanocrystalline luminescent thermometer by host sensitization. Phys. Chem. Chem. Phys. 21, 10532–10539 (2019). https://doi.org/10.1039/c9cp01806a 59. J.A. Mason, M. Veenstra, J.R. Long, Evaluating metal-organic frameworks for natural gas storage. Chem. Sci. 5, 32–51 (2014) 60. F.E. Maturi, C.D.S. Brites, E.C. Ximendes et al., Going above and beyond: a tenfold gain in the performance of luminescence thermometers joining multiparametric sensing and multiple regression. Laser Photon. Rev. 2100301 (2021). https://doi.org/10.1002/lpor.202100301 61. X. Meng, S.-Y. Song, X.-Z. Song et al., A Eu/Tb-codoped coordination polymer luminescent thermometer. Inorg. Chem. Front. 1, 757–760 (2014). https://doi.org/10.1039/C4QI00122B 62. I. Mihalcea, C. Volkringer, N. Henry, T. Loiseau, Series of mixed uranyl-lanthanide (Ce, Nd) organic coordination polymers with aromatic polycarboxylates linkers. Inorg. Chem. 51, 9610–9618 (2012). https://doi.org/10.1021/ic3005757 63. K. Miyata, Y. Konno, T. Nakanishi et al., Chameleon luminophore for sensing temperatures: control of metal-to-metal and energy back transfer in lanthanide coordination polymers. Angew Chemie Int. Ed. 52, 6413–6416 (2013). https://doi.org/10.1002/anie.201301448 64. J.E. Mondloch, M.J. Katz, W.C. Isley et al., Destruction of chemical warfare agents using metal-organic frameworks. Nat. Mater. 14, 512–516 (2015). https://doi.org/10.1038/nma t4238 65. I. N’Dala-Louika, D. Ananias, C. Latouche et al., Ratiometric mixed Eu-Tb metal-organic framework as a new cryogenic luminescent thermometer. J. Mater. Chem. C 5, 10933–10937 (2017). https://doi.org/10.1039/c7tc03223d 66. Y. Pan, H.Q. Su, E.L. Zhou et al., A stable mixed lanthanide metal-organic framework for highly sensitive thermometry. Dalt. Trans. 48, 3723–3729 (2019). https://doi.org/10.1039/c9d t00217k 67. T.Y. Popelensky, V.V. Utochnikova, How does the ligand affect the sensitivity of the luminescent thermometers based on Tb-Eu complexes. Dalt. Trans. 49, 12156–12160 (2020). https:// doi.org/10.1039/d0dt02238a 68. Y.J. Qi, Y.J. Wang, X.X. Li et al., Two d10 metal-organic frameworks as low-temperature luminescent molecular thermometers. Cryst. Growth Des. 18, 7383–7390 (2018). https://doi. org/10.1021/acs.cgd.8b00975 69. X. Rao, T. Song, J. Gao et al., A Highly Sensitive Mixed Lanthanide Metal–Organic Framework Self-Calibrated Luminescent Thermometer, 0–5 (2013) 70. M. Ren, C.D.S. Brites, S. Bao et al., A cryogenic luminescent ratiometric thermometer based on a lanthanide phosphonate dimer † 898, 8480–8484 (2015). https://doi.org/10.1039/c5tc01 468a 71. J. Rocha, C.D.S. Brites, L.D. Carlos, Lanthanide organic framework luminescent thermometers. Chem. A Eur. J. 22, 14782–14795 (2016). https://doi.org/10.1002/chem.201 600860 72. J. Rocha, L.D. Carlos, F.A.A. Paz, D. Ananias, Luminescent multifunctional lanthanidesbased metal–organic frameworks. Chem. Soc. Rev. 40, 926–940 (2011). https://doi.org/10. 1039/C0CS00130A

218

T. Amiaud and H. Serier-Brault

73. N.L. Rosi, J. Kim, M. Eddaoudi et al., Rod packings and metal-organic frameworks constructed from rod-shaped secondary building units. J. Am. Chem. Soc. 127, 1504–1518 (2005). https://doi.org/10.1021/ja045123o 74. W. Salomon, A. Dolbecq, C. Roch-Marchal et al., A multifunctional dual-luminescent polyoxometalate@metal-organic framework EuW10@UiO-67 composite as chemical probe and temperature sensor. Front. Chem. 6, 425 (2018). https://doi.org/10.3389/fchem.2018. 00425 75. X. Shen, Y. Lu, B. Yan, Lanthanide Complex Hybrid System for Fluorescent Sensing as Thermometer, 1–5. https://doi.org/10.1002/ejic.201403106 76. A. Skripka, A. Benayas, R. Marin et al., Double rare-earth nanothermometer in aqueous media: opening the third optical transparency window to temperature sensing. Nanoscale 9, 3079–3085 (2017). https://doi.org/10.1039/c6nr08472a 77. M.C. So, G.P. Wiederrecht, J.E. Mondloch et al., Metal-organic framework materials for light-harvesting and energy transfer. Chem. Commun. 51, 3501–3510 (2015) 78. N. Stock, S. Biswas, Synthesis of metal-organic frameworks (MOFs): routes to various MOF topologies, morphologies, and composites. Chem. Rev. 112, 933–969 (2012). https://doi.org/ 10.1021/cr200304e 79. L. Sun, M.G. Campbell, M. Dinc˘a, Electrically conductive porous metal-organic frameworks. Angew Chemie Int. Ed. 55, 3566–3579 (2016). https://doi.org/10.1002/anie.201506219 80. E.A. Tomic, Thermal stability of coordination polymers. J. Appl. Polym. Sci. 9, 3745–3752 (1965). https://doi.org/10.1002/app.1965.070091121 81. D.J. Tranchemontagne, J.R. Hunt, O.M. Yaghi, Room temperature synthesis of metal-organic frameworks: MOF-5, MOF-74, MOF-177, MOF-199, and IRMOF-0. Tetrahedron 64, 8553– 8557 (2008). https://doi.org/10.1016/j.tet.2008.06.036 82. V. Trannoy, A.N. Carneiro Neto, C.D.S. Brites et al., Engineering of mixed Eu 3+ /Tb 3+ metalorganic frameworks luminescent thermometers with tunable sensitivity. Adv. Opt. Mater. 9, 2001938 (2021). https://doi.org/10.1002/adom.202001938 83. C. Viravaux, O. Oms, A. Dolbecq et al., Temperature sensors based on europium polyoxometalate and mesoporous terbium metal–organic framework. J. Mater. Chem. C 9, 8323–8328 (2021). https://doi.org/10.1039/d1tc01532j 84. Y. Wan, T. Xia, Y. Cui et al., A two-photon luminescent dye-loaded metal-organic framework for physiological temperature sensing within biological windows. ChemPlusChem 82, 1320– 1325 (2017). https://doi.org/10.1002/cplu.201700438 85. J.L. Wang, C. Wang, W. Lin, Metal-organic frameworks for light harvesting and photocatalysis. ACS Catal. 2, 2630–2640 (2012) 86. S. Wang, J. Jiang, Y. Lu et al., Ratiometric fluorescence temperature sensing based on singleand dual-lanthanide metal-organic frameworks. J. Lumin. 226, 117418 (2020). https://doi. org/10.1016/j.jlumin.2020.117418 87. Z. Wang, D. Ananias, A. Carné-sánchez et al., Lanthanide—Organic Framework Nanothermometers Prepared by Spray-Drying. https://doi.org/10.1002/adfm.201500518 88. Z. Wang, D. Ananias, A. Carné-Sánchez et al., Lanthanide-organic framework nanothermometers prepared by spray-drying. Adv. Funct. Mater. 25, 2824–2830 (2015). https://doi. org/10.1002/adfm.201500518 89. Y. Wei, R. Sa, Q. Li, K. Wu, Highly stable and sensitive LnMOF ratiometric thermometers constructed with mixed ligands. Dalt. Trans. 44, 3067–3074 (2015). https://doi.org/10.1039/ C4DT03421J 90. L.-L. Wu, J. Zhao, H. Wang, J. Wang, A lanthanide(iii) metal–organic framework exhibiting ratiometric luminescent temperature sensing and tunable white light emission. CrystEngComm 18, 4268–4271 (2016). https://doi.org/10.1039/C5CE02444G 91. T. Xia, Y. Cui, Y. Yu et al., A luminescent ratiometric thermometer based on thermally coupled levels of Dy-MOF. J. Mater. Chem. C 22, 14782 (2017). https://doi.org/10.1039/C7TC00921F 92. T. Xia, Z. Shao, X. Yan et al., Tailoring the triplet level of isomorphic Eu/Tb mixed MOFs for sensitive temperature sensing. Chem. Commun. (2021). https://doi.org/10.1039/D1CC00 297J

Metal–Organic Frameworks for Luminescence Thermometry

219

93. T. Xia, T. Song, Y. Cui et al., A dye encapsulated terbium-based metal-organic framework for ratiometric temperature sensing. Dalt. Trans. 45, 18689–18695 (2016). https://doi.org/10. 1039/c6dt03674k 94. E.C. Ximendes, W.Q. Santos, U. Rocha et al., Unveiling in vivo subcutaneous thermal dynamics by infrared luminescent nanothermometers. Nano. Lett. 16 (2016). https://doi.org/ 10.1021/acs.nanolett.5b04611 95. O.M. Yaghi, H. Li, Hydrothermal synthesis of a metal-organic framework containing large rectangular channels. J. Am. Chem. Soc. 117, 10401–10402 (1995). https://doi.org/10.1021/ ja00146a033 96. Y. Yang, L. Chen, F. Jiang et al., A family of doped lanthanide metal–organic frameworks for wide-range temperature sensing and tunable white light emission. J. Mater. Chem. C 5, 1981–1989 (2017). https://doi.org/10.1039/C6TC05316E 97. Y. Yang, H. Huang, Y. Wang, A family of mixed-lanthanide metal–organic framework thermometers in a wide temperature range. Dalt. Trans. 47, 13384–13390 (2018). https://doi.org/ 10.1039/C8DT03182G 98. Y. Yang, Y. Wang, Y. Feng et al., Three isostructural Eu3+/Tb3+ co-doped MOFs for widerange ratiometric temperature sensing. Talanta 208, 3–8 (2020). https://doi.org/10.1016/j.tal anta.2019.120354 99. D. Yue, J. Zhang, D. Zhao et al., Ratiometric near infrared luminescent thermometer based on lanthanide metal-organic frameworks. J. Solid State Chem. 241, 99–104 (2016). https:// doi.org/10.1016/j.jssc.2016.06.005 100. C. Zaldo, Lanthanide-Based Luminescent Thermosensors: From Bulk to Nanoscale (Elsevier Inc., 2018) 101. D. Zhao, X. Han, S. Wang et al., 808 nm-light-excited near-infrared luminescent lanthanide metal-organic frameworks for highly sensitive physiological temperature sensing. Chem. A Eur. J. 26, 3145–3151 (2020). https://doi.org/10.1002/chem.201905216 102. D. Zhao, X. Rao, J. Yu et al., Design and synthesis of an MOF thermometer with high sensitivity in the physiological temperature range. Inorg. Chem. 54, 11193–11199 (2015). https://doi.org/10.1021/acs.inorgchem.5b01623 103. D. Zhao, D. Yue, K. Jiang et al., Ratiometric dual-emitting MOF⊃dye thermometers with a tunable operating range and sensitivity. J. Mater. Chem. C 5, 1607–1613 (2017). https://doi. org/10.1039/C6TC05203G 104. D. Zhao, J. Zhang, D. Yue et al., A highly sensitive near-infrared luminescent metal-organic framework thermometer in the physiological range. Chem. Commun. 52, 8259–8262 (2016). https://doi.org/10.1039/c6cc02471h 105. S.-N. Zhao, L.-J. Li, X.-Z. Song et al., Lanthanide ion codoped emitters for tailoring emission trajectory and temperature sensing. Adv. Funct. Mater. 25, 1463–1469 (2015). https://doi.org/ 10.1002/adfm.201402061 106. X. Zhao, X. Bu, T. Wu et al., Selective anion exchange with nanogated isoreticular positive metal-organic frameworks. Nat. Commun. 4, 1–9 (2013). https://doi.org/10.1038/ncomms 3344 107. Y. Zhou, B. Yan, Ratiometric detection of temperature using responsive dual-emissive MOF hybrids. J. Mater. Chem. C 3, 9353–9358 (2015). https://doi.org/10.1039/c5tc02004b 108. Y. Zhou, D. Zhang, J. Zeng et al., A luminescent lanthanide-free MOF nanohybrid for highly sensitive ratiometric temperature sensing in physiological range. Talanta 181, 410–415 (2018). https://doi.org/10.1016/j.talanta.2018.01.024

Luminescent Nanothermometers Operating Within Biological Windows Albenc Nexha, Maria Cinta Pujol Baiges, and Joan Josep Carvajal Martí

Abstract Luminescent thermometers are continuously replacing the traditional contact thermometers due to their ability to sense temperature down to the nanodimensional scale with high spatial and temporal resolutions. Due to this, these thermometers are being applied in multiple fields, with biomedicine being a key development. In this vein, several types of luminescent thermometers are growing as potential candidates for biomedical applications. These thermometers operate within the biological windows regions, where scattering and absorption from biological tissues, are minimized. Here, we summarize and compare the performance of all types of nanothermometers operating within the biological windows, highlighting their limitations and advantages. In addition, we emphasize strategies on how to improve the performance of these nanothermometers within the biological windows, with the goal of rendering them potential candidates for biomedical applications. Keywords Luminescent thermometers · Quantum dots · Nanodiamonds · Fluorescent gold nanoclusters · Transition metals · Lanthanide doped nanoparticles · Biological Windows · Classes of luminescent thermometry · Relative thermal sensitivity · Quantum yield · In-vivo biomedical applications

1 Biological Windows Non-contact based luminescent thermometers (LTs) are continuously growing as potential alternatives to traditional contact thermometers. Drawbacks of contact thermometers, notably at the nanodimensional scale, have urged the need for LTs. These thermometers can resolve the temperature within a medium with micro or nano spatial resolutions [1]. In addition, they exhibit temperature resolutions (minimum A. Nexha (B) INM-Leibniz Institute for New Materials, 66123 Saarbrücken, Germany e-mail: [email protected] M. C. Pujol Baiges · J. J. Carvajal Martí Departament de Química Física i Inorgànica, Universitat Rovira i Virgili, PhO2TO, 43007 Tarragona, Spain © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_6

221

222

A. Nexha et al.

temperature that a thermometer can differentiate) around < 0.1 K, followed by high thermal sensitivities (the relative change of the thermometric parameter per degree of temperature change) in the range > 1% K−1 , and temporal resolution (the minimum time that a thermometer is able to detect a temperature change within a medium) in less than 10 ms [1]. Moreover, they can operate in severe mediums, for example in biological fluids, cryogenic temperatures, fast-moving objects or electromagnetic fields, without hampering their performance or affecting the medium in which they are immersed [1]. Due to these properties, LTs are constantly being applied in multiple fields, ranging from early-diagnostic of tumour cells to brain activity, and therapeutic applications [1, 2]. For the development of a successful LT, particularly at biomedicine, the list of requirements is exhaustive. These include bright luminescence, photostability, biocompatibility, and biodegradability arising from materials with sizes preferably below 100 nm [1]. In addition, their optical properties (particularly their excitation and emission wavelengths) should permit deep-tissue penetration depths within biological mediums [1]. In this vein, the application of LTs in biomedicine is largely dependent on the choice of the excitation and emission wavelengths [3]. A proper selection of these wavelengths allows to address issues related to the limited penetration depths within biological tissues or reduce the high background fluorescent signals arising from components of the tissues. For example, exciting LTs with ultraviolet (UV) or visible (VIS) light, besides the limited penetration depths, may also induce phototoxicity within the biological tissues [4]. Further, if the luminescent materials emit within the UV or VIS wavelengths, their performance as thermometers could be highly influenced by the components of the biological tissues, which absorb and scatter quite efficiently within this range (Fig. 1a). At these regions, tissues present non-flat absorption coefficients (Fig. 1a), and also possible emissions which arise from their endogenous fluorophores such as nicotinamide adenine dinucleotide hydrate (NADH), flavin adenine dinucleotide (FAD) or collagen [5]. At larger wavelengths, the complexity is further increased as oxygenated and de-oxygenated hemoglobin, skin pigmentation or fat content, display variable extinction coefficients (Fig. 1a). In addition, from Fig. 1a, it could be noted that the scattering of biological tissues (due to the size, composition and morphology) becomes dominant over absorption for wavelengths shorter than 650 nm. All this makes tissue transmission in the visible strongly wavelength dependent. All these features, lead to incorrect and unreliable performance of LTs. On the other hand, exciting LTs with a low cost and low energy near infrared (NIR) wavelength, which induces no photobleaching, no autofluorescence and no phototoxicity upon biological tissues [4], assures deeper penetration depths, particularly when the emission of the materials is located within the so-called biological windows. Biological windows (BWs) are wavelength regions where the scattering and absorption from biological tissues are highly reduced [5]. Light scattering decreases exponentially as the wavelength increases from the UV to the NIR [6]. Due to these properties, LTs operating within these windows, are on high demand. However, also here special attention should be paid to the main absorption bands of water [7]. Water shows strong absorption bands at 980 nm, 1150, 1500 and 2000 nm, which define

Luminescent Nanothermometers Operating Within Biological Windows

223

Fig. 1 a Extinction coefficients of components of biological tissue within BWs. Data extracted from Ref. [6]. b LTs operating within BWs: quantum dots (Cd-based, carbon, and Ag2 S), nanodiamonds (ND), transition metal ions (Mn, Ni, Cr), and lanthanide doped materials (the lanthanide elements are: Er, Tm, Nd, Yb, Ho in the form of Ln3+ )

the ranges of BWs (Fig. 1a). Based on these bands, four different BWs regions are established (Fig. 1a) [1, 6]: • • • •

First biological window (I-BW) located within the 650–950 nm range Second biological window (II-BW) located within the 1000–1350 nm range Third biological window (III-BW) located within the 1400–1900 nm range Fourth biological window (IV-BW) centred at 2200 nm.

The I-BW, also known as the therapeutic window [8], is limited at shorter wavelengths from the absorption of hemoglobin, and at longer wavelengths from the absorption of water (Fig. 1a). Due to these, usually the penetration depths within this window, are limited at around 2 cm [6]. The II-BW displays a boost in terms of the signal-to-noise ratio (SNR) by a factor of 100 due to the reduced scattering of biological tissues compared to the I-BW [4]. The limitation of this window relies on the absorption bands of water. It is located within two absorption bands of water at 980 and 1150 nm (Fig. 1a), therefore a careful selection of the emission bands should be

224

A. Nexha et al.

considered before applying thermometers within this range. The III-BW, also known as short-wavelength infrared (SWIR) region [9, 10], despite being located within two of the strongest absorption bands of water, offers high penetration depths compared to the I-BW and II-BW due to the significant drop in the scattering of light by tissue [11]. Finally, the IV-BW, although still at its infancy due to the lack of excitation and detector setups [12], display similar values of light attenuation for skin as the II-BW. The discovery of new photodetectors (InSb in single or array photodiodes) and excitation laser sources (supercontinuum lasers operating in the range from 400 to 2500 nm) [11], may allow exploring the IV-BW as a new spectral domain for luminescent thermometry.

2 Luminescent Nanothermometers Within Biological Windows Luminescent thermometry explores the relationship between the temperature and the luminescent properties of a material [1, 2]. Temperature can induce changes in the intensity (E), band-shape (I ), bandwidth (γ ), spectral position (λ), lifetime (τ ), and polarization (⌃), of a luminescent material, which have led to the development of six different classes or techniques of luminescent thermometry, as explained elsewhere [1, 2]. Several types of LTs operate within BWs (Fig. 1b). These LTs include nanodiamonds (ND), fluorescent gold nanoclusters (Au), quantum dots (QDs) based on carbon (C QDs), cadmium (CdX QDs, where X = S, Se, Te), lead sulfide (PbS) and silver sulfide (Ag2 S QDs), transition metal (TM) based materials (manganese (Mn), nickel (Ni), cobalt (Co), chromium (Cr) ions), and lanthanide doped materials (Ln) such as erbium (Er), thulium (Tm), europium (Eu), neodymium (Nd), holmium (Ho) and ytterbium (Yb) ions. At the I-BW operate thermometers based on ND, Au, QDs (C QDs, and CdX QDs (X = Se, Te)), TM (Mn4+ , Co3+ and Cr3+ ) and Ln (Er3+ , Tm3+ , Nd3+ , Eu3+ and Ho3+ ) doped materials (Fig. 1b). At the II-BW are located thermometers of TM based on Ni2+ , Mn5+ and Cr4+ , QDs (PbS and Ag2 S), and Ln (Tm3+ , Nd3+ , Yb3+ ) doped materials (Fig. 1b). At the III-BW, only Ln doped materials based on Er3+ , Tm3+ and Ho3+ are developed as thermometers (Fig. 1b). Unfortunately, up to date, there are no reports on the preparation of LTs operating at the IV-BW. The luminescence for ND originates from the presence of point defect (tin, germanium, nitrogen or silicon vacancy centres) of their structure [13]. Among these defects, only the silicon based ones generate emissions located within the I-BW, attributed to their zero phonon lines [13]. Other defects display emissions only within the VIS range of the electromagnetic spectrum [13]. The advantages of these nanothermometers rely on their high thermal sensitivity and on the fact that the silicon vacancy are inner defects, hence, in theory, these thermometers are not affected by the surrounding environment [13]. However, their emissions seem to be limited only within the I-BW (Fig. 1b), and moreover, triggered by a green 532 nm laser [14, 15],

Luminescent Nanothermometers Operating Within Biological Windows

225

which is located at the VIS range, therefore limiting their penetration depths within the biological tissues. Au nanoclusters are quite appealing for applications in biomedicine due to their ultrasmall sizes (below 2 nm), chemical stability and good biocompatibility [16]. They are also highly sensitive to temperature changes, especially within the physiological range, as the intensities of their emissions drop around 70% of the initial value [16]. Despite these, their emissions are triggered by VIS excitation light, such as 580 nm, and their performance as nanothermometers is limited only to the E technique [16], which suffers from the local concentration of the emitters and the surrounding medium. C QDs display similar properties as Au nanoclusters. C QDs could sense temperature via the E technique, as well [17]. Major limitations of these thermometers are assigned to the emissions located at the border of the I-BW, and to their maximum quantum yields obtained when pumped with UV lights [17]. C QDs could be also excited with red light, such as 640 nm, and generate emission in the range from 645 to 730 nm [17]. Cd QDs display size-tunable optical properties, followed by narrow emissions and high quantum yields [18], which render these nanomaterials captivating for thermometry. In addition, issues related to their toxicity, photobleaching and emissions limited in wavelengths below 1000 nm, are already solved via the preparation of core@shell structures, in combination with PbS QDs [18, 19]. Still, limitations related to the influence of the environment, surfactants covering their surface, or ligands during functionalization on their luminescence properties should be addressed. Beside these, Cd QDs may introduce genetically-induced changes when applied in animal models [20], therefore a careful and in-depth analysis is still lacking. Ag2 S QDs overcome the limitations assigned to Cd QDs. These non-toxic QDs display peculiar properties combining emissions located at the II-BW while excited with NIR laser sources, with high quantum yields, and ability to be coated with surfactants or capping agents, with potential for biomedicine [21]. These materials offer the possibility to explore almost all classes of luminescent thermometry for sensing, such as E, I , γ , λ and τ [21]. Their emissions are highly dependent on the size of the QDs, therefore a precise control on the synthetic methodologies is required. For example, Ag2 S with sizes below 4 nm, have various emissions within the range from 500 to 1300 nm, whereas for sizes above 4 nm, they behave like a bulk unit and have a single emission within the 1000–1500 nm range [21]. TM (Ni2+ , Mn4+ /Mn5+ , Co3+ and Cr3+ /Cr4+ ) based luminescent particles are continuously growing as active materials for nanothermometry. They offer broad absorption bands, high resistance upon photobleaching, high chemical and photochemical stability, and broad emission bands located in the VIS, I-BW and II-BW (Fig. 1b). Their luminescence arises due to their incomplete filled d shells [22]. These properties are strongly affected by the crystal field of the ligands surrounding the ions. These emitters are not reported so far as thermometers within biological media, that is probably due to the fact that they are often excited with UV light sources [22]. Ln (Er3+ , Tm3+ , Eu3+ , Yb3+ , Nd3+ , and Ho3+ ) doped materials exhibit narrow emissions covering a wide range of the electromagnetic spectrum, including the

226

A. Nexha et al.

UV, VIS and all the BWs, as a function of the lanthanide ions selected and the transparency of the host material [1, 2]. Their emission lines could be generated with excitation by UV, VIS and NIR lights, with the latest one being highly applied for the development of LTs operating within the BWs. A limitation of these materials is their low absorption cross section as their 4f-4f transitions are partially forbidden due to the Laporte’s rule [1, 2]. Strategies to overcome this limitation have been already implemented including coupling with plasmonic structures, sensitization with organic dyes or conjugation with semiconductors [23, 24].

3 Performance of Luminescent Nanothermometers Despite the different nature of the luminescent materials, operating wavelengths, or the class of thermometry applied to extract the temperature, the performance of these materials as nanothermometers could be compared among them by using the thermal sensitivity or temperature resolution. Other parameters such as the spatial resolution, the temporal resolution, the repeatability and the reproducibility, could be also used to determine the performance of a nanothermometer, nevertheless, they are, barely or not reported at all. Definitions and methods on how to estimate these parameters are summarized elsewehere [1, 2]. Here, it will be shortly defined what the relative thermal sensitivity or the temperature resolution are, and how to estimate them following a defined class of nanothermometry. The thermal sensitivity could be reported in terms of absolute and relative values. The absolute sensitivity (Sabs ) is highly dependent on the experimental setup applied, and the characteristics of the luminescent material, therefore its application is limited only to nanothermometers of the same nature and tested under the same conditions [1, 2]. Sabs could be estimated by using the following expression [1, 2]: Sabs =

∂Δ ∂T

(1)

stands for the rate of change of the thermometric parameter (often labelled where ∂Δ ∂T as Δ) with temperature. Sabs is expressed in units of K−1 . The relative thermal sensitivity (Srel ), which states the maximum change of Δ for each temperature degree, is used as a figure of merit to compare the performance of nanothermometers regardless of their nature, experimental setup or operating wavelengths [1, 2]. Srel is defined as follows [1, 2]: | | 1 | ∂Δ || × 100% (2) Srel = || Δ ∂T | Srel is expressed in units of % change per degree of temperature change (% K−1 ). The temperature resolution (δT ) expresses the minimum change in temperature that a thermometer can detect [1, 2]. It is defined using the following equation:

Luminescent Nanothermometers Operating Within Biological Windows

δT =

1 δΔ Srel Δ

227

(3)

where δΔ is the uncertainty on the determination of Δ, therefore δT might be affected by the experimental detection setup used, the acquisition conditions applied, and signal-to-noise ratio (SNR) in the experiment [1, 2]. Within this chapter, we summarize and compare the performance of different types of LTs operating at the different BWs. We will use Srel as a figure of merit to compare the performance within one biological window or at different biological window ranges. We will shortly describe how the performance of a thermometer is extracted following the different classes of luminescent thermometry, and for thermometers displaying several changes on their luminescence properties when exposed to temperature, the most sensitive parameter will be pointed. In addition, when applied, strategies on how to improve the thermometric performance of a thermometer within a biological window, or which biological window to select for the highest performance of a particular thermometer, will be highlighted. Particularly, our attention will be focused on the performance of these thermometers within the physiological range of temperatures from 293 to 333 K, the range from which biomedicine benefits the most.

4 Thermometers Operating in the I-BW At the I-BW operate thermometers based on ND, Au, QDs (C QDs, and CdX QDs (X = Se, Te)), TM (Mn4+ , Co3+ and Cr3+ ) and Ln (Er3+ , Tm3+ , Nd3+ , Eu3+ , Ho3+ ) materials. In addition, rarely examples are reported where these different types of emitters are mixed, acting as nanothermometers within the I-BW. We will highlight the thermal sensing strategies for each of these materials and compare their performance in terms of Srel .

4.1 Nanodiamonds The emissions of ND arise from their structural vacancies, which could be nitrogen, tin, germanium or silicon based [13]. Nitrogen vacancy ND are the most popular ones, also applied as thermal sensors with high sensitivities. Nevertheless, they require simultaneous optical and microwave control, and due to inhomogeneous broadening, each emitter must be individually calibrated before used as thermometers [13]. In addition, nitrogen vacancy ND emit in the VIS range. On the other hand, silicon vacancy ND, manifest sharp and bright emissions, which are located in the I-BW (Fig. 1a) [14, 15]. Within the ND structure, the silicon atom being larger than the carbon atom in size, replaces two carbon atoms and is located within two

228

A. Nexha et al.

vacancies (Fig. 2a). This structure has an inversion symmetry resulting in low sensitivity to strains and allowing to achieve narrow emission bands. Silicon vacancy ND could be synthesized via a high-pressure high-temperature technique in the presence of fluorine containing organic molecules, which inhibit the production of nitrogen vacancies [14]. Particles in the size of around 50 nm with a cuboctahedral shape were produced. Upon excitation with a green 543 nm wavelength, a sharp emission peak located at around 730 nm, with a full width at half-maximum (FWHM) of approximately 4 nm is generated (Fig. 2b–d). The changes that temperature might induce in the properties of this emission peak were investigated within the range from 298 to 313 K. Temperature affects E, the intensity of the emission λ, wavelength γ is the half width of the emission of the 730 nm emission (Fig. 2b–d). This drop is due to quenching or increase in probability of non-radiative mechanisms [1].Temperature increment also induces red shift of the emission peak (Fig. 2c) [14]. This change could be attributed to thermally induced strains in the environment of the material. The width of the emission is also broadened while temperature increases. With the increase of the temperature, the width of the emission becomes broader due to the thermal vibration of the luminescent centre and its neighbouring atoms or molecules. Therefore, there are three classes of luminescent thermometry which could be used to extract thermal information from these ND. The expression for the evaluation of

Fig. 2 a Atomic structure of ND with silicon atom (Si) and two vacancies (V) in the diamond lattice of carbon (C) atoms. Luminescent nanothermometry with silicon vacancy ND using changes in: b E, c λ, and d γ , from 298 to 313 K. Data extracted from Refs. [14, 15]

Luminescent Nanothermometers Operating Within Biological Windows

229

Srel would obey Eq. 2, and Δ expresses the change of E, λ, and γ , as temperature increases from 298 to 313 K (Fig. 2b–d). Srel is expressed as follows: Srel Srel Srel

| | | 1 ∂E | | × 100% | =| E ∂T | | | | 1 ∂λ | | × 100% | =| λ ∂T | | | | 1 ∂γ | | × 100% = || γ ∂T |

(4) (5) (6)

∂γ ∂E ∂λ , ∂T and ∂T stand for the rate of the change of E, λ, and γ of the emission where ∂T with temperature T . The variation of these parameters with temperature obeys a linear function, therefore, following Eq. 2, the values of Srel would be determined from the ratio of the slope and the thermometric parameter. Within Refs. [14, 15], only the absolute values of the ) were determined. Substituting the values of the absolute sensitivisensitivities ( ∂Δ ∂T ties and the experimental data of the thermometric parameters in Eq. 2, the values of Srel would be 0.00175% K−1 , 0.77% K−1 and 1.20% K−1 , for λ, γ and E techniques, respectively. However, there are some limitations arising from these types of emitters. First, there are excited with VIS wavelengths, therefore their application in biomedicine is limited. Second, as pointed out also from Liu et al. [14] reproducibility of the data is quite complex and leads often to large deviation due to crystal strains. Crystal strains vary from ND to ND due to defects number, location of Si vacancies and shape of the ND. Based on these informations, ND are not able to precisely determine small temperature variations [14].

4.2 Fluorescent Gold Nanoclusters Au nanoclusters (Au NCs) could emit within the border of the I-BW. These fluorescent nanoclusters with diameters in the range of 1.6 nm, could be synthesized via one pot microwave-assisted strategy [16]. Under VIS excitations, such as 480 or 580 nm, these nanoclusters emit within the range from 650 to 710 nm [16, 25]. Compared to the 480 nm excitation, pumping the nanoclusters at 580 nm, generates sharper emission peaks [16, 25]. Temperature was investigated within the physiological range from 288 to 318 K, and the effect on the emission was investigated. Data suggests that temperature can induce changes in E, and also in τ (Fig. 3a–c) [16, 25], therefore two different classes of nanothermometry could be applied to extract thermal information. Lifetime thermometry could be explained briefly. Luminescence lifetime studies the time interval in which the emitted intensity decays down to 1/e of its maximum value

230

A. Nexha et al.

after a pulsed excitation [1]. The lifetime decay of the emission of these nanoclusters accelerates with the increase of the temperature (Fig. 3a). This is due to the increase of the probability of quenching mechanisms [1]. The decay curve could be fitted as a sum of three exponentials, with a short (τ1 around 24 ns), a medium (τ2 around 130 ns), and a long lifetime (600 ns < τ3 < 750 ns). Srel is evaluated using: Srel

| | | 1 ∂τ | | × 100% | =| τ ∂T |

(7)

∂τ stand for the rate of the change of τ of the emission with temperature. The where ∂T average τ exhibited a good linear relationship with the temperature (Fig. 3b). For E, the same principle as for ND could be followed. For Au nanoclusters, an overall drop of 67% in intensity was detected. Among these two methodologies to sense the temperature, the one based on the change of E display the highest Srel , with a value of around 5% K−1 at the lowest temperature under investigation. On the other hand, τ had a maximum value of around 1.05% K−1 . However, it should be admitted that for sensing temperature within biological tissues, τ would be more adequate as it is independent of the optical properties of the medium in which the nanoparticles are embedded, therefore not influenced from the biological tissues, as is the E technique [1]. Current limitations of these emitters lie on the fact that their emissions are within the border of the BWs, and they are excited with VIS wavelengths. In addition, also the quantum yield has been a drawback for these emitters. However, this can be solved by introducing them into nanofiber structures, which induced a 70-fold increase of the quantum yield [25].

4.3 Quantum Dots CdX (X = Se, Te) display the same conduct as the silicon vacancies ND when the temperature is changed. More specifically, CdSe QDs emit at 650 nm after being excited in the UV. When temperature increases, they exhibit a drop in E, a red shift of λ, and a broaden of γ of the 650 nm emission [26]. In all these examples, the variation of Δ was fitted to a linear function within the physiological range of temperatures. Nevertheless, their performance is relatively poor as they display Srel values in the range of 0.1–0.2% K−1 , regardless of the Δ used to determine their performance. Maestro et al. [27] investigated the effect of the excitation source on the thermometric performance of CdSe QDs. They compared the performance of these QDs by exciting them with VIS light (at 488 nm), and NIR light (at 800 nm). The thermometric parameters that were considered were E and λ, as the temperature was tuned within the physiological range. The excitation source had a significant impact on E, while barely changed λ [27]. For the 488 nm excitation, E of the emission suffers only a moderate reduction (25% reduction compared to its initial value), whereas for the 800 nm excitation was three times larger, reaching values up to 75% of reduction

Luminescent Nanothermometers Operating Within Biological Windows Fig. 3 a Luminescent nanothermometry with Au nanoclusters using changes in lifetime. Variation of b τ , and c E of the emission Au nanoclusters as a function of temperature. Data extracted from Refs. [16, 25]

231

232

A. Nexha et al.

Fig. 4 a Variation of E of CdSe at different temperatures, under 488 and 800 nm excitations. Data extracted from Ref. [27]. b λ of CdSe and CdTe as a function of temperature. Data extracted from Ref. [28]. c I technique with PbS/CdS/CdSe QDs. Data extracted from Ref. [29]. d Effect of temperature on the photoluminescence of C QDs. Data extracted from Ref. [17]

(Fig. 4a). These data would imply that Srel would be increased by 3-times for the 800 nm excited QDs compared to 488 nm excitation, as Srel is reversely proportional to Δ, as expressed in Eq. 2. Next, the same authors, using 800 nm excitation, compared the performance of two different derivatives of the CdX, that are the CdSe and CdTe [28]. After absorbing this energy, these materials emit at the edges of the I-BW: CdSe emits at 650 nm, and CdTe at 660 nm. The variation of λ as a function of temperature was monitored for both emitters. CdTe displayed almost a two fold improved thermometric performance compared to its CdSe analogue (Fig. 4b). Nevertheless, the thermometric performance of these materials is quite often based on E, which besides having a low Srel (below 0.5% K−1 ), suffers from the variation of the concentration of the emitters, what implies that a recalibration is required anytime they are used [1]. To overcome this limitation, the strategy of combining the emission arising from these QDs with other emissions from other QDs, such as PbS, was tested [29]. PbS/CdS/CdSe core/barrier/shell QDs were synthesized using wet chemical methodologies, followed by cationic exchange. The barrier was included to prevent the

Luminescent Nanothermometers Operating Within Biological Windows

233

recombination of PbS and CdSe. After pumping at 560 nm, two peaks located at around 670 nm and 910 nm, assigned to the CdSe shell and the PbS core, respectively, were observed. The thermometric performance was evaluated within the temperature range from 100 to 300 K. Band-shape nanothermometry (I ) was used to extract thermal knowledge of this class of nanothermometers. It can be used in fluorescent materials which possess at least two different emission bands, located within an energy gap (ΔE) in between 200 and 2000 cm−1 in a single emitter (called thermally coupled levels, TCLs), or with an ΔE higher than 2000 cm−1 arising from one or multiples emitters (called nonthermally coupled levels, NTCLs) [1]. The main advantage of this class of thermometers is its independency of signal losses and the possible fluctuations in the excitation intensity [1]. In the example of the PbS/CdS/CdSe QDs, Srel would be evaluated as follows: | | | 1 ∂I | | × 100% (8) Srel = || I ∂T | where I is the intensity ratio calculated from the integrated areas between the emis∂I is the rate of change of I with sions arising from the PbS and the CdSe QDs, and ∂T T . Within the temperature range from 100 to 300 K, the value of Srel for these dual emitting nanothermometers, was around 1.22% K−1 at the highest temperature under investigation [29]. This value is approximately 2.5 fold higher than that obtained in single based CdX (X = Se, Te) QDs, regardless of the class of nanothermometry used to extract their performance.

4.4 Carbon Quantum Dots C QDs as LTs hold great potential for applications in biomedicine as they are biocompatible, dispersible in water and exhibit low cytoxicity [17]. However, their emissions are most often encountered at the VIS wavelengths, such as in the blue or the green, and even in the yellow, after pumping with UV lights [30]. On the other hand, certain C QDs could be excited with red wavelengths around 640 nm, and generate emissions at around 645–730 nm (Fig. 4d) [17]. These C QDs are synthesized via a modified microwave-mediated one-step reaction with glutathione and formamide [17]. Particles with average sizes of 7.3 nm were obtained using this methodology. Exposing these C QDs to temperature, from 278 to 333 K, revealed a surprising trend: the luminescence was enhanced as temperature increased, as quite often observed in other emitting materials (Fig. 4d). Comparing E of the pluminescence at the initial temperature (278 K) with the final temperature (333 K) under investigation, revealed a 3.5-fold increase. In terms of Srel , following Eq. 4, a maximum value of 3.71% K−1 was obtained [17]. Besides this high value of Srel , it should be stated that the performance of C QDs is still based on the E technique, which suffers from the variation of concentration of thermal probes. In addition, a crucial issue with these emitters remains with the

234

A. Nexha et al.

understanding of the origin of their luminescence, although some first trials into this direction are underway [31].

4.5 Transition Metal Doped Materials TM doped particles have been recently employed as LTs. TM offer broad optical absorption bands, high resistance upon photobleaching, high chemical and photochemical stability, and broad emissions bands which are located within the VIS, I-BW and II-BW [22, 32]. These emissions mainly arise due to the incomplete filling of the d shells of these elements, therefore allowing for multiple d-d transitions to take place. These properties are strongly affected by the crystal field of the ligands [22, 32]. TM employed as nanothermometers within the I-BW involve the use of Cr3+ [33–35], Mn4+ [36–38], and Co3+ [39]. Due to the 3d 3 electronic configuration, Cr3+ ions exhibit emissions in the I-BW. After excitation with UV or blue wavelengths, Cr3+ doped materials usually exhibit two bands: a sharp band at around 700 nm, assigned as the R band, and a very broad band centred at 800 nm, assigned as the B band (Fig. 5a). For thermal sensing, two strategies could be applied: the variation of the intensity of the R versus the B band, or the comparison of the intensity of the splitting of the R band into two peaks (Fig. 5a) [33–35]. The thermometric performance of these materials can be modulated via a rational tailoring of the crystal field around the ions [34]. The crystal field is tuned based on the concentration of these ions: a high concentration, lead to a smaller crystal field which according to the Tanabe-Sugano energy level diagram [40], implies a smaller ΔE in between the transitions involved in thermal sensing. Therefore, to get a better thermal response, a small concentration of the ions should be selected, which could lead to a larger ΔE [34]. ΔE is directly proportional to Srel , as the thermometric performance was extracted from the Boltzmann law [34]. Based on this, and considering the ratio of intensities between the B and R bands, the smallest concentration of ions (0.1% of Cr3+ ) displayed the highest Srel compared to other two higher concentrations (0.5 and 3%) [34]. Cr3+ doped Bi2 Ga4 O9 particles were synthesized via solid state reaction. Upon 442 nm excitation, these particles generated the typical emission bands of Cr3+ : the R and B bands [33]. Within the temperature range of 77–450 K, two alternatives for thermal sensing were studied: (i) the intensity ratio between the two R peaks (699 and 710 nm), or (ii) the intensity ratio between the B and R bands. It was concluded that within this temperature range, the ratio between the two R peaks displays the highest Srel with a value of 2.74% K−1 at 80 K (Fig. 5b) [33]. However, within the physiological range of temperatures, the ratio between the B and R bands has almost a two times higher sensitivity than the ratio between the two R peaks (Fig. 5b). It was studied the possibility of using persistent luminescence for thermometry arising from these ions. In this vein, Cr3+ ions were embedded in the La2 MgHfO6 host and pumped with UV light, at 254 nm [35]. Here, the ratio among the two R peaks (although not as clearly distinguished as in the example of Bi2 Ga4 O9 particles)

Luminescent Nanothermometers Operating Within Biological Windows

235

Fig. 5 a Emission spectra of Cr3+ doped materials under excitation with blue light. Data extracted from Ref. [33]. b Srel of Cr3+ doped materials based on the intensity ratio between the peaks of the R line and the ratio between the R and B lines. Data extracted from Ref. [33]. c Emission spectra of Mn4+ doped materials under 400 nm excitation. Data extracted from Ref. [36]. d Srel of Mn4+ doped materials as a function of temperature. Data extracted from Refs. [36–38]. Effect of calcination on the sensitivity of Co3+ doped CaAl2 O4 materials analysed via: e E and f τ thermometry. Data extracted from Ref. [39]

236

A. Nexha et al.

was used to extract the performance within the temperature in the range from 295 to 373 K. Three luminescence parameters changed as the temperature increased: I of the two peaks, γ of the whole R band, and τ of the persistent emission [35]. Among these three, τ of the persistent emission displayed the highest Srel with a value of 1.7% K−1 at 463 K (the authors monitored the emission until it completely disappeared at above 463 K). However, this approach comes with several limitations: high temperatures (above 473 K) have to be applied to trigger the combinatory effect of excited Cr3+ ions and the defect traps, which lead to the persistent luminescence, and a precise time control is required in the interval when stopping the excitation and the moment the data were recorded [35]. On the other hand, it was analysed also the luminescence’s lifetime under 430 nm pulsed excitation at different temperatures, and concluded that this methodology, not only is more reliable than the persistent lifetime, but also more sensitive, reaching a maximum value of Srel of 1.93% K−1 at 573 K [35]. Related to Mn4+ ions applied as thermometers within the I-BW, their performance is based on a single emission located within the range from 700 to 730 nm (Fig. 5c), depending on the type of the host where the ions are embedded. The emission line is assigned to the electronic transition from the 2 E excited state back to the 4 A ground state. It is generated after pumping the ions with blue light [36–38]. Mn4+ ions are embedded into SrTiO3 [36], Mg2 TiO4 [38], and Li2 MgHfO4 [37], all synthesized using post-calcination processes. In all hosts, a low concentration of the Mn4+ ions (0.005–0.1 mol%) was added to properly modulate the crystal field strength. Two classes of luminescent nanothermometry were applied to extract the thermometric performance: E of the 700–730 nm band (for Li2 MgHfO4 host), or τ of this emission (for SrTiO3 and Mg2 TiO4 hosts). Among these hosts, SrTiO3 displayed the highest sensitivity with a value of 3.57% K−1 at 318 K, almost 3-times and 4-times higher compared to the one obtained in Li2 MgHfO4 and Mg2 TiO4 , respectively (Fig. 5d). Nanothermometers based on Co3+ ions have been also reported, although rarely, with emissions within the I-BW. A typical example involves embedding these ions in the CaAl2 O4 host [39]. Upon UV excitation at 266 nm, Co3+ ions emit at nearly 800 nm, assigned to their transition from the 3 T1 excited state to the 1 A1 ground state. Two classes of thermometry have been applied to determine the thermometric performance, mainly E and τ . First, the concentration of the transition metal ions was optimized to exhibit the highest intensity of the emission band, and the value of Srel by using E. Data suggested that 0.02% of Co3+ metal ions generate the brightest emissions and a slightly higher Srel compared to the other concentrations tested [39]. The CaAl2 O4 host is also know to switch from the hexagonal to the monoclinic crystalline phase when calcined at high temperatures from 1123 to 1273 K [39]. In this way, the crystal field around the ion would change, which would drastically affect the thermometric performance. Taking this into account, the authors investigated, via both classes of thermometry, the effect of the calcination temperature on the performance of these emitters within different ranges of temperatures [39]. Concerning E, for the hexagonal phase of the host, the maximum sensitivity was in the range of 2.70% K−1 at 275 K (Fig. 5e). After calcinating the host at different temperatures (1123, 1173, 1273, 1373 K), the new monoclinic host was exposed at the same temperature range

Luminescent Nanothermometers Operating Within Biological Windows

237

as the hexagonal one, and the thermometric performance was extracted following the same principle. Now, the sensitivity increased up to 3.77% K−1 at 340 K, 3.23% K−1 at 237 K, and 4.74% K−1 at 123 K, for 1173 K, 1273 K and 1373 K calcination temperatures, respectively (Fig. 5e). On the other hand, calcination at 1123 K, didn’t display any major improvement in the sensitivity (Fig. 5e). These experiments were attempted also for τ . Here, the sensitivity changed as follows: 0.93% K−1 at 423 K, 1.50% K−1 at 380 K, 1.34% K−1 at 256 K and 1.86% K−1 at 263 K, for particles obtained at 1123 K, 1173 K, 1273 K, 1373 K (Fig. 5f), respectively [39]. Surprisingly, the authors did not provide data on the sensitivity of the hexagonal host studied under τ . In addition, no explanation was provided on why the maximum sensitivities were archived at different temperatures for the different calcination temperatures [39]. Nevertheless, it could be concluded from data on Fig. 5e, f that the sensitivity of the E technique is approximately 2-times higher than that of the τ technique. TM nanothermometers might exhibit quite high sensitivities, regardless of the thermometry technique applied for sensing, probably because their electronic transitions are highly affected from the surrounding environment compared to other types of thermometers. Despite this, up to date, there are no reports on the applications of these materials into biomedicine, that is attributed to the fact that these materials are pumped with UV or VIS light [22, 32].

4.6 Lanthanide Doped Materials Different types of lanthanides emitting ions could operate as nanothermometers within the I-BW. Among them, Tm3+ and Nd3+ ions constitute the most popular ones, mainly due to their emission bands located in the range 700 nm–800 nm [1], respectively. Other lanthanides operating within this region are Er3+ and Eu3+ ions. A common feature of these ions, which separates them from all the luminescence materials introduced up to now, is their ability to absorb NIR light, such as 808 nm (absorbed by Tm3+ or Nd3+ ions), or 980 nm (absorbed by Yb3+ ), 1210 nm (absorbed by Tm3+ ) or 1530 nm (absorbed by Er3+ ) [1]. Therefore, when embedded within a host, the lanthanide ions could act simultaneously as sensitizer (able to absorb the excitation source), and as activators (able to trigger emissions) [1]. For example, the absorption cross section of Tm3+ at 808 nm is relatively low, thus these ions are most often encountered together with Yb3+ and excited at 980 nm in order to maximize their SNR [1]. On the other hand, Nd3+ ions have a large absorption cross section at 808 nm, therefore they could be found as the only lanthanide ion in a host matrix [1]. In general, the thermometric performance of lanthanide doped materials is defined on the difference of ΔE between the electronic transitions involved in temperature sensing. Thus, two different approaches could be identified in the literature: (i) nanothermometers operating via TCL, and (ii) nanothermometers operating via NTCL [1]. TCL best describes the performance of single emitting lanthanides, mostly co-doped with Yb3+ as a sensitizer. On the other hand, NTCL are applied for dual emitting centres arising from different lanthanide ions.

238

4.6.1

A. Nexha et al.

Single Emitting Lanthanide Doped Materials

Nd3+ Single Nd3+ doped materials operating in the I-BW are mainly based on three different TCLs assigned to electronic transitions responsible for the generation of the emission lines in the I-BW [1]. These TCLs are 4 F7/2 /4 F3/2 (TCL 1), 4 F7/2 /4 F5/2 (TCL 2) and 4 F5/2 /4 F3/2 (TCL 3), which are all responsible for the generation of emissions within this biological window [1]. These emissions could be triggered by pumping the materials with green or NIR light [1]. Shortly, Nd3+ absorbs the energy of the excitation source, and promote the electrons to the high-lying levels, from which non-radiatively the 4 F7/2 and 4 F5/2 levels are populated (Fig. 6a). From the 4 F7/2 level back to its ground state, the first emission line labelled as λem1 , located at around 750 nm, is generated. From the 4 F5/2 level two processes can take place: first, another emission line labelled as λem2 , located at around 800 nm is generated, and second, a non-radiative process can populate the 4 F3/2 level which upon relaxing to the ground state, can generate a third emission line labelled as λem3 at around 900 nm [1]. The 4 F3/2 level is composed also of two different Stark sublevels, which could generate emissions at around 900 nm, labelled as λem3 (1) and λem3 (2) (Fig. 6a), respectively, located at around 860 nm and 890 nm in fluoride hosts, for example [41]. When exposed to different temperatures, these luminescence materials obey the Boltzmann thermalization law, which strictly implies that ΔE is directly proportional to the sensitivity [1]. For example, Nd3+ doped in an oxyfluoride host was tested as luminescent thermometer based on the three TCLs [42]. The values of ΔE for these TCLs were as follows: 1910 cm−1 , 1020 cm−1 and 960 cm−1 , for TCL 1, TCL 2, TCL 3, respectively [42]. Among these three TCLs, clearly the TCL 1 has the highest ΔE, which implied also the highest sensitivity. Thus, in the temperature range from 303 to 623 K, the maximum sensitivities were 3.27% K−1 , 2.05% K−1 and 1.92% K−1 at the lowest temperature for TCL 1, TCL 2 and TCL 3 (Fig. 6b), respectively [42]. The same conclusion was also drawn from Nd3+ doped TiO2 nanoparticles which compared the performance of TCL 1 and TCL 3 [43]. Therefore, the sensitivity is limited to the value of ΔE, and the only way to improve the sensitivity of these materials is to have them operating within very low temperatures. For example, Nd3+ doped yttrium aluminum garnet (YAG), based on the TCL 2 scheme, display a maximum sensitivity around 2% K−1 at 200 K, while in the physiological range of temperatures it dropped down to 0.30% K−1 [44]. However, thermometers operating at cryogenic temperatures are not attractive for biomedical applications. Another option for nanothermometry with these single emitting Nd3+ materials is the one based on the 4 F3/2 → 4 I9/2 transition (Fig. 6a), which usually generates emissions within the 800 and 950 nm range, depending on the type of the host [1]. Here, the Stark sublevels of the 4 F3/2 electronic state are used to extract the thermometric performance. However, their ΔE are quite small, in the range from 50 to 270 cm−1 , therefore strictly speaking the Boltzmann law does not properly hold. Different types of Nd3+ doped hosts have been used for nanothermometry based on this transition, such as LiLuF4 [41], ALaP4 O12 (where A = Li, K, Na and Rb) [45],

Luminescent Nanothermometers Operating Within Biological Windows

239

Fig. 6 a Simplified energy diagram of Nd3+ doped materials pumped with green light, and selected emissions used for luminescence thermometry. b Srel of single emitting Nd3+ luminescent thermometers operating in the I-BW: in turquoise, yellow and brown, thermometers operating based on TCL 1 (TiO2 [43], YOF [42]), TCL 2 (YOF [42], and YAG [44]), and TCL 3 (YOF [42], Bi2 SiO5 [46], LiLuF4 [41], and LiLaP4 O12 [45]), respectively. c Luminescence of Tm3+ co-doped with Ho3+ in the monoclinic KLu(WO4 )2 host, excited at 808 nm. Data extracted from Ref. [47]. d Srel of Ho, Tm:KLu(WO4 )2 as a function of the concentration of Tm3+ and Ho3+ ions. Data extracted from Ref. [48]

or Bi2 SiO5 [46]. However, the values of Srel are in the range of 0.11–0.62% K−1 (Fig. 6b) in brown colour) [1]. Hence, for the I-BW region, the three TCL, especially TCL 1 seems to be a more promising strategy for sensitive thermometers based on Nd3+ compared to the Stark sublevels of the 4 F3/2 → 4 I9/2 transition. However, one should also consider the signal discriminability. These levels lie very close to each other, and thus the probability of overlapping bands is high, which might induce errors while recording the data [1].

Tm3+ The performance of single emitting Tm3+ doped nanothermometers within the I-BW is most often based on the emissions arising from TCL levels located at around 700 nm

240

A. Nexha et al.

and 800 nm, assigned to the 3 F2,3 → 3 H6 and 3 H4 → 3 H6 transitions, respectively [1]. A few examples report also on the applicability of the Stark sublevels from the different electronic levels involved in the 3 H4 → 3 H6 transitions for nanothermometry [1]. These emission lines could be generated by pumping the materials with NIR laser sources, including 980, 1210, and 1319 nm [1]. The 1210 and 1319 nm excitation sources are not often applied, however, they might hold great interest for biomedical applications as both excitations have a good penetration depth in the biological tissues, higher when compared to the traditional 980 nm excitation [1]. Tm3+ doped yttrium orthoaluminate perovskite (YAP) was excited at 1210 nm, and the intensity ratio between the 700 and the 800 nm bands was used to extract the performance of the thermometer [49]. The experimental data were fitted to a Boltzmann law as these two emissions arise from TCL levels. On the other hand, Tm3+ doped NaNbO3 particles were excited at 1319 nm, and their thermometric performance was based on the different emissions arising from the different Stark sublevels of the electronic levels involved in the 3 H4 → 3 H6 transition [50]. Curiously, here the experimental data were fitted to a simple linear function. Among these two thermometers, the one excited at 1210 nm, achieved a maximum sensitivity of 2.64% K−1 at 324 K [49], compared to the 0.80% K−1 at 303 K of the one excited at 1319 nm [50]. It is important to highlight here the difference in the signal discriminability: while for the material pumped at 1319 nm the two emission peaks were clearly overlapped and difficult to distinguish, the one excited at 1210 nm, allowed easily record the two emission bands. However, a drawback is related to their low SNR, as can be observed for example for the 700 nm emission band while excited with the 1210 nm laser [49]. That is why Tm3+ materials are usually co-doped with Yb3+ , which can absorb at 980 nm, and via energy transfer (ET) populates the electronic levels of Tm3+ ion [1]. Concerning the combination of Tm3+ with Yb3+ and their implication in luminescent thermometry, the number of articles is exhaustive. We will summarize some of the main results. There are different classes of luminescent thermometry applied to extract thermal knowledge from Tm3+ and Yb3+ co-doped materials, including I , γ , and τ . For the I technique, obviously the intensity ratio between the 700 and 800 nm emissions, and rarely also the Stark sublevels of the 800 nm band, have been used [1]. Tm3+ and Yb3+ ions are usually found in hosts such as vanadates [51, 52], phosphates [53, 54], tungstates [55], silicates [56], and fluorides [57, 58]. The maximum sensitivities of these materials range from 2 to 3% K−1 , achieved within the physiological range of temperatures [1]. Among these hosts, phosphates appear to be superior. In addition, comparing the data on the sensitivities, phosphate hosts composed of rare earth materials, such as LaPO4 (3% K−1 ), and YPO4 (2.33% K−1 ) [53], appear to act as more sensitive thermometers when compared to phosphate hosts composed of post-transition metals, such as BiPO4 (2.14% K−1 ) [54]. A critical point for the application of these nanothermometers is the low SNR of the 700 nm emission. That is why plenty of studies are devoted to the decrease of the non-radiative processes within these materials and the improvement of the quantum yield by coating the materials with an extra shell layer of silica, found in materials like GdVO4 @SiO2

Luminescent Nanothermometers Operating Within Biological Windows

241

[51], NaYF4 @NaYF4 @SiO2 [58], and Bi2 SiO5 @SiO2 [56]. There have been also attempts to use these types of nanothermometers for high temperatures up to 1000 K, which although out of the interest for biomedicine, could be appealing for industrial processes [52]. However, the maximum Srel was achieved at room temperature with a value of 2.9% K−1 , while at 1000 K, this value dropped significantly down to 0.25% K−1 [52]. Related to the γ technique, Tm3+ , Yb3+ were found in YOF nanoparticles, and the bandwidth of the highest band at 800 nm was investigated as a function of the temperature from 12 to 300 K [59]. The experimental data were fitted to a simple exponential equation and the maximum Srel was 0.80% K−1 at 300 K [59]. The lifetime technique was applied to Tm3+ , Yb3+ co-doped hexagonal NaYF4 nanocrystals [60]. The lifetime of the 800 nm emission was studied as a function of temperature within the physiological range. The sensitivity was studied as a function of the concentration of Tm3+ and the excitation pulse width. The maximum value of Srel was 0.92% K−1 for the 1% concentration of the ions and 200 μs pulse width excitation [60]. In summary, among all the thermometric classes for single Tm3+ emitting materials operating in the I-BW, the I technique results the most effective one. However, it should be taken into account that co-doping Tm3+ , Yb3+ into hosts might induce a strong heating effect in tissues as the 980 nm excitation wavelength is efficiently absorbed by water [1].

Eu3+ Eu3+ doped materials can act as LTs operating in the I-BW due to their emission line in the deep red region, around 710 nm, which is assigned to the 5 D0 → 7 F4 transition. Nevertheless, reports on this type of nanothermometer are scarce in the literature. Only Eu3+ doped Y2 O3 particles have been analysed up to date for this purpose [61]. Despite this, these nanothermometers come with a peculiar feature: they are able to act as primary thermometers, that are thermometers whose calibration curve could be predicted based on the knowledge of a thermodynamic law [61]. In this way, the thermometric performance of these thermometers is independent of the medium in which they operate, and there is no need for calibration. Their thermometric performance was based on the I technique of the deep red emission, when resonantly excited via the 7 F2 and 7 F1 states [61]. Their performance was evaluated within the range of temperatures from 180 to 280 K. Excitation via the 7 F1 state generated a slightly higher sensitivity with a value of 1.70% K−1 at the lowest temperature under investigation [61], which is perhaps not appealing for biomedicine. But the most important fact is the ability to act as primary thermometers, which if applied in other types of thermometers with higher sensitivities in the physiological range of temperatures, could be highly captivating.

242

A. Nexha et al.

Er3+ Single Er3+ doped materials are not often explored as LTs within the I-BW. These thermometers operate on the principle of the I technique using the Stark sublevels of the red emission (at around 660 nm) [62, 63], and the emissions located at around 800 and 850 nm [64], and NTCL located in the red region and at around 800 nm [65]. These emission lines could be generated via excitation with VIS wavelength, such as the green at 532 nm [64], or via excitation at NIR wavelengths, such as 800 nm [62, 63], and 1530 nm [65]. In terms of sensitivity, that of the materials based on the I technique of the emissions located at 800 and 850 nm, is the highest one, approximately 2-times higher when compared to the others, reaching a maximum Srel of 1.39% K−1 at room temperature [1]. As expected, the performance of the thermometers based on the Stark sublevels of the red emission, is relatively poor as their ΔE is very small (in between 200 and 300 cm−1 ), which could lead also to poor signal discriminability.

4.6.2

Dual Emitting Lanthanide Doped Materials

To overcome limitations in the signal discriminability and to improve SNR and the thermometric performance, Ln doped thermometers are better prepared with dual emitting centres, i.e. two lanthanide ions act as activators and their emissions are used for thermal sensing [1]. In these systems, one of the activators has also the role of the sensitizer, or another lanthanide such as Yb3+ is added to fulfil this role. For the example of Tm3+ doped materials, they are mostly co-doped with Ho3+ , and Er3+ ions, in the absence or presence of Yb3+ , depending on the selection of the excitation source. When Tm3+ is encountered together with Ho3+ , 808 nm is used as the excitation wavelength, which is absorbed by Tm3+ ions and then due to ET processes, the electronic levels of Ho3+ are populated, which after relaxing radiatively to an intermediate or the ground state, generate emissions in the I-BW and other BWs [9, 47, 48]. When together with these ions also Yb3+ is included, the overall system is excited with a 980 nm laser source, and Yb3+ , which act as sensitizer is absorbing this energy, and via ET processes is populating the electronic levels of Tm3+ and Ho3+ [66, 67]. It should be mentioned here that compared to single Tm3+ , or to Tm3+ , Yb3+ co-doped materials, whose performance is based on poor SNR of one of the emissions of Tm3+ [51], or by deconvoluting the different emissions via Lorentzian function [50], the intensities of the emissions involved in Ho3+ , Tm3+ co-doped materials, are nearly balanced in intensity (Fig. 6c) [47, 48]. Clearly, in this type of nanothermometers, NTCL govern the thermometric performance. Between excitations at 808 and 980 nm, the first excitation wavelength not only is the most beneficial one as it does not suffer from absorption by water molecules contained in the biological tissues, but also significantly improve the performance of the thermometers, especially within the physiological range of temperatures. Tm3+ was encountered together with Ho3+ in monoclinic tungstate hosts such as KLu(WO4 )2 and their thermometric performance within the I-BW was based on

Luminescent Nanothermometers Operating Within Biological Windows

243

I between the 700 nm emission of Tm3+ and the 755 nm emission of Ho3+ [47, 48]. The experimental data were fitted to a simple order exponential function. Two conclusions could be drawn from the performance of these thermometers. First, the concentration of the emitting lanthanide ions influenced drastically the performance as they could affect the ET (and back energy transfer, BET) and non-radiative processes taking place within them. Therefore, among different concentrations tested, 1% Ho3+ and 15% Tm3+ displayed the highest sensitivity (Fig. 6d) [48]. Second, the morphological characteristics of the host can affect the sensitivity as well. Monoclinic KLu(WO4 )2 host are quite often synthesized via the modified sol–gel Pechini (P) methodology, which leads to big agglomerates with average sizes of 2 μm [48]. Solvothermal methodologies (conventional autoclave (CA) and microwaveassisted (MW)) produced particles with average sizes below 20 nm and no defined morphology [68]. Finally, applying thermal decomposition (TD) methodology, particles with well-defined rod shape and sizes around 1 μm were obtained [47]. Using the same concentration of Ho3+ and Tm3+ within these hosts, the results from the thermometric performance led to the conclusion that big agglomerated the particles from P method had the highest sensitivity [47]. Although in depth studies are required to draw more systematic conclusions, the size and shape (and all other parameters related to those, such as surface area, surface-to-volume ratio and possible existence of defects) play a pivotal role on the performance of these thermometers. Concerning the combination of Tm3+ with Er3+ , they are encountered always in the presence of Yb3+ which acts as sensitizer and via multiple ET processes can populate the levels of the two activators [1]. The thermometric performance of these materials was based on I between the 700 nm emission of Tm3+ and the red emission of Er3+ . The experimental data were fitted to polynomial functions [69, 70]. Nevertheless, there are not many reports on these type of nanothermometers due to the following reasons: (i) their performance is relatively poor compared to the Tm3+ and Ho3+ codoped materials, their maximum Srel were achieved at high temperatures, for example for Tm3+ , Er3+ , Yb3+ :NaLuF4 at 500 K with a value of 0.76% K−1 [69], or for Tm3+ , Er3+ , Yb3+ :LuF3 at 363 K with a value of 0.95% K−1 [70]; (ii) they were pumped with a 980 nm laser, and (iii) the Er3+ emission in the red, not only is located within the border of the I-BW, but also exhibited a poor quantum yield when compared to the green emission of Er3+ more traditionally used for luminescence thermometry purposes. Concerning Nd3+ based nanothermometers operating on the principle of the dual emitting centres, these ions are usually combined with Yb3+ (with dual role of sensitizers and activator for both ions), Eu3+ (as activator) and Er3+ (as activator). Nd3+ , Yb3+ co-doped materials were excited with a 977 nm laser, which is absorbed by Yb3+ and then via the so-called phonon assisted energy transfer (PAET) process [72], the electronic levels of Nd3+ are populated. This process is successful if the maximum phonon energy of a host is suitable for covering the ΔE between the levels of these ions. Thus, the emissions of Nd3+ ion, including the three TCL levels with emissions in the I-BW, and also the emission of Yb3+ generated by the radiative relaxation back to the ground state, can be generated. As a note, the common excitation at 980 nm, absorbed by Yb3+ ions, generated an emission which is usually located

244

A. Nexha et al.

within 1000–1100 nm. However, in hosts which are able to sustain PAET processes, these emissions are encountered within 900–970 nm [1]. Each combination of the emissions generated by the three TCL levels with the emission of Yb3+ (at 920 nm) has been tested for nanothermometry purposes. Among all the combinations, the emission arising from TCL 2 and the 920 nm emission of Yb3+ displayed the highest sensitivity, which however is still smaller than that of a single emitting Nd3+ based thermometer using the emission generated from the TCL 1 level [72]. When combined with Eu3+ ions, the thermometric performance of these thermometers is based on I of the deep red emission of Eu3+ and the emission generated from the TCL 2 level of Nd3+ , which are simultaneously excited with green light [73, 74]. The other two emissions arising from the TCLs of Nd3+ have poor SNR, therefore they are not reliable for thermal sensing. In these mixtures, as the temperature is increased, the intensity of the Eu3+ did not almost change, while the intensity of the emission of Nd3+ increased [73, 74]. Thus, the emission of the Eu3+ ion could act as a reference. The thermometric performance of these thermometers is better than the one obtained when using the emission arising from the TCL 3 of Nd3+ , but still lower than that obtained when using the emission from the TCL 1 of single emitting Nd3+ ions [73, 74]. Recently, a novel strategy has been applied to boost the thermal response of dual emitting lanthanide LTs. Within a one-dimensional sandwiched nanostructure were integrated dual emitting centres with different responses to temperature: (i) Nd3+ emitting at 803 nm, whose emission was thermally enhanced, and (ii) Er3+ emitting at 654 nm, whose emission was thermally quenched [71]. Yb3+ could absorb the energy of the 980 nm excitation source and populate the energy levels of the two emitting ions. The design was as follows: a structure with NaYF4 co-doped with Er3+ and Yb3+ , an inactive intermediate structure of NaYF4 , and another structure of NaYF4 :Yb3+ , Nd3+ (Fig. 7a). This structure can guarantee the ligand-capped surface for thermal enhancement in the Yb3+ -Nd3+ section and could potentially block cross relaxation (CR) mechanisms (CR are processes that lead to quenching of luminescence). These mechanisms are not efficiently blocked in simple triple doped or core@shell structures [71]. In addition, also a large concentration of the sensitizer enhanced the emission, with Yb3+ concentrations as large as 60% [71]. As the sandwiched structure was exposed to elevated temperatures (from 298 to 413 K), the anti-Stokes emission of Nd3+ (803 nm) and the multiphoton upconversion of Er3+ evolved in opposite directions of enhancement and quenching (Fig. 7b), respectively. The intensity ratio between the Nd3+ emission and the Er3+ emissions were used for thermal sensing. The concentration of these ions within the sandwiched structure was optimized based on the values of the sensitivities obtained. From the data, the optimal concentrations were selected as follows NaYF4 , 4% Er3+ , 40% Yb3+ /NaYF4 /NaYF4 :60% Yb3+ ,4% Nd3+ [71]. This structure generated the highest sensitivity value within the I-BW with a value of 9.6% K−1 at room temperature (Fig. 7c). Nevertheless, being excited at 980 nm, might imply difficulties for applications in biomedicine. Similar designs of structures with the ability to absorb other wavelengths which do not suffer from self-absorption, are highly desirable.

Luminescent Nanothermometers Operating Within Biological Windows Fig. 7 a Sandwiched structure with dual emitting centres whose emissions are thermally enhanced and thermally quenched when exposed to temperature. b Luminescence within the sandwiched structure as a function of temperature. Data extracted from Ref. [71]. c Srel of the sandwiched structure with different concentrations of the lanthanide ions, in the following order Yb3+ , Nd3+ @ Yb3+ , Er3+ . Data extracted from Ref. [71]

245

246

A. Nexha et al.

4.7 Mixed Luminescent Materials The different types of fluorescent materials presented within this biological window, could be also combined within the same structure, and tested as potential nanothermometers. This strategy has been applied for TM combined with Ln [1], with peculiar luminescence properties. Within these mixtures, the luminescence of Ln ions almost did not change with temperature, while the luminescence of TM is highly affected and continuously quenched due to the sensitive profile of the d-d transitions [1, 22]. Therefore, within these mixtures, the luminescence of Ln is acting as a reference signal. A typical example of the mixture of Ln/TM materials involves the use of the emission arising from TCL1 of Nd3+ , and the emissions arising from different TM such as Cr3+ at 690 nm [44], Mn4+ at 675 nm [75], Ti4+ at 669 nm [76], and Fe3+ at 720 nm [77]. These emission lines were generated by exciting the mixtures at 590 nm (for Nd3+ and Cr3+ co-doped materials), 355 nm (for Nd3+ and Mn4+ codoped materials), 266 nm (for Nd3+ , Ti4+ and Fe3+ co-doped materials), or two distinct excitation wavelengths (266 nm for Fe3+ and 808 nm for Nd3+ ). The intensity ratio between the TM emissions and the emission arising from Nd3+ ions, was used to calculate the sensitivity. Among these thermometers, the ones based on Nd3+ , Ti4+ and Cr3+ co-doped materials display sensitivity values of 3.5% K−1 , while the ones co-doped with Fe3+ and Mn4+ reach maximum values of 0.5% K−1 . Although these types of thermometers could improve substantially the sensitivity compared to those based on single Nd3+ emitters, it should be noted that the maximum Srel were recorded either at high temperatures (at around 475 K for Ti4+ and Cr3+ ) or at cryogenic temperatures (at around 200 K for Fe3+ and Mn4+ ). In addition, a crucial disadvantage of these thermometers is that they have to be pumped with UV and VIS light, and therefore are not suitable for biomedical applications. As a summary for all types of nanothermometers operating in the I-BW, combining thermally quenched and thermally enhanced emissions arising from Ln materials appears to be the most promising strategy for increasing the thermometric performance (Fig. 8). TM (either single emitting or in combination with Ln) also hold a great potential when looking for high Srel values (Fig. 8). Regardless of these factors, it should be admitted that in both examples, the excitation sources are a crucial limitation: lanthanides with thermally quenched and enhanced emissions are usually pumped at 980 nm, which matches with the absorption band of water, while TM (and their additional combinations) often are pumped with UV lights, which display phototoxicity and restricted penetration depths within biological tissues. In view of these aspects, novel design strategies that could take profit of other excitation sources, which do not match the absorption bands of water and can penetrate deeper in tissues, should be developed. Materials which are able to absorb these types of excitations (such as 808, 1200 or 1500 nm) are already available, for example considering Nd3+ , Tm3+ or Er3+ doped and co-doped materials [1].

Luminescent Nanothermometers Operating Within Biological Windows 10

247

71

Srel, % K-1

8

6 16

4

2

76

17 39

42

36

61 14

29

0 100

150

200

250

300

53 47

350

33

400

450

500

T, K Fig. 8 Srel of all types of nanothermometers operating within the I-BW: ND [14] (in black); Au [16] (in gold); QDs based on CdX [29], and C [17] (in red); TM based on Mn4+ [36], Co3+ [39], and Cr3+ [33], (in brown); Ln based on single Nd3+ [42], single Tm3+ [53], single Eu3+ [61], codoped Ho3+ , Tm3+ [47], and Er3+ , Nd3+ emitters in a sandwich structure [71] (in turquoise); and the mixture of Ln/TM [76], (in violet). Numbers portray the references from which the data were extracted

5 Thermometers Operating in the Second Biological Window At the II-BW operate thermometers based on TM (Ni2+ , Mn5+ and Cr4+ ), QDs (PbS and Ag2 S), and Ln (Tm3+ , Nd3+ , Yb3+ ) materials (Fig. 1b). Compared to the I-BW, here the strategy of mixing two different types of emitters, is rarely reported. A few examples explored the possibility of mixing QDs with Ln materials. Within this biological window, we will attempt to underline the most promising systems for nanothermometry, and what strategies to follow to boost their sensitivities.

5.1 Quantum Dots QDs could be synthesized via several synthetic methodologies, ranging from thermal decomposition, hot-injection and co-precipitation, which lead to different sizes and shapes. The optical properties of these materials are highly tunable as a function of their morphological characteristics. This property, combined with the stability against photobleaching and the easy functionalization of their surfaces to provide biocompatibility, render these types of materials appealing for luminescent thermometry and further, for biomedical applications [32].

248

A. Nexha et al.

Within II-BW, two types of QDs are reported as LTs, that are PbS (operating in the range of 1060–1270 nm) and Ag2 S (operating in the range of 1200–1235 nm). Temperature can affect almost all luminescence features of these QDs, for example it affects E, I , γ , λ, and τ for Ag2 S QDs within the physiological range [21, 32]. Concerning PbS QDs, only E is affected. PbS QDs are often prepared in a core@shell@shell type of structure on the form of PbS@CdS@ZnS to ensure complete isolation from environment of the core, and to reduce the trap state with the goal of generating a high SNR of their emission located at 1270 nm [78, 79]. This emission line is assigned to the first exciton transition of PbS QDs [78, 79]. The intensity of this emission drops around 50% of its initial value when exposed to temperatures within the physiological range. The calculated Srel is in the range of 1% K−1 [78]. Strategies used to boost the sensitivity of these QDs involve combining them with different emitters within the same biological window. PbS QDs were combined with Nd3+ doped materials, which emit at 1060 nm (assigned to the 4 F3/2 → 4 I11/2 transition) [79]. Both emissions have been triggered by excitation at 808 nm. When exposed to temperature, opposite trends on the intensity of the emissions are observed: the intensity of the emission of Nd3+ doped materials were not significantly affected by temperature, while the intensity of the PbS emissions droped continuously. The performance of this mixed structure was evaluated by I among the emissions of Nd3+ and PbS. The calculated value of Srel was around 2.5% K−1 at room temperature [79]. Therefore, mixing these two different types of emitters, increased approximately 2.5-times the thermometric performance compared to the use only of PbS emitting thermometers, constituting a successful strategy towards the development of more sensitive thermometers. The luminescence properties of Ag2 S QDs, focused on the 1200 nm line, change substantially with temperature, which allowed multithermal readouts. The emission line is generated after exciting at 808 nm and is assigned to the intrinsic bandgap of the material [21, 32]. Among all these luminescent thermometry classes (E, I , γ , λ, and τ ), the one based on the overall drop of the intensity of the emission, resulted the most sensitive one (Fig. 9). It achieved a maximum Srel of 5% K−1 at the room temperature [83]. Nevertheless, it should be admitted that this sensing approach is not reliable and in addition, the fluorescent material itself has a low quantum yield. Therefore, in this vein, two strategies should be highlighted. The first one relies on the preparation of a novel type of Ag2 S QDs, which are called Ag2 S superdots. These superdots are made of Ag@Ag2 S cores covered by an AgCl shell. The synthetic approach behind the preparation of these superdots relies on photochemistry using a femtosecond laser irradiation [84]. A peculiar feature of these superdots is their high quantum yield, reaching a maximum value of around 80%, approximately 80times higher when compared to other Ag2 S structures [21]. This boost in quantum yield was assigned to the covering shell, which, according to the authors, drastically reduced the structural defects. In terms of sensitivity, the maximum value achieved was 3.9% K−1 at room temperature, determined by the E technique [84]. The second strategy that is worth highlighting here is the one that explores the applicability of the τ technique for these QDs. The lifetime decreased to a 70% of its initial value as

Luminescent Nanothermometers Operating Within Biological Windows

249

the temperature increased [85, 86]. In terms of sensitivity, it was around 4% K−1 at room temperature, comparable to the most sensitive thermometry classes for these QDs (Fig. 9). Despite this, the main benefit of this class of thermometry relies on the fact that is not affected from the different medium or biological tissues where the emitting materials are embedded. In this sense, these Ag2 S QDs were coated with different ligands to render them soluble in apolar or polar solvents [85]. The sensitivity was affected by this surface functionalization, which proved the robustness and reliability of lifetime nanothermometers (where sensitivity is independent of their surface properties). It is worth underlying for QDs operating in the II-BW that they hold great promises for biomedical applications as they are pumped at 808 nm, and can display high sensitivities at room temperature, using multiple thermal sensing techniques, among them, the lifetime, which is unsensitive to the surroundings of the luminescent thermometer.

Fig. 9 a Srel of Ag2 S QDs thermometers operating within the II-BW based on: E, I , λ and τ . Numbers portray the references from which the data were extracted. b Changes of the luminescence of Mn5+ doped materials as a function of temperature: E, I , and λ. Data extracted from Ref. [80]. c Effect of the grain size on the thermometric performance of Ni2+ doped SrTiO3 thermometers. Data extracted from Ref. [81]. d Effect of the host material (ATiO3 , where A = Sr, Ca, Ba, Mg) on Srel of Ni2+ doped thermometers. Data extracted from Ref. [82]

250

A. Nexha et al.

5.2 Transition Metal Doped Materials Main TM operating within the II-BW are Mn5+ [80], Ni2+ [81, 82], and Cr4+ [87]. As pointed out also for the thermometers operating in the I-BW, also here there are no reports on the application of these materials into biological tissues. However, compared to the results reported in the I-BW, within the II-BW, the emission of these TM is generated while pumping with NIR lasers. For example, the emission bands of Mn5+ at 1190 and 1230 nm, are triggered by excitation at 808 nm [80]. On the other hand, Cr4+ ions have multiple absorption bands from 600 to 800 nm, which allows to excite these ions within this range. Upon 730 nm excitation, an emission band at 1230 nm was generated for Cr4+ doped materials [87]. Only Ni2+ ions have to be excited with UV light, such as 375 nm, to trigger the emissions in the II-BW (at 1245 nm), and even in the III-BW, depending on the type of host in which this ion was embedded [81, 82]. Going into more details about Mn5+ doped materials and their application as luminescent thermometers within this region, they were embedded into irregular barium phosphate Ba3 (PO4 )2 nanoparticles via an aerosol flame approach combined with calcination at 1073 K for 2 h [80]. The emission bands of these ions, when exposed to temperature display multiple changes: E of the main emission band at 1190 nm, I of the 1190 and 1230 nm bands, and λ of the 1190 nm band [80]. About λ, also the other emission band displayed changes, however the authors neglected it as it had a low SNR (Fig. 9b). In fact, also I was based on this 1230 nm band, therefore strictly speaking, this low signal might affect the accuracy of the measurements. Among all these parameters, I was the most sensitive one, however the value of sensitivity was just 0.43% K−1 at room temperature [80]. Besides this, these ions are among the few transition metals whose emissions can be triggered within the NIR wavelengths, therefore they might hold great promises for biomedical applications. Ni2+ ions were doped into ATiO3 structures (where A = Sr, Ca, Ba, or Mg). These hosts were all synthesized by a P method, followed by calcination at different temperatures within the range from 1123 to 1373 K over a period of 8 h [81, 82]. The first efforts were based on the Ni2+ doped SrTiO3 structures, where the role of the grain size of the particles in the thermometric performance, was investigated. Tuning the calcination temperature, nanoparticles with average sizes from 31 to 46 nm were obtained [81]. Taking profit from the decrease of E of the 1245 nm emission, per each particle size, the values of Srel were extracted. In general, with the increase of the size of the particles, a better thermometric performance was achieved (Fig. 9c). Maximum sensitivities (around 4% K−1 ) were obtained for temperatures higher than 373 K. It was also explored the possibility of using λ to detect temperature changes, however, this approach displayed a poor performance with Srel of around 0.25% K−1 [81]. The effect of the different cations on the host structure, i.e. ATiO3 with A = Sr, Ca, Ba, or Mg, was elucidated. It is important to note that different types of cations change the crystalline phase of the final product. Thus, when A = Sr and Ca, the final products crystallized in the cubic system, with A = Ca the final product crystallized in the orthorhombic system, and with A = Mg it crystallized in the trigonal system

Luminescent Nanothermometers Operating Within Biological Windows

251

[82]. As a consequence, the wavelength of the emission band of Ni2+ ions will change. For A = Sr and Ca, the emissions were located within the II-BW, at 1210 nm and 1320 nm, respectively. On the contrary, for A = Ba and Mg, the emissions were located within the III-BW, at 1540 nm and 1700 nm, respectively. For simplicity, we are comparing here the performance of these thermometers within this section, although they are assigned to different BWs. To determine their performance, λ was taken into account [82]. Overall, SrTiO3 had the lowest performance within all the temperature range considered 123–483 K (Fig. 9d). On the other extreme, BaTiO3 exhibited the highest sensitivity at around 5% K−1 , however recorded at 480 K. If we focus within the physiological range of temperatures, CaTiO3 reached a sensitivity around 3% K−1 , while MgTiO3 of 2.3% K−1 , and BaTiO3 around 1% K−1 [82]. Cr4+ doped materials, although not often explored, might be appealing for luminescent thermometry, as they could be pumped in two different wavelengths within the NIR range, that are within the 600–800 nm range, and at 1050 nm. Among these, the 600–800 nm range has a higher absorption coefficient, therefore by pumping at these wavelengths would generate a more efficient luminescence. These ions were embedded in Ca2 Al2 SiO7 nanoparticles following a P methodology [87]. The effect of the temperature in the 1230 nm emission band of Cr4+ ions resulted in changes in E, λ, and τ of this emission [87]. Among these methodologies, λ was the most sensitive one, achieving a sensitivity of 0.61% K−1 , very close to that obtained by the τ technique (0.59% K−1 ) [87]. If we compare the thermometric performance of these thermometers, with other TM operating in the II-BW, they are less sensitive, therefore, new strategies, inspired also in the fact that they could be pumped in NIR (including I-BW and II-BW), should be investigated.

5.3 Lanthanide Doped Materials Yb3+ and Nd3+ ions are the most used lanthanide ions within the II-BW. Their popularity relies on the ability to absorb light within the I-BW and downshift this energy into emissions in the II-BW. The performance of Yb3+ doped thermometers is based on the emission of this ion at around 1000 nm, either single doped or combined with other lanthanide emitting ions, such as Tm3+ or even Nd3+ . Regarding the performance of Nd3+ , these ions have two emission bands in the II-BW, at around 1050 and 1300 nm (Fig. 1b). In addition, these emissions could be also combined with those arising from other lanthanide emitting ions, for example Ho3+ ions.

5.3.1

Yb3+

Yb3+ doped material acting as thermometers in the II-BW operate based on single emitting Yb3+ materials or dual emitting materials combined with other lanthanide ions [1]. Related to the single emitting Yb3+ materials, their performance is very poor as it is based on Stark sublevels of the emission line of Yb3+ in the II-BW, whose

252

A. Nexha et al.

ΔE is very small. In addition, to trigger the emission of Yb3+ in the II-BW, UV excitation sources are applied. For example, Yb3+ were embedded into mesoporous organosilica, that absorbs light at 350 nm, and via charge transfer state, provide an electron to the emitting ion.88 Two Stark sublevel based emissions were used, located at around 980–990 and 1015–1025 nm. Their maximum sensitivity was 0.17% K−1 , clearly not suitable for biomedical applications [88]. Recently, this emission line of Yb3+ ion is used to sense temperature via the τ technique [89]. An inert core@active shell@inert shell structure of NaYF4 @NaYF4 :Yb3+ , Nd3+ @CaF2 was prepared via thermal decomposition and also functionalized with polyacrylic acid to make it water soluble. The lanthanide ions were incorporated within the middle active shell structure to reduce the deactivation processes from both the crystal defects in the core and the surface quenching centres. After pumping at 800 nm, the lifetime of the emission of Yb3+ ion at 1000 nm was manifested as a function of temperature. Before that, the concentration of both lanthanide ions within the middle active shell were optimized based on the intensity of the emission generated and the thermometric performance. These concentrations can affect the ET and BET processes between Yb3+ to Nd3+ , and the energy migration process between Yb3+ ions. The optimized concentrations of the ions were 20% Yb3+ and 60% Nd3+ . In terms of sensitivity, the maximum Srel , was 1.40% K−1 at room temperature. What is more interesting, this sensitivity was stable against continuous intense laser irradiations or harsh pH from 1 to 12 and was not affected by the concentration of the nanoparticles. Clearly, the lifetime of the emission of these ions is more sensitive for temperature determination than the intensity ratio based on the Stark sublevels. When this emission line of Yb3+ is combined with that arising from other lanthanide emitting ions, there is a boost in sensitivity. Usually, the combinations are Yb3+ with Tm3+ (emission at 1230 nm), or Yb3+ with the two emissions of Nd3+ ions within the II-BW (located at around 1050 and 1300 nm). What is novel in these combinations, is the fact that the role of the sensitizer is now assigned to either Tm3+ or Nd3+ ions, taking profit of their absorption bands from 690 to 808 nm, respectively. In this way, the heating problem of the 980 nm excitation source is overcome. Within these examples, the importance of the type of structures is highlighted. For example, SNR in a simple core and a core@shell structures was examined. Often simple core structures display low SNR as a result of their interaction with organic molecules, which increases the probability of non-radiative processes to take place. The core@shell structure not only can block the non-radiative processes, but also allow to manipulate ET processes by incorporating the lanthanide ions at different locations within the structure. The combination of Yb3+ and Tm3+ emissions is more sensitive to temperature changes when compared to the combination of Yb3+ with Nd3+ emissions. Yb3+ and Tm3+ emissions were combined in a core@shell structure based on Er3+ , Yb3+ :LaF3 @Tm3+ , Yb3+ :LaF3 nanocrystals. These types of nanocrystals were synthesized via wet chemical methodologies, having a final size of around 32 nm [91]. Within this structure, the core and the shell are both active, meaning that they are both composed of activators, whose emissions are used for thermal sensing. The

Luminescent Nanothermometers Operating Within Biological Windows

253

SNR and the thermometric performance of these structures were compared with the simple core of the form Er3+ , Yb3+ , Nd3+ :LaF3 structures. Under 690 or 808 nm excitation wavelengths, the emission lines of Yb3+ located at 1000 nm, and the one of Tm3+ located at 1230 nm, were used for thermal sensing. Shortly, the energy of the excitation source was absorbed by Tm3+ ions, which promote their electrons to the 3 F2,3 excited state. From there, CR process 3 F2,3 ;3 H6 → 3 F4 ;3 H5 , populates the 3 H5 level of this ion. A radiative decay from this level, back to the ground state, generates the emission of Tm3+ ion in the II-BW. In addition, the 3 H5 level of Tm3+ is resonant in energy with the 2 F5/2 level of Yb3+ , allowing an ET process to take place and populate that level of Yb3+ . A radiative relaxation to the ground state, generates the emission of Yb3+ within the II-BW. Further ET processes between the shell and the core, can lead to the population of Er3+ ions, which can generate an emission in the III-BW, such as the 1550 nm emission (Fig. 10a). If we compare the performance of the active core@active shell structure versus the simple core within the II-BW, the core@shell structure had a maximum value of sensitivity of 3.9% K−1 , which was approximately 3-times higher when compared to that of the simple core structure. The authors assigned this boost in sensitivity to the efficient ET and quenching rates within the core@shell structure. The emission of Yb3+ can be combined with the Nd3+ emission (at 1060 nm) in a core@shell structure of the form Er3+ , Yb3+ :NaYF4 @Nd3+ , Yb3+ :NaYF4 . For temperature sensing in the II-BW, the emission of the two activators in the outer shell was used. These emissions are triggered by excitation at 808 nm, which is absorbed by the Nd3+ ions in the outer shell, therefore Nd3+ ions have a dual role in this structure: sensitizers and activators. The intensity ratio of the Nd3+ emission versus the Yb3+ emission, generated a maximum sensitivity of 2.1% K−1 at 370 K [92]. Other types of more complex structures such as Tm3+ , 3+ 3+ 3+ 3+ 3+ 3+ Yb :SrF2 @Y :SrF2 @Er , Nd , Yb :SrF2 @Nd :SrF2 , have been also used as nanothermometers [90]. Their performances were based on the emissions arising from the Nd3+ ions located in shell 3 and Yb3+ ions located in shell 2. An important implication on the performance of these thermometers was the concentration of Er3+ ions located in shell 2. With the increase of the concentration of these ions, several ET processes between Yb3+ and Er3+ take place, which quench the intensity of the emission of Yb3+ ions as the probability of the Yb3+ → Er3+ ET process is increased. By this quenching process, the thermometric performance is favoured. Increasing the concentration of Er3+ ions from 0 to 8%, continuously increased the slope of I between the Nd3+ and Yb3+ emissions (Fig. 10b). This slope is proportional to the Srel , therefore, the maximum performance was achieved for 8% Er3+ ions [90], reaching a value of 1.62% K−1 within the physiological range of temperatures (Fig. 10c).

254

A. Nexha et al.

Fig. 10 a Mechanisms of the generation of the emission lines within the II-BW of Er3+ , Yb3+ :LaF3 @Tm3+ , Yb3+ :LaF3 nanocrystals under 690 nm excitation. Effect of the concentration of Er3+ ions on the: b I (normalized values) and c Srel of complex SrF2 structure. Data extracted from Ref. [90]. d Mechanisms of the generation of the emission lines within the II-BW of Er3+ , Ho3+ , Yb3+ :NaGdF4 @Yb3+ :NaGdF4 @Nd3+ , Yb3+ :NaGdF4 @NaGdF4 multishell nanostructures under 806 nm excitation

Luminescent Nanothermometers Operating Within Biological Windows

5.3.2

255

Nd3+

As on the example of Yb3+ doped thermometers, Nd3+ doped thermometers operating within the II-BW are based on single emitting Nd3+ materials, and materials which combine the emission of Nd3+ with those of other lanthanide ions (or QDs as described in Sect. 5.1). Single emitting Nd3+ doped thermometers operate based on the 1060 and 1350 nm emission lines. The Stark sublevels of the 1060 nm are very close to each other, and their thermometric performance is limited by ΔE. That is why their maximum Srel is within the range from 0.15 to 0.50% K−1 , depending on the type of host where the ions are embedded [1]. Besides this, quite often is not possible to distinguish the Stark levels of these emissions lines, leading to low signal discriminability. In order to address this issue, Skripka et al. implemented a highly sensitive InGaAs NIR detector, which not only allowed to properly separate the energy levels, but also increased the sensitivity up to 0.49% K−1 [41], although still not high enough for biomedical applications. On the other hand, the performance of the dual emitting thermometers based on Nd3+ emissions and those of other emitters (such as Ln or QDs) is more sensitive that the single emitting materials. Dual emitting thermometers reached Srel values as high as 1.6% K−1 and above. Here, we discuss the example of another complex multishell structure which is based on the Nd3+ emission within the II-BW. This structure is composed of an active core (Er3+ , Ho3+ , Yb3+ :NaGdF4 ), an inner shell composed of Yb3+ :NaGdF4 , a second active shell containing Nd3+ , Yb3+ :NaGdF4 , and a final inert shell of NaGdF4 [93]. The multishell structure was prepared via a thermal decomposition methodology, followed by a surface functionalization to render them water soluble and appropriate for biomedical applications. This is the reason why the third inert shell was included, to reduce the de-excitation channels arising from the water molecules. The structure was pumped at 806 nm. Shortly, this wavelength is absorbed by Nd3+ ions in shell 2, which after promoting their electrons to the excited state, relaxes non-radiatively to the 4 F3/2 level, which now relaxes radiatively to the 4 I13/2 level to generate the emission at 1340 nm (Fig. 10d). From the 4 F3/2 level, several multiple ET processes can populate the excited levels of Ho3+ (5 I7 ) and Er3+ (4 I11/2 ) at the core. From this excited level, Ho3+ can relax radiatively to the ground state generating the emission located at 1180 nm. On the other hand, Er3+ can generate emissions in the III-BW (Fig. 10d). With the increase of temperature from 293 to 323 K, the intensity of the emission of Ho3+ increased due to phonon assisted ET Yb3+ (2 F5/2 ) → Ho3+ (5 I6 ) process, while the intensity of the emission of Nd3+ decreased due to the phonon-assisted ET Nd3+ (4 F3/2 ) → Yb3+ (2 F5/2 ) process [93]. The intensity ratio between these two emissions achieved a maximum sensitivity of 1.17% K−1 at room temperature [93]. As a summary about thermometers operating within the II-BW, the focus is switched to the development of thermometers operating within the physiological range of temperatures (Fig. 11). In terms of sensitivities, QDs based on Ag2 S materials appear to exhibit the highest Srel (Fig. 11). In addition, they could be applied in different classes of luminescent thermometry techniques, such as E, I , γ , λ and

256

A. Nexha et al. 5 4

91

83

Srel, % K-1

82

3 2 90

1

79

93

89 78 80

0

87

41

300

350

400

450

T, K Fig. 11 Srel of different types of LTs operating within the II-BW: QDs based on Ag2 S [83], and PbS [78] (in red); TM based on Mn5+ [80], Ni2+ [82], and Cr4+ [87], (in grey); Ln based on single Nd [41], single Yb [89], co-doped Yb, Tm [91], co-doped Nd, Yb [90], co-doped Nd, Ho [93], (in turquoise); and mixed QDs/Ln [79] (in violet) materials. Numbers portray the references from which the data were extracted

τ . What makes these materials more appealing for biomedical applications are their properties such as ultrasmall sizes (around 10 nm), being able to absorb NIR excitation sources, and high quantum yields (in the range of 80%). Especially for the last point, Ln, which could compete with Ag2 S QDs, struggle to improve it, that is why a lot of efforts are focused on the development of core@shell structure or multistructures, to reduce the non-radiative processes and increase the SNR.

6 Thermometers Operating in the Third Biological Window At the III-BW, only Ln doped materials based on Er3+ , Tm3+ and Ho3+ have been analysed as luminescent thermometers operating in this range up to date (Fig. 1b). The performance of Er3+ is mainly assigned to its emission located at around 1550 nm, while Tm3+ , although displays two emissions around at 1450 and 1800 nm, is most often found co-doped with Ho3+ ions, which emit at 1960 nm [9, 47, 68]. This type of co-doping is attractive as it allows the development of self-assessed photothermal agents, i.e. materials that are able to generate heat and simultaneously monitor the heat based on the intensity ratio between their emissions (1800 nm of Tm3+ and 1960 nm of Ho3+ ).

Luminescent Nanothermometers Operating Within Biological Windows

257

6.1 Er3+ The Stark sublevels of the 4 I13/2 → 4 I15/2 electronic transition of Er3+ ions are used to extract the thermometric performance within the III-BW. These Stark sublevels are located very close to each other, that is why the sensitivity is lower than 0.5% K−1 [1]. Up to date, however, it looks like the attention is switched towards increasing the SNR of this band instead of improving the separation of the emission lines due to the different Stark sublevels. Several strategies have been implemented. For example, in LiErF4 @LiYF4 structures [94], different amounts of Ce3+ ions were introduced and their effect on the brightness of the emission and the thermometric performance was analyzed. It was concluded that 1 mol% of these ions displayed the brightest emission, and also the highest Srel . Nevertheless, the values are in the range of 0.40–0.50% K−1 . Also Mn2+ ions have a positive feedback on the brightness of the Er3+ emission in the III-BW when embedded into the hexagonal NaScF4 host [95]. The sensitivity as thermometers is similar to those reported before. Similarly, but focused on the role of the hosts, Savchuk et al. [10] investigated doped Er3+ ions in different hosts and observed the effect of the brightness and the thermometric performance. Er3+ ions were embedded into hosts like oxides, double oxides, oxyfluorides and fluorides. Among these hosts, fluorides exhibited the brightness emission (Fig. 12a). Nevertheless, the thermometric performance was very poor, around 2- to 5-times lower compared to that of LiErF4 @LiYF4 nanoparticles or Er3+ :NaScF4 thermometers.

6.2 Ho3+ , Tm3+ Co-doped Ho3+ , Tm3+ materials could be excited at 808 nm as Tm3+ ions are able to absorb the energy of this wavelength. There could be also another option for excitation of these materials by adding Yb3+ as sensitizer and pumping them at 980 nm. However, besides the well-known absorption problem by water molecules, when pumping at 980 nm, these materials display poor thermometric sensitivities with values around 0.1% K−1 [1]. When the emissions of Ho3+ and Tm3+ are combined, a higher thermometric performance could be achieved. Tm3+ ions can absorb the energy of the 808 nm laser source and promote their electrons to the 3 H4 excited state. From this level, a radiative decay to the 3 F4 level led to the emission at 1450 nm. Another decay from the 3 F4 level to the ground state, generates the second emission of Tm3+ ions at 1800 nm (Fig. 12b). In addition, the 3 F4 level is resonant in energy with the 5 I7 level of Ho3+ , allowing ET and BET processes to take place. In this way, the 5 I7 energy level of Ho3+ is populated, which after decaying to the ground state generates the emission of Ho3+ ions at 1960 nm (Fig. 12b). As discussed in the I-BW, these ions are encountered in the monoclinic KLu(WO4 )2 host and synthesized via four different methodologies which renders

258

A. Nexha et al.

Fig. 12 a Intensity of the emission of Er3+ ions doped into different types of hosts. Adapted from Ref. [10]. b Mechanisms of the generation of the emission lines in the III-BW of Ho3+ , Tm3+ co-doped materials. c Emissions in the III-BW of Ho3+ , Tm3+ doped KLu(WO4 )2 nanoparticles. Adapted from Ref. [47]. d Intensity ratios used for luminescent thermometry in the III-BW for Ho3+ , Tm3+ co-doped KLu(WO4 )2 nanoparticles. Adapted from Ref. [9]

different morphological characteristics to the final product: P with particles around 2 μm with no defined shape [9], CA and MW with sizes below 20 nm and no shape [68], and TD with defined rod shape and sizes in the range of 1 μm [47]. Regardless of the synthetic methodology applied, all Ho3+ , Tm3+ co-doped particles generated emissions within the III-BW (Fig. 12c). Among them, the brightest emissions were generated from the particles synthesized via the P methodology (Fig. 12c). On the contrary, well-defined rods obtained by the TD methodology, exhibit the lowest intensity. There are three options of luminescent thermometry within the III-BW from these Ho3+ , Tm3+ co-doped materials, based on I of: (i) the 1450 and 1800 nm emissions of Tm3+ ions, (ii) the 1450 nm emission of Tm3+ ions and the 1960 nm of Ho3+ ions, and (iii) the 1800 nm emission of Tm3+ ions and the 1960 nm emission of Ho3+ ions. Among them, the highest slope as a function of the temperature was exhibited by the intensity ratio between the 1800 nm emission of Tm3+ and the 1960 nm emission of Ho3+ (Fig. 12d) [9]. Therefore, in order to compare the

Luminescent Nanothermometers Operating Within Biological Windows 0.9 0.8 9

0.7

Srel, % K-1

Fig. 13 Srel of different types of LTs operating within the III-BW: single Er [10, 94], (in gold); single Tm [10], (in red); and co-doped Ho, Tm materials obtained by different synthetic methodologies (P—[9], CA—[68], MW—[68] and TD—[47]) (in turquoise). Numbers portray the references from which the data were extracted

259

0.6 0.5 0.4

0.2 0.1 0.0 290

47

94

0.3

68 68

10 10

10

300

310

320

330

340

T, K

performance of these thermometers as a function of the morphological characteristics, this intensity ratio was chosen. From the comparison, it could be concluded that as the size of these particles decreased, their Srel decreased [47], therefore among all these thermometers, the ones synthesized by the P methodology displayed the maximum sensitivity around 1% K−1 at room temperature (Fig. 13). It is argued that this trend is due to the fact that by decreasing the size of the particles, it increases the interaction between the emitting ions and the ligands attached to the surface of the particles, which in turn, quenched the luminescence of the particles and therefore also detrimentally affected to their thermometric performance [47]. Within this biological window, clearly the Ho3+ , Tm3+ co-doping strategy appears to be more sensitive compared to Er3+ based thermometers (Fig. 13). Nevertheless, their sensitivities are around 1% K−1 , which compared to other thermometers in other biological windows is still lower. It should be emphasized that the benefits of Ho3+ , Tm3+ co-doping relies on their ability to produce self-assessed photothermal agents [9, 47, 68].

7 Applications of Thermometers Operating Within the Biological Windows The development of different sensing techniques and highly sensitive thermometers has continuously led to application of these luminescent thermometers within the different BWs. Generally, I thermometry was applied quite often to monitor biomedical processes, such as in-vivo acquisition of 2D subcutaneous thermal images, in-vitro temperature sensing, or as ex-vivo self-assessed photothermal agents [1]. However, it has been demonstrated from some types of emitting materials (especially Ag2 S QDs) that when their luminescence spectra are acquired inside biological

260

A. Nexha et al.

tissues within BWs, spectral distortions might take place, which imply erroneous temperature readouts [96]. Although methods to improve and reduce these errors have been adapted [96], new alternatives for fully reliable thermal readings are still needed. To this end, recently a trend towards the application of τ . nanothermometry for biomedical applications is continuously being explored. Tan et al. [89] explored τ nanothermometry of the 1000 nm emission of Yb3+ ion arising from NaYF4 @NaYF4 :20% Yb3+ , 60% Nd3+ @CaF2 nanoparticles to diagnose in vivo murine inflammation. First, the applicability for in vivo sensing of these emitters was tested by studying the lifetime at different scenarios. For example, the long-time photostability of the emitters was tested. They were continuously pumped at 800 nm at high power densities (around 2 W/cm2 ) and their lifetime remained invariant over 5 h. Repeatability was also tested over 5 cycles of cooling and heating from room temperature to 313 K. No thermal hysteresis was observed, proving good repeatability of the data. It was also crucial to understand the variation of the lifetime as a function of the concentration of the particles and the pH of the medium. For example, tumours tissues have extracellular acidic microenvironments (pH 6.5–6.9) [89]. Both these parameters did not substantially affect the lifetime. Based on these pretesting proves, the nanothermometers were applied to sense the temperature within an in vivo model based on thermographic images. To this end, first 100 μL of a 10% concentration of the fluorescent particles into a yeast solution was prepared, and then simultaneously injected into the back of the mouse to induce inflammation. An inflammation in mouse is usually accompanied by an increase of the body temperature. As a reference, a second injection with pure saline buffer was injected into a second mouse. Also, a subcutaneous injection of the nanoparticles (100 μL, 10 mg/mL) was applied to the back of the two mice. Later, after 16 h post injection, using a high resolution InGaAs camera, the NIR-II luminescence images generated by the luminescent nanoparticles were captured. These images proved the presence of the nanothermometers inside the inflamed and healthy (normal) mouse (Fig. 14b). At the same time, also images with a thermal camera were acquired, which demonstrated a slight increment (≈ 2.6 ºC) in the skin temperature of the mouse. Then, time gating methodology was adopted to obtain the luminescence lifetime images of both mice. From the images, the lifetime of the nanothermometers subjected to the yeast injection was on average 0.68 ms, while in the control mouse the average lifetime was calculated to be around 0.70 ms. This leads to a 2.8% reduction of the lifetime of the nanothermometers within the inflamed mouse, which is a result of the increase of the temperature in the subcutaneous injection. From the calibration curve of the thermometers, the reduction of lifetime could be correlated to a temperature increment of 2.3 ºC, in agreement with the temperature change registered with the thermal camera. In this way, thermographic images based on the lifetime of the Yb3+ emission located in the II-BW could be used to detect in vivo inflammation. Besides these results, one crucial problem of Ln materials is their low brightness [3], compared to other types of emitters, especially Ag2 S QDs. Typically, lanthanides have brightnesses around 200–300 M−1 cm−1 [3], which at a certain degree might

Luminescent Nanothermometers Operating Within Biological Windows

261

Fig. 14 a Principle of thermographic lifetime technique for in vivo thermal sensing. b Luminescence intensity imaging of the nanoparticles within the inflamed and normal mouse. c Thermal images of the inflamed and normal mouse, and d thermographic luminescence lifetime-hued imaging. Data adapted from Ref. [89]

affect their reliability. On the other hand, Ag2 S QDs display a brightness 4- to 5times higher than that of lanthanide doped materials [86]. Taking profit of this high brightness, Ag2 S quantum dots have been applied to remotely monitor the absolute temperature within internal organs such as liver by using the principle of τ nanothermometry [86]. Before actual application in the in vivo model, the reliability of the emission of Ag2 S QDs in the II-BW was tested. Experiments involved the repeatability by exposing the particles at different cycles of cooling and heating and monitoring the variation of their lifetimes. Data were reproducible up to 99%. Next, the effect of a testing biological tissue such a phantom tissue was explored. Tuning the thickness of the tissue (2 and 4 mm), did not significantly influence the lifetime of the emissions. After the pre-testing, the in vivo application in the liver of a mouse was explored. Liver was selected as an internal organ, because it has been observed previously that after intravenous injection, Ag2 S QDs were accumulated in this organ [84]. First, the quantum dots dispersed in phosphate-buffered saline (PBS, 100 μL of a 0.5 mg/mL dispersion) were injected intravenously. Second, bacterial lipopolysaccharides (LPS) with a concentration of 10 mg/kg, were injected to induce inflammation (Fig. 15a). Inflammation was confirmed by a significant increase in the hepatic gene expression of several proinflammatory markers such as tumor necrosis factor-alpha (TNFα), Interleukin-6 (IL-6), Interleukin-1β (IL-1β), and inducible nitric oxide synthase (iNOS) (Fig. 15b). On the other hand, the levels of these markers in the control mice did not change. Before the inflammatory process, the lifetime of Ag2 S QDs was recorded to determine the basal temperature and the thermal stability of the liver. The anesthetized mouse was illuminated with 800 nm laser pulses of 10 ns duration, with a spot size of 0.1 cm2 . The luminescence generated by the QDs was collected by a set of lenses

262

A. Nexha et al.

Fig. 15 a Experimental setup for real-time thermal monitoring of liver during inflammation, b gene expression of different types of proinflammatory markers, c evolution of the fluorescence lifetime before LPS injection as a function of time, d time evolution of the fluorescence lifetime after LPS injection as a function of time, e time evolution of liver temperature after LPS injection (brown circles) and rectal temperature (red squares). PMT stands for photomultiplier. All data adapted from Ref. [86].

and spectrally filtered by a monochromator. The decay of the intensity was finally recorded by an infrared photomultiplier (PMT) connected to a digital oscilloscope (Fig. 15a). The lifetime was 239 ns (Fig. 15c), which rendered a temperature in the liver of 38.7 ºC with a temperature fluctuation of ± 0.3 °C. After verifying that the liver temperature was stable, the lifetime of the QDs was continuously monitored while LPS was administrated intravenously. It was detected a continuous decrease of the lifetime of the emission after injection of LPS (Fig. 15d). Taking profit from the calibration curve of the nanothermometer, the temperature of the liver as function of time could be extracted (Fig. 15e). Within the first hour, the temperature of the liver tended to increase, and it later remained almost constant. The body core temperature (data recorded via a rectal probe and shown in Fig. 15e), started rising only 20 min after the drug administration, contrary to the immediate increase in the liver temperature. For the liver, the increase of temperature is related to the blood perfusion and the local metabolic activity because of prostaglandin release.

Luminescent Nanothermometers Operating Within Biological Windows

263

8 Conclusions Different types of LTs are continuously being developed to be used within the different BWs. Among all these materials, Ln doped materials and QDs based on Ag2 S nanoparticles are the ones attracting more attention due to their ability to absorb and emit within BWs ranges, particularly at the II-BW or III-BW. Another feature of these materials, not mentioned within this chapter, is their ability to generate heat within the medium in which they are embedded. Therefore, these materials not only generate heat, but also sense the temperature induced by the heat that they generate as they act also as luminescent nanothermometers, allowing to develop in this way selfassessed photothermal agents, which are appealing for hyperthermia applications [1]. Among other materials, Cr3+ doped materials should be investigated further as they have absorption bands at the I-BW and emit in the II-BW. Other luminescent thermometers such as ND, Au NPs and C QDs are still limited to excitations in the UV or VIS with emissions in the border of the I-BW. In terms of thermometry techniques, clearly two of them are the most applied and reliable options: I and τ thermometry. Band-shape is not dependent on the concentration of the materials and has an emission as reference. However, the spectral distortion when emissions are acquired within biological media is hampering their use [96]. On the other hand, lifetime thermometry does not suffer from this drawback and is more reliable for temperature sensing within biological tissues. However, up to date, the materials used based on this technique, are only Ln materials or Ag2 S QDs, with the latter being more promising as they exhibit brighter emissions, therefore allowing for easier readouts. Finally, we highlight a few challenges that should be addressed. First, the largescale synthesis of the luminescent nanoparticles remains a drawback. In addition to this, reproducibility of the synthesis and the influence of batch-to-batch synthesis to the thermometric performance should be carefully investigated. Second, more efforts should be devoted to the preparation of fully operative LTs within the BWs: excited within BWs, and emissions in BWs, which in rare examples is applied. Third, crucial for biological applications, is the brightness of the luminescent materials: higher brightness led to more reliable results. Regarding the brightness, almost all luminescent materials are not bright enough, probably due to their low absorption cross sections and due to the synthetic methodologies applied (the final products are coated with organic surfactants which due to their vibrational modes quench the photoluminescence of the fluorescent particles). Last, more efforts should be placed into lifetime thermometry, since this technique, as already mentioned, is free of spectral distortions when the spectra are recorded inside the biological tissues and is more reliable than the band-shape thermometry. However, it appears to be limited, for the moment, only to Ln materials and Ag2 S QDs.

264

A. Nexha et al.

List of Symbols E I γ λ τ ⌃ Sabs Δ T Srel δT δΔ/Δ ΔE

Intensity (parameter symbol for intensity based thermometry) Intensity (parameter symbol for Band-shape thermometry Bandwidth (parameter symbol for bandwidth thermometry) Spectral position or wavelength Lifetime Polarization thermometry Absolute thermal sensitivity Thermometric parameter Absolute temperature Relative thermal sensitivity Temperature resolution Uncertainty of the thermometric parameter Energy gap

List of Abbreviations LTs UV VIS NADH FAD NIR BWs I-BW II-BW III-BW IV-BW SNR SWIR ND Au NCs QDs C QDs CdX QDs TM Ln FWHM TCLs NTCLs YAP

Non-contact based luminescent thermometers Ultraviolet Visible Nicotinamide adenine dinucleotide hydrate Flavin adenine dinucleotide Near infrared Biological windows First biological window Second biological window Third biological window Fourth biological window Signal-to-noise ratio Short-wavelength infrared region Nanodiamonds Fluorescent gold nanoclusters Quantum dots Carbon based quantum dots Cadmium based quantum dots Transition metal Lanthanide ions Full width at half-maximum Thermally coupled levels Non-thermally coupled levels Yttrium orthoaluminate perovskite

Luminescent Nanothermometers Operating Within Biological Windows

ET BET P CA MW TD PAET CR PBS LPS TNF-α IL-6 IL-1β iNOS PMT

265

Energy transfer Back energy transfer Sol-gel Pechini methodology Conventional autoclave solvothermal methodology Microwave-assisted solvothermal methodology Thermal decomposition methodology Phonon assisted energy transfer Cross relaxation Phosphate-buffered saline Lipopolysaccharides Tumor necrosis factor-alpha Interleukin-6 Interleukin-1β Inducible nitric oxide synthase Photomultiplier

References 1. A. Nexha, J.J. Carvajal, M.C. Pujol, F. Díaz, M. Aguiló, Nanoscale 13, 7913–7987 (2021) 2. C.D.S. Brites, P.P. Lima, N.J.O. Silva, A. Millán, V.S. Amaral, F. Palacio, L.D. Carlos, Nanoscale 4, 4799–4829 (2012) 3. A. Bednarkiewicz, L. Marciniak, L.D. Carlos, D. Jaque, Nanoscale 12, 14405–14421 (2020) 4. E. Hemmer, A. Benayas, F. Légaré, F. Vetrone, Nanoscale Horizons 1, 168–184 (2016) 5. A.M. Smith, M.C. Mancini, S. Nie, Nat. Nanotechnol. 4, 710–711 (2009) 6. S.L. Jacques, Phys. Med. Biol. 58, R37–R61 (2013) 7. R. Weissleder, Nat. Biotechnol. 19, 316–317 (2001) 8. L. Shi, L.A. Sordillo, A. Rodríguez-Contreras, R. Alfano, J. Biophotonics 9, 38–43 (2016) 9. A. Nexha, J.J. Carvajal, M.C. Pujol, F. Díaz, M. Aguiló, J. Mater. Chem. C 8, 180–191 (2020) 10. O.A. Savchuk, J.J. Carvajal, P. Haro-Gonzalez, M. Aguiló, F. Díaz, J. Alloy. Compd. 746, 710–719 (2018) 11. D.J. Naczynski, M.C. Tan, M. Zevon, B. Wall, J. Kohl, A. Kulesa, S. Chen, C.M. Roth, R.E. Riman, P.V. Moghe, Nat. Commun. 4, 2199 (2013) 12. B. del Rosal, I. Villa, D. Jaque, F. Sanz-Rodríguez, J. Biophotonics 9, 1059–1067 (2016) 13. M.H. Alkahtani, F. Alghannam, L. Jiang, A. Almethen, A.A. Rampersaud, R. Brick, C.L. Gomes, M.O. Scully, P.R. Hemmer, Nanophotonics 7, 1423–1453 (2018) 14. W. Liu, M.N.A. Alam, Y. Liu, V.N. Agafonov, H. Qi, K. Koynov, V.A. Davydov, R. Uzbekov, U. Kaiser, T. Lasser, F. Jelezko, A. Ermakova, T. Weil, Nano Lett. 22, 2881–2888 (2022) 15. S. Choi, V.N. Agafonov, V.A. Davydov, T. Plakhotnik, ACS Photonics 6, 1387–1392 (2019) 16. L. Shang, F. Stockmar, N. Azadfar, G.U. Nienhaus, Angew. Chem. 52, 11154–11157 (2013) 17. J.R. Macairan, D.B. Jaunky, A. Piekny, R. Naccache, Nanoscale Adv. 1, 105–113 (2019) 18. H. Zhao, A. Vomiero, F. Rosei, Small 16, 2000804 (2020) 19. A. Teitelboim, D. Oron, ACS Nano 10, 446–452 (2016) 20. K.M. Tsoi, Q. Dai, B.A. Alman, W.C.W. Chan, Acc. Chem. Res. 46, 662–671 (2013) 21. Y. Shen, J. Lifante, E. Ximendes, H.D.A. Santos, D. Ruiz, B.H. Juárez, I. Zabala Gutiérrez, V. Torres Vera, J. Rubio Retama, E. Martín Rodríguez, D.H. Ortgies, D. Jaque, A. Benayas, B. del Rosal, Nanoscale 11, 19251–19264 (2019)

266

A. Nexha et al.

22. M. Drami´canin, Luminescence Thermometry, ed. by M. Drami´canin (Woodhead Publishing, 2018), pp. 113–157. https://doi.org/10.1016/B978-0-08-102029-6.00006-3 23. R. Marin, D. Jaque, Chem. Rev. 121, 1425–1462 (2021) 24. R. Marin, D. Jaque, A. Benayas, Nanoscale Horizons 6, 209–230 (2021) 25. W. Zhang, D. Lin, H. Wang, J. Li, G.U. Nienhaus, Z. Su, G. Wei, L. Shang, Bioconjug. Chem. 28, 2224–2229 (2017) 26. S. Li, K. Zhang, J.M. Yang, L. Lin, H. Yang, Nano Lett. 7, 3102–3105 (2007) 27. L.M. Maestro, E.M. Rodríguez, F.S. Rodríguez, M.C.I. la Cruz, A. Juarranz, R. Naccache, F. Vetrone, D. Jaque, J.A. Capobianco, J.G. Solé, Nano Lett. 10, 5109–5115 (2010) 28. L.M. Maestro, C. Jacinto, U.R. Silva, F. Vetrone, J.A. Capobianco, D. Jaque, J.G. Solé, Small 7, 1774–1778 (2011) 29. J. Liu, H. Zhang, G.S. Selopal, S. Sun, H. Zhao, F. Rosei, ACS Photonics 6, 2479–2486 (2019) 30. F. Yuan, T. Yuan, L. Sui, Z. Wang, Z. Xi, Y. Li, X. Li, L. Fan, Z. Tan, A. Chen, M. Jin, S. Yang, Nat. Commun. 9, 2249 (2018) 31. J.R. Macairan, T.V. de Medeiros, M. Gazzetto, F. Yarur Villanueva, A. Cannizzo, R. Naccache, J. Colloid Interface Sci. 606, 67–76 (2022) 32. A. Nexha, M. Cinta Pujol, J. Josep Carvajal, SPIE (2021). https://doi.org/10.1117/3.260432 6.ch2 33. M. Back, E. Trave, J. Ueda, S. Tanabe, Chem. Mater. 28, 8347–8356 (2016) 34. Y. Wang, H. Suo, L. Li, G. Wang, W. Deng, W. Ding, Z. Wang, P. Li, Z. Zhang, Phys. B 625, 413496 (2022) 35. E. Glais, M. Pellerin, V. Castaing, D. Alloyeau, N. Touati, B. Viana, C. Chanéac, RSC Adv. 8, 41767–41774 (2018) 36. W.M. Piotrowski, K. Trejgis, M. Dramicanin, L. Marciniak, J. Mater. Chem. C 9, 10309–10316 (2021) 37. Y. Qin, F. Zhong, Y. Bian, S. Hariyani, Y. Cao, J. Brgoch, T. Seto, M.G. Brik, A.M. Srivastava, X. Wang, Y. Wang, Inorg. Chem. 61, 8126–8134 (2022) 38. F. Venturini, M. Baumgartner, S.M. Borisov, Sensors 18, 668 (2018) 39. K. Kniec, A. Kochanowska, L. Li, M. Suta, L. Marciniak, J. Mater. Chem. C (2022). https:// doi.org/10.1039/D2TC00952H 40. Y. Tanabe, S. Sugano, J. Phys. Soc. Jpn. 9, 766–779 (1954) 41. A. Skripka, A. Morinvil, M. Matulionyte, T. Cheng, F. Vetrone, Nanoscale 11, 11322–11330 (2019) 42. W. Xu, H. Zhao, Z. Zhang, W. Cao, Sens. Actuators B Chem. 178, 520–524 (2013) 43. S. Acosta, L.J. Borrero-González, P. Umek, L.A.O. Nunes, P. Guttmann C. Bittencourt 21, 5306 (2021) 44. L. Marciniak, A. Bednarkiewicz, J. Drabik, K. Trejgis, W. Strek, Phys. Chem. Chem. Phys. 19, 7343–7351 (2017) 45. Ł Marciniak, A. Bednarkiewicz, D. Hreniak, W. Strek, J. Mater. Chem. C 4, 11284–11290 (2016) 46. D. Chen, Y. Liang, S. Miao, J. Bi, K. Sun, J. Lumin. 234, 117967 (2021) 47. A. Nexha, J.J. Carvajal, M.C. Pujol, F. Díaz, M. Aguiló, J. Mater. Chem. C 9, 2024–2036 (2021) 48. O.A. Savchuk, J.J. Carvajal, C.D.S. Brites, L.D. Carlos, M. Aguilo, F. Diaz, Nanoscale 10, 6602–6610 (2018) 49. M.A. Hernández-Rodriguez, A.D. Lozano-Gorrín, V. Lavín, U.R. Rodríguez-Mendoza, I.R. Martín, Opt. Express 25, 27845–27856 (2017) 50. A.F. Pereira, J.F. Silva, A.S. Gouveia-Neto, C. Jacinto, Sens. Actuators B Chem. 238, 525–531 (2017) 51. O. Savchuk, J.J. Carvajal Marti, C. Cascales, P. Haro-Gonzalez, F. Sanz-Rodríguez, M. Aguilo, F. Diaz, Nanomaterials 10, 993 (2020) 52. M. Runowski, P. Wo´zny, N. Stopikowska, I.R. Martín, V. Lavín, S. Lis, ACS Appl. Mater. Interfaces 12, 43933–43941 (2020) 53. M. Runowski, A. Shyichuk, A. Tymi´nski, T. Grzyb, V. Lavín, S. Lis, ACS Appl. Mater. Interfaces 10, 17269–17279 (2018)

Luminescent Nanothermometers Operating Within Biological Windows

267

54. N. Wang, Z. Fu, Y. Wei, T. Sheng, J. Alloy. Compd. 772, 371–380 (2019) 55. L. Li, F. Qin, Y. Zhou, Y. Zheng, H. Zhao, Z. Zhang, J. Lumin. 202, 301–308 (2018) 56. E. Casagrande, M. Back, D. Cristofori, J. Ueda, S. Tanabe, S. Palazzolo, F. Rizzolio, V. Canzonieri, E. Trave, P. Riello, J. Mater. Chem. C 8, 7828–7836 (2020) 57. R. Wu, J. Zhou, L. Lei, S. Zhang, Z. Xiao, J. Zhang, S. Xu, Chem. Phys. Lett. 667, 206–210 (2017) 58. Y.Y. Bu, X.H. Yan, Appl. Phys. B 123, 59 (2017) 59. I.E. Kolesnikov, A.A. Kalinichev, M.A. Kurochkin, E.V. Golyeva, E.Y. Kolesnikov, A.V. Kurochkin, E. Lähderanta, M.D. Mikhailov, Sci. Rep. 7, 18002 (2017) 60. M.E. Raab, S.L. Maurizio, J.A. Capobianco, P.N. Prasad, J. Phys. Chem. B 125, 13132–13136 (2021) 61. A.S. Souza, L.A.O. Nunes, I.G.N. Silva, F.A.M. Oliveira, L.L. da Luz, H.F. Brito, M.C.F.C. Felinto, R.A.S. Ferreira, S.A. Júnior, L.D. Carlos, O.L. Malta, Nanoscale 8, 5327–5333 (2016) 62. K. Lu, X. Sun, L. Xu, B. Jiang, J. Ren, J.J. Carvajal, E. Zhao, L. Liu, J. Zhang, J. Alloy. Compd. 842, 155602 (2020) 63. A.C. Brandão-Silva, M.A. Gomes, Z.S. Macedo, J.F.M. Avila, J.J. Rodrigues, M.A.R.C. Alencar, J. Phys. Chem. C 122, 20459–20468 (2018) 64. P. Haro-González, I.R. Martín, L.L. Martín, S.F. León-Luis, C. Pérez-Rodríguez, V. Lavín, Opt. Mater. 33, 742–745 (2011) 65. K. Lu, Y. Yi, L. Xu, X. Sun, L. Liu, H. Li, Nanomaterials 10, 24 (2020) 66. M. Ding, M. Zhang, C. Lu, Mater. Lett. 209, 52–55 (2017) 67. H. Lu, H. Hao, Y. Gao, D. Li, G. Shi, Y. Song, Y. Wang, X. Zhang, Microchim. Acta 184, 641–646 (2017) 68. A. Nexha, M.C. Pujol, J.J. Carvajal, F. Díaz, M. Aguiló, Nanomaterials 11, 485 (2021) 69. H. Lu, H. Hao, G. Shi, Y. Gao, R. Wang, Y. Song, Y. Wang, X. Zhang, RSC Adv. 6, 55307–55311 (2016) 70. M. Ding, D. Chen, C. Lu, J. Xi, Z. Ji, Z. Xu, Mater. Lett. 189, 5–8 (2017) 71. C. Mi, J. Zhou, F. Wang, G. Lin, D. Jin, Chem. Mater. 31, 9480–9487 (2019) 72. A.F. Silva, F. Elan, E.L. Falcão-Filho, L.J.Q. Maia, C.B. de Araújo, J. Mater. Chem. C 5, 1240–1246 (2017) 73. Z. Zhao, F. Hu, Z. Cao, F. Chi, X. Wei, Y. Chen, C.-K. Duan, M. Yin, Ceram. Int. 43, 14951– 14955 (2017) 74. I.E. Kolesnikov, D.V. Mamonova, A.A. Kalinichev, M.A. Kurochkin, V.A. Medvedev, E.Y. Kolesnikov, E. Lähderanta, A.A. Manshina, Nanoscale 12, 5953–5960 (2020) 75. K. Trejgis, L. Marciniak, Phys. Chem. Chem. Phys. 20, 9574–9581 (2018) 76. J. Drabik, B. Cichy, L. Marciniak, J. Phys. Chem. C 122, 14928–14936 (2018) 77. K. Kniec, M. Tikhomirov, B. Pozniak, K. Ledwa, L. Marciniak, Nanomaterials 10, 189 (2020) 78. A. Benayas, F. Ren, E. Carrasco, V. Marzal, B. del Rosal, B.A. Gonfa, Á. Juarranz, F. SanzRodríguez, D. Jaque, J. García-Solé, D. Ma, F. Vetrone, Adv. Func. Mater. 25, 6650–6659 (2015) 79. B. del Rosal, E. Carrasco, F. Ren, A. Benayas, F. Vetrone, F. Sanz-Rodríguez, D. Ma, Á. Juarranz, D. Jaque, Adv. Func. Mater. 26, 6060–6068 (2016) 80. P.M. Gschwend, D. Niedbalka, L.R.H. Gerken, I.K. Herrmann, S.E. Pratsinis, Adv. Sci. 7, 2000370 (2020) 81. C. Matuszewska, K. Elzbieciak-Piecka, L. Marciniak, J. Phys. Chem. C 123, 18646–18653 (2019) 82. C. Matuszewska, L. Marciniak, J. Lumin. 223, 117221 (2020) 83. Y. Shen, H.D.A. Santos, E.C. Ximendes, J. Lifante, A. Sanz-Portilla, L. Monge, N. Fernández, I. Chaves-Coira, C. Jacinto, C.D.S. Brites, L.D. Carlos, A. Benayas, M.C. Iglesias-de la Cruz, D. Jaque, Adv. Func. Mater. 30, 2002730 (2020) 84. J. Lifante, Y. Shen, I. Zabala Gutierrez, I. Rubia-Rodríguez, D. Ortega, N. Fernandez, S. Melle, M. Granado, J. Rubio-Retama, D. Jaque, E. Ximendes, Adv. Sci. 8, 2003838 (2021) 85. H.D.A. Santos, D. Ruiz, G. Lifante, C. Jacinto, B.H. Juarez, D. Jaque, Nanoscale 9, 2505–2513 (2017)

268

A. Nexha et al.

86. Y. Shen, J. Lifante, I. Zabala-Gutierrez, M. de la Fuente-Fernández, M. Granado, N. Fernández, J. Rubio-Retama, D. Jaque, R. Marin, E. Ximendes, A. Benayas, Adv. Mater. 34, 2107764 (2022) 87. X. Chen, S. Liu, K. Huang, J. Nie, R. Kang, X. Tian, S. Zhang, Y. Li, J. Qiu, Chem. Eng. J. 396, 125201 (2020) 88. A.M. Kaczmarek, D. Esquivel, J. Ouwehand, P. Van Der Voort, F.J. Romero-Salguero, R. Van Deun, Dalton Trans. 46, 7878–7887 (2017) 89. M. Tan, F. Li, N. Cao, H. Li, X. Wang, C. Zhang, D. Jaque, G. Chen, Small 16, 2004118 (2020) 90. P. Cortelletti, A. Skripka, C. Facciotti, M. Pedroni, G. Caputo, N. Pinna, M. Quintanilla, A. Benayas, F. Vetrone, A. Speghini, Nanoscale 10, 2568–2576 (2018) 91. E.C. Ximendes, U. Rocha, T.O. Sales, N. Fernández, F. Sanz-Rodríguez, I.R. Martín, C. Jacinto, D. Jaque, Adv. Func. Mater. 27, 1702249 (2017) 92. L. Marciniak, K. Prorok, L. Francés-Soriano, J. Pérez-Prieto, A. Bednarkiewicz, Nanoscale 8, 5037–5042 (2016) 93. A. Skripka, A. Benayas, R. Marin, P. Canton, E. Hemmer, F. Vetrone, Nanoscale 9, 3079–3085 (2017) 94. C. Hazra, A. Skripka, S.J.L. Ribeiro, F. Vetrone, Adv. Opt. Mater. 8, 2001178 (2020) 95. G. Xiang, X. Liu, Q. Xia, X. Liu, S. Xu, S. Jiang, X. Zhou, L. Li, D. Wu, L. Ma, X. Wang, J. Zhang, Talanta 224, 121832 (2021) 96. Y. Shen, J. Lifante, N. Fernández, D. Jaque, E. Ximendes, ACS Nano 14, 4122–4133 (2020)

Luminescence Thermometry for in vivo Applications Erving Ximendes

Abstract This chapter presents a comprehensive account of the evolution of luminescent nanoparticles for thermometry, from their initial use in semitransparent organisms to their application in small animal models. It highlights the most notable experiments that have expanded the boundaries of applicability of this field and led to the successful application of these thermal probes for studying, diagnosing, and treating various tissues. In addition, the chapter elucidates the advantages of luminescent nanoparticles and presents a critical analysis of the potential areas for improvement.

1 Introduction Temperature is a parameter whose importance can be acknowledged by all people. It is simply too important for our daily lives. Even our senses help us to infer its impact. Case in point, we instinctively know what the current season is, or if we left the food for too much time inside a microwave. It is no surprise then that, through a physiological lens, it is advantageous for animals to maintain a reasonable and stable body temperature. After all, biochemical processes can be hastened. This, in turn, can result in more rapid conversion of food to energy, faster conduction of nerve impulses, improved coordination, faster growth and an increased rate of healing. When dealing with temperature measurements under in vivo or in vitro conditions, however, most of the commonly utilized thermometers are not capable of providing a thermal reading due to either their relatively big size or to the possible invasive and harmful contact that they could have on a living system. Hence, different methods of temperature sensing needed to be developed to circumvent such issues. Luminescence thermometry (LTh) was one of them. Since many materials used for LTh had a considerable sensitivity in the physiological temperature range (35–45 E. Ximendes (B) nanoBIG, Departamento de Física de Materiales, Universidad Autónoma de Madrid, Madrid, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_7

269

270

E. Ximendes

°C) range, they were readily recognized as minimally invasive thermal probes suitable for biological applications [1]. Compared with the traditional methods, LTh has the advantages of remote observation, short acquisition periods and the possibility of having real-time information. As the reader has been familiarized with the different types of LThs in the previous chapters, we will dedicate the subsections of the current one to describe how the biological applicability of luminescent thermometers has evolved over time. These subsections were organized according to the level of complexity of each of the organisms/animals studied. As the reader will notice, luminescence thermometry seem to be under a rather constant expansion of its applicability in the biological domain.

2 Semitransparent Organisms At the early infancy of luminescence thermometry, its applicability in living biological systems was restricted to small semitransparent organisms. Such a constraint had been imposed by the then common use of excitation and emission light in the visible range, where penetration of light is low [2]. Therefore, the use of these probes was restricted to small semitransparent organisms, such as fly larvae and the nematode Caenorhabditis elegans (C. elegans). A representative example of this level of application can be found in the work of Donner et al. [3], where the authors proposed a technique, based on measuring the fluorescence polarization anisotropy (FPA) of green fluorescent protein (GFP), for in vivo intracellular temperature mapping. Due to the biocompatibility of GFP, the authors were able to cross the chasm separating in vitro from in vivo measurements. As another example of use of such method, the authors performed an experiment in which the temperature was mapped in real time due to a local heat generation within the C. elegans. A practical example which concurred with the approach used photothermal therapies (Fig. 1).

Fig. 1 Study of C. elegans with gold nanorods and green fluorescent protein. a Bright field, fluorescence (green) and two photon (yellow) image of a living C. elegans after incubation with both GFP and GNRs. b Time evolution of the internal temperature in a C. elegans subjected to 800 nm laser irradiation (starting at t = 20 s). Adapted with permission Copyright 2013, American Chemical Society Donner et al. [4]

Luminescence Thermometry for in vivo Applications

271

3 Insect Models In an attempt to move to more complex organisms, Miyagawa et al. have developed a novel Eu3+ -based thermometers. They had two sources of signal: rhodamine 800 as reference and a polymer incorporating Eu-tris(dinaphthoylmethane)-bistrioctylphosphine oxide as a thermal sensing unit. The problem they tried to solve was the determination of heat produced by the muscle of a living beetle when activated during pre-flight preparation. While the literature had reported that an increment in temperature was to be expected, no careful numerical assessment of this increment had been done. The authors showed that their polymer successfully measured in vivo temperature variations through the muscle of the beetle. The ratio between the signals provided by Eu3+ and Rh800 was evaluated in 5 distinct regions of the muscle (Fig. 2a, b) and showed a trend similar to the one recorded by an infrared thermographic camera [5]. The same research group later improved such results by developing nanosheets with the same ions. They had better flexibility and transparency, enabling their attachment onto the uneven surfaces of the beetle without the aid of a glue (Fig. 2c). The Eu3+ -Rh800 stacked nanosheets were then used to record the endogenous heat production transfer between muscle fibers. Such a task would be impossible to be achieved through an infrared thermographic camera due to limitations in terms of spatial resolution. The results presented by Miyagawa lead to a deeper understanding of physiological processes taking place before, during, and after the beetle’s preflight preparation (Fig. 2d) [6].

4 Murine Models 4.1 Control of Photothermal Therapy When you combine two different materials within the nanoscale, a hybrid nanoparticle (HNP) is made. This type of NPs have attracted a great deal of attention due to the concomitant use of functionalities of their constituents, which, in turn, can result in multifunctional platforms [7, 8]. Zhu et al., for instance, produced compact HNPs capable of self-monitored heating [9]. The authors took profit from the core–shell technique to synthesize a structure such as the one shown in Fig. 3a, i.e. a core of Er3+ /Yb3+ co-doped NaLuF4 nanocrystal surrounded by a passive shell of undoped NaLuF4 covered with an external carbon shell. The core acted as a thermal sensor based on the spectral analysis of the Er3+ emission (bands centered at around 525 and 545 nm) [8, 10]. Fig. 3b shows the thermal dependence of the ratio between the fluorescence intensities at 545 and 525 nm, which allowed for a straightforward determination of temperature with a resolution of 0.5 °C. The carbon shell, on the other hand, acted as a heating unit when excited by a 730 nm laser. To demonstrate the potential of their HNPs, the authors incubated them in living cells, which were then subjected to irradiation at 980 and 730 nm in order to excite both the core and

272

E. Ximendes

Fig. 2 Study of a Dicronorrhina derbyana beetle. a Temporal thermal profile performed by an IR thermometer. b Corresponding temporal profiles of the normalized intensity ratio in the distinct regions of the muscle. c Bright-field image of a Dicronorrhina derbyana beetle. The triangle marks the location of the stacked nanosheets attached to the dorsal longitudinal muscle. d) Temperature mapping (three stimulations) of the muscle before (left), during (middle), and after (right) beetle’s preflight preparation using the stacked nanosheets. Adapted with permission Copyright 2016, American Chemical Society

the shell in order to achieve self-monitored heating. It was shown that the heating efficiency of the carbon shell was so high that it could increase the intracellular temperature up to 60 °C and, consequently, induce cell death.

Fig. 3 Temperature-controlled photothermal therapy with hybrid nanostructures. a Schematic representation of the HNPs designed by X. Zhu et al. for simultaneous in vivo thermal sensing and heating. b Calibration curve of the core sensing unit showing the temperature dependence of the intensity ratio R under excitation at 980 nm. Dots are experimental data and dashed line is a guide for the eyes. c Two photon excited emission spectra generated by the HNPs incorporated into a tumor in a nude mouse obtained after a 3-min-long 730 nm laser irradiation at three different power densities. Adapted with permission Copyright 2016, Nature Publishing Group

Luminescence Thermometry for in vivo Applications

273

Encouraged by such promising results, cancer cells were incubated with HNPs into nude mice to induce the growth of malignant tumors and, subsequently, perform their treatment. The emission spectra of intratumorally located HNPs corresponding to various 730 nm laser power densities are depicted in Fig. 3c. It becomes clear that the higher the 730 nm power density, the lower the emitted intensity at 545 nm, which, according to the HNPs calibration, suggests that a relevant intratumoral heating is taking place. The authors then demonstrated that it is possible to have precise control over photothermal treatments by profiting from the multifunctionality of HNPs. While certainly a very promising proof of concept, the main limitation of this work is the need of two different excitations. The second is the use of a luminescent sample emitting in the visible region, which, as we will discuss later, has a negative impact on the penetration depth that can be achieved. A different scenario emerged when it was reported what later came to be known as the infrared biological windows (BWs) [11]. They correspond to wavelength ranges in which the absorption and scattering of light in biological tissues are minimized. As such, the penetration depth of both excitation and emission could be maximized, and, in principle, one could think of applications going from the subcutaneous to even deeper levels. As a consequence, several luminescent thermometers operating in these wavelength ranges have been synthesized (Fig. 4). An example of acquisition of murine intratumoral temperature by means of thermally-sensitive PbS/CdS/ZnS QDs was demonstrated by del Rosal et al. [12]. Such samples were used as multifunctional agents (i.e., with therapeutic and thermal

Fig. 4 Number of works involving luminescence thermometry in the near infrared in the past two decades

274

E. Ximendes

sensing capabilities). The authors profited from the photothermal conversion efficiency of the QDs in such a way that with only one excitation laser beam (808 nm), the QDs could be simultaneously induced to emit light (from which the thermal dynamics of the system could be evaluated) and to generate heat (which would ablate the tumors). For such, the authors subjected the tumor-bearing mice to a 4-min long irradiation after making an intratumoral injection of a dispersion of the nanoparticles. While performing the treatment, an infrared camera was used to record the temperature-sensitive emission. Using then the calibration of the nanoparticles, the intratumoral temperature increment was determined (Fig. 5a). The authors have found that the injection site had a significantly higher temperature than the one of the surface by the end of each treatment (Fig. 5b). With such work, the authors then suggested that photothermal therapies could highly benefit from the feedback provided by luminescent thermometers. As discussed in previous chapters, depending on the ions and doping concentrations selected, rare-earth doped nanoparticles have the potential of possessing a good thermal sensitivity as well as high photothermal conversion efficiency. For the case of Nd3+ -doped NPs, when the concentration of Nd3+ ions is high, there is an increment in the absorbed pump power and a decrement in the fluorescence quantum yield. As consequence, heat is generated. Under such conditions, one is then tempted to find the proper balance between light and heat generation through the study Nd3+ content withing the NP. Carrasco et al. has exploited this concept with LaF3 :Nd3+ NPs [11– 14]. After a thorough investigation of how the heating efficiency and the emission intensity depended on the Nd3+ concentration, the authors were able to select 5.6 at% doped LaF3 :Nd3+ as the optimal value for achieving multifunctionality during an in vivo photothermal treatment. The authors have chosen the ratio between the emitted intensities at 865 and 885 nm as their thermometric parameter. The internal

Fig. 5 Temperature-controlled photothermal therapy with PbS/CdS/ZnS QDs and rare-earth doped NPs. a in vivo intratumoral temperature increment, as calculated from the QD emission quenching and surface temperature increment, as measured with a thermographic camera for different irradiation power densities. Adapted with permission Copyright 2016, Wiley–VCH. b Time evolution of the intratumoral temperature as obtained from the analysis of the luminescence of the intratumorally injected NPs and tumor surface temperature as measured by infrared thermal imaging. The evolution of the surface temperature of the control tumor is included for comparison. c Post-treatment size evolution of both treated and control tumors. The inset shows the optical image of the surface scar left at the tumor site 15 days after the therapy. Adapted with permission Copyright 2015, Wiley–VCH

Luminescence Thermometry for in vivo Applications

275

and surface temperatures of a tumor during treatment are shown in Fig. 5c as a function of the irradiation time. And once again it was a discrepancy between them has been observed, which only adds to the importance of an accurate sensing at the injection site for the minimization of side effects caused by extra heating. The success of LaF3 :Nd3+ NPs as self-monitored photothermal agents can be seen in Fig. 5d. While a valuable proof-of-concept, one needs to emphasize that the injection of the dispersion of NPs was made in the tumor itself. If one intends to be more general in terms of application (to deeper tumors, for instance), then compelling strategies need to be developed for targeting the nanostructures to tumors.

4.2 Control of Magnetic Hyperthermia A recent example of a nanocapsule composed of both magnetic and Ag2 S NPs to overcome the general limitations of control over magnetic hyperthermia (MHT) has been recently shown by the group of Daniel Jaque [15]. The newly produced nanocapsules acted as multimodal contrast agents under several imaging techniques (MRI, CT, photoacoustic and near-infrared fluorescence imaging) and could, therefore, benefit from all these modalities. To verify their full potential, the authors conceived an experiment for achieving thermally monitored in vivo MHT. For such, an intradermal injection of the nanocapsules was made in a CD1 mouse which had been previously anesthetized and placed inside a custom-built system (Fig. 6a). This system, in turn, had an optical window for the acquisition of the fluorescence image of the animal while the magnetic was being applied. The 808 nm excitation intensity was kept at a constant value of 10 mW cm−2 . Under such conditions, the laser-induced heating was minimized. The near-infrared fluorescence images of the animal obtained before and 5 min into the application of an AC magnetic field (20 kA m−1 , 140 kHz) showed a similar distribution of the nanocapsules (Fig. 6b). However, a quenching of the emission was observed within that timeframe, which indicated a local heating at the site of the intradermal injection. The dynamics of the near-infrared signal generated by the nanocapsules showed a relative decrease of approximately 15% (Fig. 6c). After switching off the AC magnetic field, the fluorescence signal gradually recovered its original value. By utilizing the calibration curve of the nanocapsules, the drop in fluorescence intensity was translated to temperature units. An increment of temperature close to 2.5 °C was estimated. The thermal resolution of the readout, in turn, was determined to be 0.2 °C. Contrary to the works described in the previous paragraphs, the dynamics of the intradermal temperature during the MHT treatment agreed well with the one measured with an infrared thermographic camera. However, such an agreement was expected due to the superficial nature of the intradermal injection. As a matter of fact, the small thickness of the dermis (5 °C) in rats. Figure 1 shows a non-comprehensive summary of the available data in brain temperature fluctuations caused by certain behaviours, external stimuli or drugs. While a more detailed analysis of brain temperature, its regulation mechanisms and its sensitivity to various stimuli is beyond the scope of this book chapter, review articles by Kiyatkin et al., Wang et al., and Bertolizio et al., are all excellent sources for further information on the topic [2, 3, 33, 34].

302

B. del Rosal

Fig. 1 Brain temperature fluctuates. Brain temperature changes observed under different external stimuli, behavioural changes and drug administration in animal models and humans. Data extracted from references [35–45]

2.2 Thermal Modulation of Neural Activity Neural stimulation is a well-established technique based on the local application of electric fields to modulate neural activity. It is widely used to address symptoms associated to neurological diseases and restore neural function [46, 47]. Temperature modulates the capacitance of the neuronal cell membrane and regulates the permeability of certain ionic channels, making it possible to use localised heating to modulate neural activity. This has gained interest in recent years as a contactless alternative to electrical stimulation methods to treat neurological disorders [48]. Electrical stimulation is an invasive process, since it requires the surgical implantation of electrodes, which have limited long-term biocompatibility [49]. The most researched contactless stimulation technique is infrared neural stimulation, which relies on infrared (1.45–2.2 µm) laser pulses to create a thermal transient that stimulates neural activity [50, 51]. Infrared stimulation can be used to suppress neural activity, too—a net temperature increase in the tissue caused by a highfrequency laser pulse train inhibits electrical activity [52]. Nanoparticle-mediated thermal stimulation (both photothermal and magnetothermal) can increase the selectivity of this technique and minimize off-target heating and potential tissue damage [4, 53]. Luminescence thermometry has potential to further develop these highly selective stimulation processes, since the temperature changes occurring during them are still not well-characterised.

2.3 Techniques to Study Brain Temperature Despite the physiological relevance of brain temperature and its potential applications in diagnosis and therapy, there is still a large knowledge gap regarding its fluctuations, distribution and role in disease development. Addressing that gap would require a

Contactless Luminescence Nanothermometry in the Brain

303

technique capable of probing brain temperature in live, preclinical animal models. All the data summarised earlier in this section have been obtained using invasive instruments (thermocouples or fiber-optic thermometers) that can only record point measurements, or magnetic resonance thermometry. Thermocouples and fiber-optic thermometers are very sensitive and can report temperature changes in real time, but they need to be placed in direct contact with the tissue, which leads to tissue damage upon insertion and prevents their application for long-term measurements [54, 55]. They are also unsuitable for studying the effect of some external stimuli (for example, laser light) as they can absorb part of the light and act as local heat sources [56]. In contrast, magnetic resonance thermometry is non-invasive and entirely contactless, but it cannot provide temperature information in real time and its spatial resolution (∼mm) is quite low compared to that that can be achieved with optical methods [57].

3 NIR-II Luminescence Thermometry of the Brain Luminescence nanothermometry combines several features that make it seemingly ideal for probing brain temperature: it is contactless, it is not limited to point measurements and can allow real-time sensing at high spatial resolution. While luminescence nanothermometry has gained traction in preclinical research, research on its application for brain thermometry is still very limited due to two the intrinsic challenges associated to fluorescence-based sensing in in vivo models. Delivering nanothermometers to a particular target area in the brain is technically complex and only small volumes (< 30 µL for mice) can be injected [58]. This complicates imaging at acceptable signal-to-noise ratios for accurate sensing, especially in deep brain locations.

3.1 Light-Tissue Interaction Light attenuation is one of the major limitations of any in vivo luminescence-based sensing; and is particularly critical in the case of the brain since luminescent probes will always be located at relatively large tissue depths. Both absorption and scattering contribute to light attenuation, although absorption is relatively low across both NIRI and NIR-II windows except for water absorption bands at 980 and 1450 nm [9]. Scattering, which is highly dependent on tissue properties and wavelength, is thus the main contributor to attenuation of light in the NIR. For almost all biological tissues, including all those that light must penetrate through to image fluorescent probes located in the brain—skin, subcutaneous tissue, cranial bone brain, as shown schematically—the scattering coefficient follows a decreasing trend with increasing wavelengths. Figure 2 shows the reduced scattering coefficients across the visible and NIR for several biological tissues across the visible and NIR.

304

B. del Rosal

Fig. 2 Wavelength depencence of tissue scattering. Reduced scattering coefficient in the 400– 1700 nm spectral range for various biological tissues including the brain and the cranial bone. Reproduced with permission [59]. Copyright 2017, Springer Nature

Mathematically, the scattering coefficients can be represented as: 

μs (skin) = 0.11λ−4 + 1.61λ−0.22 

μs (subcutaneous tissue) = 0.96λ−0.68 

μs (cranial bone) = 1.72λ−0.65 

μs (brain tissue) = 4.72λ−2.07 where the wavelength λ is expressed in µm and the reduced scattering coefficient in mm−1 [59]. Compared to other biological tissues, the brain has a very large scattering coefficient with a very pronounced wavelength dependence. This indicates that choosing fluorescent probes that emit well within NIR-II is critical for fluorescence imaging and sensing in the brain—even more so than for applications in other organs and tissues. To get a clearer picture of the impact the choice of wavelength will have on the fluorescence signal we can detect, we can look at the light attenuation due to scattering in a mouse model considering average tissue thicknesses—250 µm for the skin, 100 µm for the subcutaneous tissue and 340 µm for the cranial bone [60, 61]. If we use 808 nm light for optical excitation, 72% of the excitation light will be scattered before reaching the surface of the brain—and 52% of that light will be scattered within 100 µm into the brain. Virtually no excitation light would reach 1 mm within the brain—only 0.06% of the initial excitation intensity. For longer wavelengths, the attenuation due to scattering through the first three layers of tissue (skin, subcutaneous tissue, cranial bone) is not dramatically different, with 37% of

Contactless Luminescence Nanothermometry in the Brain

305

the excitation light reaching the surface of the brain. Reaching deeper within the brain becomes possible—28% of the light (instead of 52%) will be scattered within 100 µm into the brain, and about 4% of the initial excitation light could reach 1 mm within the brain. This highlights the importance of using NIR-II wavelengths for fluorescence imaging and sensing in the brain. The numbers above are only illustrative of that fact and should not be interpreted as fully accurate values. Absorption has been considered negligible for the calculations, which is not entirely accurate even though the scattering is responsible for most light attenuation happening in the infrared. Furthermore, the thickness of the different tissue layers is highly dependent on mouse age, strain and sex, so the numbers used for the calculations above only represent indicative values. NIR-II fluorescence has demonstrated potential for high-resolution transcranial imaging at relatively large tissue depths. Hong and coworkers used the NIR-II emission band of single-walled carbon nanotubes to image brain vasculature in mice, and could resolve very small (∼10 µm) blood vessels in the brain through the scalp and skull in mice at tissue depths up to 2.9 mm below the skin surface [62]. This suggests that NIR-II nanothermometers could be used for contactless sensing of brain temperature. While many NIR-II fluorescent NPs, including lanthanide-doped nanocrystals and semiconductor NPs of various compositions have a NIR-II temperature-sensitive emission [63–65], only Ag2 S NPs have been used for brain thermometry so far.

3.2 Ag2 S Nanoparticles in Brain Thermometry Ag2 S NPs combine several properties that make them good candidates for brain thermometry: their emission is centered at ∼1200 nm, their thermal sensitivity is relatively large (up to 4%/°C) and they lack highly toxic heavy metals (Pb, Cd) in their composition, unlike other NIR-II semiconductor NPs [66]. This has made them one of the most researched NIR-II fluorescent probes for deep-tissue in vivo imaging since they were first used for this purpose [67, 68]. Ag2 S NPs allow temperature sensing based on multiple parameters of their photoluminescence—the intensity and position of their PL emission peak and their radiative lifetime all change with temperature, as shown in detail in Fig. 3 [69, 70]. So far, only intensity-based thermometry has been applied to sensing brain temperature in vivo using luminescent nanothermometers. This technique is very easy to implement from an experimental point of view and requires no post-processing, which makes it ideal for cases when real-time sensing is required. It is however inherently limited to sensing temperature changes under a given excitation intensity—no actual temperature values can be extracted from the emission intensity. Our group used an intensity-based thermometry approach with Ag2 S NPs injected intracerebrally in a mouse model to detect sub-degree changes in brain temperature, achieving a thermal resolution of ± 0.2 °C [72]. We monitored the NIR-II emission of the NPs in real-time to detect local changes in brain temperature while simultaneously monitoring core and skin surface temperature using a rectal probe and a

306

B. del Rosal

Fig. 3 Ag2 S as NIR-II nanothermometers. a Emission spectra of Ag2 S NPs for temperatures in the 22–44 °C range, evidencing a temperature-induced quenching and peak redshift. b–d Temperature-related changes in peak position (b), intensity (c) and ratio between intensities at 1225 and 1175 nm (d), respectively. The thermal sensitivity is included for all three parameters. Reproduced with permission [71]. Copyright 2020, Wiley–VCH. e Radiative decay curves of Ag2S NPs for temperatures between 20 and 50 °C, from which the temperature-dependent lifetimes in f are calculated. Reproduced with permission [70]. Copyright 2022, Wiley-WCH

thermographic camera. We studied changes in brain temperature across a series of proof-of concept experiments which involved applying physical and pharmacological stimuli—whole-body cooling, drug-induced coma—known to cause changes in brain temperature (see Sect. 2). The results, summarised in Fig. 4, were consistent with previously reported data. In the case of whole-body cooling (Fig. 4a), we observed that the temperature of the brain dropped only 3 °C during the cooling process and returned its baseline value 30 min after active cooling stopped. Meanwhile, core temperature (measured with a rectal probe) and skin temperature (measured with a thermographic camera) dropped by almost 18 °C and took about 90 min to return to baseline values. Looking at the rate at which the temperature decreased during cooling (Fig. 4b), we see that the brain had a much slower cooling rate—not quite reaching -0.2 °C, whereas the skin cooled down up to three times faster. This is consistent with the existence of specific thermoregulation mechanisms for the brain, as reported previously [73]. In the case of the drug-induced coma, we observed that pentobarbital administration caused a decrease in brain temperature close to 1 °C, while skin temperature remained unchanged (Fig. 4c). This drop in temperature is also consistent with existing data, which attribute this change in temperature to the effect of barbiturates in brain metabolic and electrical activity [44].

Contactless Luminescence Nanothermometry in the Brain

307

Fig. 4 Contactless NIR-II brain thermometry. a Time evolution of core, skin and brain temperatures of a mouse during a whole-body cooling experiment. Stages I, II and III correspond to thermal stabilisation, whole-body cooling by ethanol application and return to baseline temperature. b Time derivative of skin and brain temperatures calculated from the data in a. c Temperature change caused by barbiturate administration in a mouse. The shaded regions correspond to three different patterns in the electroencephalogram. Reproduced with permission [72]. Copyright 2018, Wiley–VCH

4 Limitations and Challenges The results shown in the previous section indicate that NIR-II nanothermometry can measure brain temperature and overcome the limitations of currently available thermometry methods. For luminescence thermometry to become a valuable technique in neuroscience, future research in the field should address the limitations of the work described in Sect. 3.

4.1 Measurement Type and Accuracy Using a thermometry approach that enables measuring absolute temperatures is essential to get a full understanding of brain temperature distribution and fluctuations. Intensity-based measurements such as the one described in Sect. 3 have very limited applicability and only allow for recording temperature changes over short periods of time, during which the nanothermometer concentration can be considered constant [12]. Ratiometric luminescence thermometry could, in principle, address this issue without requiring any major changes to the imaging approach—only a fast-change filter wheel to capture fluorescence images in two different wavelength bands. Ratiometric thermometry is possible with Ag2 S NPs since their emission peak changes with temperature—the ratio between two different ranges of the emission band does change with temperature, too. However, the wavelength dependence of light-tissue interaction changes the shape of the emission band, which complicates using a ratiometric approach for temperature sensing in vivo [74]. As shown in Fig. 5a, Ag2 S

308

B. del Rosal

Fig. 5 Tissue interference in NIR-II thermometry. a Emission spectra of Ag2 S NPs as measured in aqueous solution and in the liver in vivo. b Temperature dependence of the radiative lifetime of Ag2S NPs through different thicknesses of tissue phantom. Reproduced with permission [70]. Copyright 2022, Wiley–VCH

NPs display very different emission band shapes depending on their location—in solution or within a tissue. This would render any in solution temperature calibration useless for in vivo ratiometric thermometry. Since light-tissue interaction does not affect the radiative lifetime of fluorescent nanothermometers in vivo, this is most likely the way to go for in vivo brain thermometry despite requiring a more complex setup for time-resolved detection. Jaque’s group recently applied Ag2 S NPs to lifetime-based thermometry in a mouse model of liver inflammation and observed no changes in the detected lifetime associated to changes in tissue depth (Fig. 5b) [70]. Beyond measurement accuracy, thermal resolution is another issue that may limit the applicability of luminescence nanothermometry in neuroscience. Since physiologically relevant temperature changes in the brain are often smaller than 1 °C, thermal accuracies equal or better than 0.1 °C would be required. Brighter NIRII nanothermometers would allow improved thermal accuracies thanks to better signal-to-noise ratios. Strategies based on wet chemistry or photochemistry have shown potential to generate Ag2 S NPs with enhanced quantum yields, although their application for in vivo thermometry remains unexplored [75, 76]. Besides material development, mathematical analysis tools could boost thermal resolution, as shown by Maturi et al. [77]. They applied multiple linear regression to multiparametric in vivo temperature data reported previously by Shen et al., who had simultaneously tracked changes in overall intensity, emission peak position and intensity ratio to achieve accurate in vivo temperature sensing [71]. The combination of multiparametric sensing and multiple linear regression produced a record thermal uncertainty of 0.05 K. This accuracy could allow precisely monitoring even very small fluctuations in brain temperature.

Contactless Luminescence Nanothermometry in the Brain

309

4.2 Delivering Nanoparticles to the Brain A major advantage of luminescence nanothermometry compared to thermocouples and fiber-optic thermometers is its contactless nature—that does not necessarily make it non-invasive, though. The work described in Sect. 3 relied on intracerebral injection to deliver the nanothermometers to a target area in the brain. This process involves drilling a hole in the skull of the animal—which is immobilised in a stereotaxic frame—and slowly injecting a small volume of NPs dispersed in saline solution. However, direct injection remains the simplest way to reach target areas in the brain, since it entirely bypasses the blood–brain barrier [78]. NPs injected intravenously are not only susceptible to uptake by the reticuloendothelial system, but also incapable of cross an intact blood–brain barrier. In that case, a method to disrupt blood–brain barrier permeability (ultrasound or laser-induced heating) would be required in conjunction to systemic injection [79]. Specific antibodies or peptides conjugated to the NP surface can also allow NPs to cross the blood–brain barrier [80]. Intrathecal and intranasal routes—both of which bypass the blood–brain barrier—also allow nanoparticle delivery to the brain [81, 82]. The minimally invasive nature of intranasal administration makes it, in principle, particularly attractive for nanothermometer delivery.

5 Conclusion and Future Perspectives Brain temperature has attracted increasing interest as a relevant clinical parameter. However, many questions remain around brain temperature, including its role in neurological diseases. This lack of knowledge is mostly due to the lack of techniques that can probe brain temperature with sufficient spatiotemporal resolution in preclinical models. The results outlined in this book chapter show the potential of NIR-II nanothermometry to measure brain temperature in a contactless manner, in real time and at high spatial resolutions. Currently, luminescence nanothermometry is unable to detect temperature changes of physiological significance in the brain with sufficient accuracy, although research efforts in material development and mathematical analysis will most likely address that limitation in the near future. Sensing temperature in deep-brain locations may still be particularly challenging due to the large extinction coefficient of the brain. Research efforts focused on developing bright nanothermometers with excitation and emission bands in the NIR-II will be essential. Further developments in imaging technology and analysis could expand the application of NIR-II nanothermometry to freely moving animals. The fundamental neuroscience knowledge gained with NIR-II nanothermometry could ultimately lead to new diagnostic and therapeutic strategies for neurological disorders.

310

B. del Rosal

References 1. Organization, W.H.: Neurological Disorders: Public Health Challenges. World Health Organization (2006) 2. E.A. Kiyatkin, Brain temperature homeostasis: physiological fluctuations and pathological shifts. Front. Biosci. 15, 73 (2010) 3. H. Wang et al., Brain temperature and its fundamental properties: a review for clinical neuroscientists. Front. Neurosci. 8, 307 (2014) 4. E. Colombo et al., Nanoparticles: a challenging vehicle for neural stimulation. Front. Neurosci. 10, 105 (2016) 5. D. Jaque et al., Fluorescent nanothermometers for intracellular thermal sensing. Nanomed. 9(7), 1047–1062 (2014) 6. Tanimoto, R., et al.: Detection of temperature difference in neuronal cells. Scientific Rep. 6 (2016) 7. D.A. Simpson et al., Non-neurotoxic nanodiamond probes for intraneuronal temperature mapping. ACS Nano. 11(12), 12077–12086 (2017) 8. J.S. Donner et al., Imaging of Plasmonic heating in a living organism. ACS Nano. 7(10), 8666–8672 (2013) 9. A. Bashkatov et al., Optical properties of human skin, subcutaneous and mucous tissues in the wavelength range from 400 to 2000 nm. J. Phys. D Appl. Phys. 38(15), 2543 (2005) 10. B. del Rosal et al., In vivo autofluorescence in the biological windows: the role of pigmentation. J. Biophotonics 9(10), 1059–1067 (2016) 11. S. Diao et al., Biological imaging without autofluorescence in the second near-infrared region. Nano Res. 8(9), 3027–3034 (2015) 12. B. del Rosal et al., In vivo luminescence nanothermometry: from materials to applications. Adv. Optical Mat. 5(1), 1600508 (2016) 13. X. Zhu et al., Temperature-feedback upconversion nanocomposite for accurate photothermal therapy at facile temperature. Nat. Commun. 7, 10437 (2016) 14. E. Carrasco et al., Intratumoral thermal reading during photo-thermal therapy by multifunctional fluorescent nanoparticles. Adv. Func. Mater. 25(4), 615–626 (2015) 15. E.C. Ximendes et al., Unveiling in vivo subcutaneous thermal dynamics by infrared luminescent nanothermometers. Nano Lett. 16(3), 1695–1703 (2016) 16. H.D. Santos et al., In vivo early tumor detection and diagnosis by infrared luminescence transient nanothermometry. Adv. Func. Mater. 28(43), 1803924 (2018) 17. C.W. Kuzawa et al., Metabolic costs and evolutionary implications of human brain development. Proc. Natl. Acad. Sci. 111(36), 13010–13015 (2014) 18. J.D. Hardy, Physiology of temperature regulation. Physiol. Rev. 41(3), 521–606 (1961) 19. Swan, H.: Thermoregulation and Bioenergetics: Patterns for Vertebrate Survival (Book), 442 p. American Elsevier Publishing Co., Inc., New York (1974) 20. L. Nybo, N.H. Secher, B. Nielsen, Inadequate heat release from the human brain during prolonged exercise with hyperthermia. J. Physiol. 545(2), 697–704 (2002) 21. M. Zhu, et al., How the body controls brain temperature: the temperature shielding effect of cerebral blood flow. J. Appl. Physiol. 101(5), p. 1481–1488 (2006) 22. M. Zhu, J.J. Ackerman, D.A. Yablonskiy, Body and brain temperature coupling: the critical role of cerebral blood flow. J. Comp. Physiol. B 179(6), 701–710 (2009) 23. B. Nielsen, L. Nybo, Cerebral changes during exercise in the heat. Sports Med. 33(1), 1–11 (2003) 24. A. Bouchama, J.P. Knochel, Heat stroke. N. Engl. J. Med. 346(25), 1978–1988 (2002) 25. N. Malamud, W. Haymaker, R.P. Custer, Heat stroke. a clinico-pathologic study of 125 fatal cases. Milit. Surg. 99(5), 397–449 (1946) 26. H.S. Sharma, P.J. Hoopes, Hyperthermia induced pathophysiology of the central nervous system. Int. J. Hyperthermia 19(3), 325–354 (2003) 27. M.W. Dewhirst et al., Basic principles of thermal dosimetry and thermal thresholds for tissue damage from hyperthermia. Int. J. Hyperthermia 19(3), 267–294 (2003)

Contactless Luminescence Nanothermometry in the Brain

311

28. S.M. Thompson, L.M. Masukawa, D.A. Prince, Temperature dependence of intrinsic membrane properties and synaptic potentials in hippocampal CA1 neurons in vitro. J. Neurosci. 5(3), 817–824 (1985) 29. N.R. Hardingham, A.U. Larkman, The reliability of excitatory synaptic transmission in slices of rat visual cortex in vitro is temperature dependent. J. Physiol. 507(1), 249–256 (1998) 30. E.A. Kiyatkin, H.S. Sharma, Permeability of the blood–brain barrier depends on brain temperature. Neurosci. 161(3), 926–939 (2009) 31. K.N. Fountas et al., Intracranial temperature. Neurocrit. Care 1(2), 195–199 (2004) 32. J. Delgado, T. Hanai, Intracerebral temperatures in free-moving cats. Am. J. Physiol.-Legacy Cont. 211(3), 755–769 (1966) 33. G. Bertolizio, L. Mason, B. Bissonnette, Brain temperature: heat production, elimination and clinical relevance. Pediatr. Anesth. 21(4), 347–358 (2011) 34. H. Wang et al., Thermal regulation of the brain-an anatomical and physiological review for clinical neuroscientists. Front Neurosci. 9, 528 (2015) 35. E.A. Kiyatkin, Brain temperature fluctuations during physiological and pathological conditions. Eur. J. Appl. Physiol. 101(1), 3–17 (2007) 36. S. Onitsuka et al., Ice slurry ingestion reduces human brain temperature measured using noninvasive magnetic resonance spectroscopy. Sci. Rep. 8(1), 1–7 (2018) 37. X.-F. Yang, J.H. Chang, S.M. Rothman, Intracerebral temperature alterations associated with focal seizures. Epilepsy Res. 52(2), 97–105 (2002) 38. Y. Takeda et al., Effects of pharyngeal cooling on brain temperature in primates and humans: a study for proof of principle. J. Am. Soc. Anesthesiol. 117(1), 117–125 (2012) 39. R.A. Bola, E.A. Kiyatkin, Brain temperature effects of intravenous heroin: State dependency, environmental modulation, and the effects of dose. Neuropharm. 126, 271–280 (2017) 40. M.O. Baud, P.J. Magistretti, J.M. PETIT, Sustained sleep fragmentation affects brain temperature, food intake and glucose tolerance in mice. J. Sleep Res. 22(1), 3–12 (2013) 41. K. Yamada et al., Moyamoya patients exhibit higher brain temperatures than normal controls. NeuroReport 21(13), 851–855 (2010) 42. J.-M. Ramirez, L.P. Folkow, A.S. Blix, Hypoxia tolerance in mammals and birds: from the wilderness to the clinic. Annu. Rev. Physiol. 69, 113–143 (2007) 43. A. Pietroboni et al., Brain temperature in multiple sclerosis. Mult. Scler. J. 20(7), 894–896 (2014) 44. E.A. Kiyatkin, P.L. Brown, Brain and body temperature homeostasis during sodium pentobarbital anesthesia with and without body warming in rats. Physiol. Behav. 84(4), 563–570 (2005) 45. E. Kiyatkin, R. Mitchum Jr., Fluctuations in brain temperature during sexual interaction in male rats: an approach for evaluating neural activity underlying motivated behavior. Neuroscience 119(4), 1169–1183 (2003) 46. J.S. Perlmutter, J.W. Mink, Deep brain stimulation. Annu. Rev. Neurosci. 29, 229–257 (2006) 47. R.L. Testerman, M.T. Rise, P.H. Stypulkowski, Electrical stimulation as therapy for neurological disorders. IEEE Eng. Med. Biol. Mag. 25(5), 74–78 (2006) 48. S. Luan, et al., Neuromodulation: present and emerging methods. Front. Neuroeng. 7(27) (2014) 49. C. Marin, E. Fernández, Biocompatibility of intracortical microelectrodes: current status and future prospects. Front. Neuroeng. 3, 8 (2010) 50. J.D. Wells et al., Application of infrared light for in vivo neural stimulation. J. Biomed. Opt. 10(6), 064003 (2005) 51. J.M. Cayce et al., Infrared neural stimulation of primary visual cortex in non-human primates. Neuroimage 84, 181–190 (2014) 52. A.R. Duke et al., Transient and selective suppression of neural activity with infrared light. Sci. Rep. 3, 2600 (2013) 53. Y. Wang, L. Guo, Nanomaterial-enabled neural stimulation. Front. Neurosci. 10 (2016) 54. S. DeBow, F. Colbourne, Brain temperature measurement and regulation in awake and freely moving rodents. Methods 30(2), 167–171 (2003)

312

B. del Rosal

55. S. Musolino et al., Portable optical fiber probe for in vivo brain temperature measurements. Biomed. Opt. Express 7(8), 3069–3077 (2016) 56. F. Manns et al., In situ temperature measurements with thermocouple probes during laser interstitial thermotherapy (LITT): quantification and correction of a measurement artifact. Lasers Surg. Med.: Off. J. Am. Soc. Laser Med. Surg. 23(2), 94–103 (1998) 57. V. Rieke, K.B. Pauly, MR thermometry. J. Magnetic Res. Imag. 27(2), 376–390 (2008) 58. S. Shimizu, Chapter 32–Routes of administration, in The Laboratory Mouse. ed. by H.J. Hedrich, G. Bullock (Academic Press, London, 2004), pp.527–542 59. G. Hong, A.L. Antaris, H. Dai, Near-infrared fluorophores for biomedical imaging. Nat. Biomed. Eng. 1, 0010 (2017) 60. K. Calabro et al., Gender variations in the optical properties of skin in murine animal models. J. Biomed. Opt. 16(1), 011008 (2011) 61. H. Soleimanzad, H. Gurden, F. Pain, Optical properties of mice skull bone in the 455-to 705-nm range. J. Biomed. Opt. 22(1), 010503 (2017) 62. G. Hong et al., Through-skull fluorescence imaging of the brain in a new near-infrared window. Nat. Photonics 8, 723 (2014) 63. A. Nexha et al., Lanthanide doped luminescence nanothermometers in the biological windows: strategies and applications. Nanoscale 13(17), 7913–7987 (2021) 64. Q. Dai et al., Size-Dependent Temperature Effects on PbSe Nanocrystals. Langmuir 26(13), 11435–11440 (2010) 65. B. del Rosal et al., Infrared-emitting QDs for thermal therapy with real-time subcutaneous temperature feedback. Adv. Func. Mater. 26(33), 6060–6068 (2016) 66. Y. Shen et al., Perspectives for Ag2S NIR-II nanoparticles in biomedicine: from imaging to multifunctionality. Nanoscale 11(41), 19251–19264 (2019) 67. Y. Zhang et al., Ag2S quantum dot: a bright and biocompatible fluorescent nanoprobe in the second near-infrared window. ACS Nano. 6(5), 3695–3702 (2012) 68. P. Jiang et al., Water-soluble Ag2S quantum dots for near-infrared fluorescence imaging in vivo. Biomat. 33(20), 5130–5135 (2012) 69. D. Ruiz et al., Ag/Ag2S nanocrystals for high sensitivity near-infrared luminescence nanothermometry. Adv. Func. Mater. 27(6), 1604629 (2017) 70. Y. Shen et al., Reliable and remote monitoring of absolute temperature during liver inflammation via luminescence-lifetime-based nanothermometry. Adv. Mater. 34(7), 2107764 (2022) 71. Y. Shen et al., Ag2S nanoheaters with multiparameter sensing for reliable thermal feedback during in vivo tumor therapy. Adv. Func. Mater. 30(49), 2002730 (2020) 72. B. del Rosal et al., In vivo contactless brain nanothermometry. Adv. Func. Mater. 28(52), 1806088 (2018) 73. H. Wang et al., Thermal regulation of the brain—an anatomical and physiological review for clinical neuroscientists. Front. Neurosci. 9, 528 (2016) 74. Y. Shen et al., In vivo spectral distortions of infrared luminescent nanothermometers compromise their reliability. ACS Nano. 14(4), 4122–4133 (2020) 75. H.D. Santos et al., Ultrafast photochemistry produces superbright short-wave infrared dots for low-dose in vivo imaging. Nat. Commun. 11(1), 1–12 (2020) 76. I.Z. Gutierrez, et al., Boosting the Near-Infrared Emission of Ag2S Nanoparticles by a Controllable Surface Treatment for Bioimaging Applications. ACS Applied Materials & Interfaces (2022) 77. F.E. Maturi et al., Going above and beyond: a tenfold gain in the performance of luminescence thermometers joining multiparametric sensing and multiple regression. Laser Photonics Rev. 15(11), 2100301 (2021) 78. S. Pasha, K. Gupta, Various drug delivery approaches to the central nervous system. Expert Opin. Drug Deliv. 7(1), 113–135 (2010) 79. J. Stockwell et al., Novel central nervous system drug delivery systems. Chem. Biol. Drug Des. 83(5), 507–520 (2014) 80. J. Xie et al., Nanomaterial-based blood-brain-barrier (BBB) crossing strategies. Biomat. 224, 119491 (2019)

Contactless Luminescence Nanothermometry in the Brain

313

81. K.T. Householder et al., Fate of nanoparticles in the central nervous system after intrathecal injection in healthy mice. Sci. Rep. 9(1), 12587 (2019) 82. S.U. Islam et al., Intranasal delivery of nanoformulations: a potential way of treatment for neurological disorders. Molecules 25(8), 1929 (2020)

Optical Trapping of Luminescent Nanothermometers Lucía Labrador-Páez and Patricia Haro-González

Abstract Optical trapping is a remote non-invasive tool that generates optical forces by means of tightly focused light. This technique offers the opportunity to isolate, move at will, and monitor single nanoparticles such as for example luminescent nanothermometers. In this chapter we introduce the fundamentals of optical trapping and its advantages for luminescence nanothermometry applications. Then, we present the latest advances in the optical trapping of lanthanidedoped inorganic nanoparticles and metallic plasmonic nanoparticles for luminescent nanothermometry applications.

1 Introduction Optical trapping is a tool that allows to remotely manipulate single particles by means of light. Arthur Ashkin developed this technique in the 1980s and got awarded the Nobel prize in Physics in 2018 for it. The main application field of optical tweezers is biomedicine, where molecules, organelles, and cells can be manipulated and studied [1, 2]. Inorganic particles can be trapped too [3]. The isolation of a single luminescent nanothermometer by an optical trap will allow to measure temperature locally, in a less invasive way, and reducing the luminescence background. In this chapter, we will introduce the optical trap as a tool and explore how it has been used in the field of luminescent thermometry, i.e., when a single temperature-sensitive luminescent particle is optically trapped and used as local contactless thermometer.

L. Labrador-Páez Physical Chemistry Department, Chemical Sciences Faculty, Complutense University of Madrid, 28040 Madrid, Spain e-mail: [email protected] P. Haro-González (B) Nanomaterials for Bioimaging Group, Material Physics Department, Autonomous University of Madrid, 28048 Madrid, Spain e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_10

315

316

L. Labrador-Páez and P. Haro-González

2 Fundamentals of Optical Trapping By focusing a laser beam with a high numerical aperture (NA) objective, it is possible to form an optical trap. If there is a particle moving in its surroundings, it can experience an optical force that attracts it and holds it into the optical trap. These optical forces appear due to a momentum exchange between the focused laser beam and the trapped particle via a scattering process. The optical force exerted on the particle can be decomposed into two components: the scattering force (F scat ), which pushes the particle in the direction of propagation of the beam, and the gradient force (F grad ) that attracts the particle towards regions of higher intensity. For stable optical trap, the axial gradient force should match the scattering force [3–5]. For a full understanding of the optical forces involved in a trapping experiment, it is necessary to solve the Maxwell’s equations with the appropriate boundary conditions. The Lorenz-Mie Theory was the first step into that direction. It describes the scattering of a plane wave by a spherical particle for arbitrary particle size, refractive index, and wavelength. However, the Lorenz-Mie Theory cannot describe the typical Gaussian laser beam shape. The calculation of optical forces of Gaussian beams and the Generalized Lorenz-Mie Theory can achieve arbitrarily shaped beams. For spherical particles, the problem is usually solved by approximations, which depend on the radius of the object used in the experiment and either the wavelength or the waist size of the trapping beam. When the size of the particle (a) is much smaller than the wavelength of the beam (a > λ), i.e. in the Mie regime, the ray optics model provides a simple and intuitive description of these interactions. In the Rayleigh regime (a > λ), the force exerted by the laser beam on a spherical particle is computed using the ray optics approach. When the beam interacts with the particle, rays change their direction according to the laws of geometrical optics. The difference between the refractive index of the particle (n) and the medium (nm ), is the main factor determining the magnitude of the optical forces. The scattering and gradient forces that suffers the particle in a homogeneous and non-dispersive medium are given by:   T 2 [cos(2θi − 2θr ) + R cos(2θi )] nP 1 + R cos(2θ) − Fscat = c 1 + R 2 + 2R cos(2r)    T 2 [sin(2θi − 2θr ) + R cos(2θi )] nP R sin(2θ) − Fgrad = c 1 + R 2 + 2R cos(2r) 

(3)

(4)

where P is the laser power and T and R are the polarization-dependent Fresnel coefficients for transmission and reflection, respectively. θi and θr are the incidence and reflexion angles, and c is the speed of the light in vacuum. For an unpolarized light, the Fresnel coefficients are considered   as the average of the s and p polarizations   i.e., R = Rs + R p /2 and T = Ts + T p /2. When the particle size is comparable to the trapping light wavelength, (a~λ), neither the ray optics nor the point-dipole approach are valid. The full electromagnetic field equations solution of the problem is required. Most objects that are useful or interesting in optical trapping, in practice, tend to fall into this intermediate size range. For example, microspheres used alone or as handles to manipulate other objects are typically in this range, which is the same size range as biological specimens that can be trapped directly, e.g., bacteria, yeast, and organelles of larger cells.

3 How the Field Started The application of optical trapping for luminescent thermometry began with a work in 2012 where Haro-González et al. [6] demonstrated that a lanthanide-doped luminescent microparticle was both temperature sensitive and manipulable by an optical trap. Shortly after, in 2013, they assessed the heating produced by an optical trap in the solvent and in an optically trapped cell by measuring the temperature-sensitive luminescence of non-trapped quantum dots dispersed in the medium [7]. Later, in 2015, the first applications of optically trapped thermometers were found. Pauzauskie´s group and collaborators managed to characterize the heating and optical refrigeration of a lanthanide-doped optically trapped microcrystal by its own temperaturedependent luminescence [8]. The same year, the field was expanded to the use of metal nanoparticles. Andres-Arroyo et al. [9] estimated the heat produced by an optically trapped gold nanoparticle by measuring its scattering spectrum combining

318

L. Labrador-Páez and P. Haro-González

dark field spectroscopy and optical trapping techniques. In 2016, the optical trapping of luminescent thermometers first found an application in biomedicine. The heat profile next to a cell undergoing a photothermal treatment was measured by the emission of an optically trapped lanthanide-doped upconversion nanoparticle by Rodriguez-Sevilla et al. [10]. By then, the field was finally stablished and the number of published works increased.

4 Optimization of the Optical Trapping of Luminescent Nanothermometers Any particle with a refractive index greater than that of the surrounding medium (see above Section Fundamentals of Optical Trapping) and a non-negligible polarizability can be optically trapped. However, in the case of nanoparticles there are some difficulties if the exerted optical forces are not high enough to counteract the destabilizing effects such as thermal fluctuations or being the scattering component of the optical force larger than the gradient one [11, 12]. There are two strategies to optimize the optical manipulation of nanoparticles: the reduction of the optical trap volume and the modification of the surface of the nanoparticle. The nanoparticle surface can be modified to increase the optical forces, as the magnitude of the optical forces applied on nanoparticle depends on the ratio of the refractive index of the particle to the refractive index of the medium. It depends on the particle polarizability, which is intimately related to its surface charge. Nanoparticles present a superficial charge, which interacts with the solvent’s ions. These charges distribute around the particle forming an electric double layer characterized by the ζ potential. It has been experimentally demonstrated that the optical forces acting on nanoparticles depend more strongly on the electrostatic properties than on their volume [13, 14]. In addition, surface modification of the nanothermometers can enhance their luminescence properties and their colloidal stability, while providing the possibility of subsequent bioconjugation. Several methods for surface modification have been reported [15, 16]. An alternative strategy to optimize optical trapping forces is based on the reduction of the optical trap volume to increase the electric field gradient. Near-field techniques have offered a significant increase in trapping efficiency and improved the thermal stability of nanoparticles, covering the drawbacks of conventional tweezers. Some examples are plasmonic tweezers, slot waveguides, photonic crystal resonators, or photonic nanojet [11, 17, 18]. Among these near-field techniques, photonic nanojet has demonstrated to be easily implemented in optical trapping as it does not require complex nanofabrication processes nor setup customization. Under certain conditions, microspheres (acting as microlenses) can produce a narrower beam with a waist under the Abbe diffraction limit called a photonic nanojet. This reduction of the optical trap size facilitates the optical manipulation and detection of single nanoparticles [11].

Optical Trapping of Luminescent Nanothermometers

319

5 Desing of the Experimental Setup Figure 1a schematically represents a basic optical trapping setup, which is formed by the optical trap and the imaging and particle detection devices.

5.1 The Optical Trap High NA objective lenses are the basis of optical traps (see Fig. 1a), as they ensure small trap size (see Eq. 5) and, thus, large optical force [3]. Single-mode Gaussian lasers will ensure the smallest focus diameter [4]. Moreover, infinity corrected microscope objective lenses will facilitate the imaging of the optical trap [20].

Fig. 1 Basic single-beam optical tweezers setup for luminescent temperature nanosensors trapping. a Schematic representation of a basic setup. b Position detection by means of a quadrant photodiode (QPD) in the case of particles larger that trapping wavelength. Reprinted with permission from [3]. Copyright 2017 The Royal Society of Chemistry. c Position detection by means of a QPD by back focal plane interferometry (for particles smaller than trapping wavelength). Reprinted with permission from [19]. Copyright 2016 Elsevier. d Detail of setup modification for the detection of the luminescence generated by the trapped particle. Reprinted with permission from [3]. Copyright 2017 The Royal Society of Chemistry

320

L. Labrador-Páez and P. Haro-González

d=

2λ πNA

(5)

When selecting the optical trapping wavelength, its effect on the optical trap size must be considered (see Eq. 5) [21, 22]. Moreover, avoiding water absorption bands when choosing the trap laser wavelength will reduce the heating produced by the optical trap laser absorption in experiments in water dispersion or biological media [23]. Wavelengths around 800, 830, and 1064 nm are the best candidates for applications of optical trapping in nanomedicine [23–25]. Laser power will also increase the induced heating. The optical force is linearly proportional with laser power, but an increase in the trapping power could also lead to a temperature loading [24, 25]. The particularities of each experiment will determine the proper choice of the trapping wavelength and laser power. Most temperature sensing experiments can be performed by using a single optical trap that will manipulate the nanothermometer and excite its luminescence (see Fig. 1d). However, multiple traps will make more complex experiments possible [26]. Multiple traps can be created, for instance, by merging different laser sources in one objective, by using time-shared traps, by splitting a laser beam into two beams with orthogonal polarization, and by modulating a laser beam by means of interferometric patterns or holographic liquid crystals [4, 20, 27–29]. The relative position between the optical trap and the sample must be controlled. Piezoelectric stages allow controlled movement of the sample relative to the laser (see Fig. 1a). Alternatively, acousto-optic deflectors can tilt the trapping beam, which results in a displacement of the trap position relative to the sample [4, 28, 30–32].

5.2 Imaging and Nanoparticle Detection All trapping studies rely on tracking and/or imaging of the trapped particle. Back focal plane interferometry can provide the position of an optically trapped nanoparticle from the interference between the light scattered and the trapping beam (see Fig. 1) [4, 28, 33]. A quadrant photodiode located in the back focal plane of the objective lens collimating the outgoing laser light is required (see Fig. 1). Tracking the trajectory of the trapped nanoparticles by simply video imaging is possible for luminescent ones. However, a mass finding algorithm and a pixel-real distance calibration are required [28]. Since high frame rate cameras are needed for this purpose, intense illumination of the trap and optimized luminescence quantum yield are convenient to achieve enough contrast. The acquisition of the luminescence of the trapped nanosensor is commonly performed by using the trapping objective lens as collecting lens for the luminescence. Then the collected light is spectrally filtered, detected by a camera, and, if required, analyzed in real time by a spectrophotometer (see Fig. 1d).

Optical Trapping of Luminescent Nanothermometers

321

6 Luminescence Thermometry with Optically Trapped Dielectric Nanoparticles 6.1 Optical Trapping of Luminescent Nanothermometers Among dielectric nanoparticles, only lanthanide-doped luminescent nanoparticles have been employed successfully for this type of experiments until now. Lanthanidedoped nanoparticles usually consist of fluoride or oxide materials as host matrices and lanthanide ions as active emitters. They can act as luminescent nanothermometer due to their temperature-dependent luminescent properties [34, 35]. Commonly, fluorescence intensity ratio technique is used for temperature measurement using lanthanide-doped particles [36, 37]. This method is based in comparing the intensity of two adjacent emission bands originated by two energy levels close in energy. These emission bands are thermalized as they can have high probability of undergoing non-radiative transitions that redistribute the population between them. One of the most used lanthanide ions for thermal sensing is Er3+ (see Fig. 2a and b). The local temperature is extracted from the analysis of the emission from its thermally coupled states: 4 S3/2 and 2 H11/2 . Er3+ -doped microparticles have been used for thermometry in optical trapping [10]. However, the relatively weak forces caused by the optical trapping of dielectric nanoparticles result in some limitations. Any strong interaction will usually override the weak forces in the optical trap, resulting in the particles being removed out of the trap during manipulation. To overcome these limitations, some strategies are based on trapping bigger clusters of nanoparticles and on the construction of microspheres containing nanoparticles in its interior, such as show in Fig. 2e–f [26, 37]. However, since the spatial resolution of temperature measurement is dependent of the size of the trapped particle, it is desirable to trap the smallest possible particles. One of the smallest single nanoparticle trapped to measure temperature has a size of ~150 nm [37]. Figure 2c shows the thermal image of a cluster of gold nanorods on a glass substrate imaged with the laser-trapped Er2 O3 luminescent nanothermometer. The white dotted line on Fig. 2c shows the location where a temperature profile shown in Fig. 2d is selected. These results provide a spatial resolution around the size of the nanoparticle. Other luminescence properties are temperature-dependent and can be used to measure temperature by means of optically trapped sensors [6, 23, 36]. However, most examples are related to microsensors due to the limitation described before [38].

322

L. Labrador-Páez and P. Haro-González

Fig. 2 a Simplified energy level diagram showing the temperature sensitive energy levels, 2 H11/2 and 4 S3/2 , and the corresponding excitation and emission processes. b Emission spectra of Er3+ ions at two different temperatures. The intensity ratio I 1 /I 2 is highly temperature-sensitive. Reprinted with permission from [22]. Copyright 2016 John Wiley and Sons. c Thermal image of a hot gold nanostructures cluster obtained by scanning the optically trapped Er2 O3 nanothermometer. The white dotted line represents the place where the temperature profile shown in (d) is located. The solid red line is a Gaussian fit to the temperature profile. Reproduced with permission from [37]. Copyright 2016 Springer Nature. e–f Microscopic images of composite microsphere containing the luminescent nanoparticles in e bright field, and f upconverted emission under excitation at 980 nm. Reprinted with permission from [26]. Copyright 2017 American Chemical Society

6.2 Spinning of Dielectric Nanoparticles for Temperature Sensing While optical trapping relies on the transfer of linear momentum from light to the particle, rotation may also be engineered by transferring angular momentum from light to the particle, which is transformed into mechanical angular momentum. This

Optical Trapping of Luminescent Nanothermometers

323

Fig. 2 (continued)

can be either spin angular momentum (associated with the light’s circular polarization) or orbital angular momentum (associated with inclined wavefronts of the light) [39, 40]. Particles tend to align with their more polarizable axis parallel with the electric field to minimize the potential energy, so their orientation may be manipulated by trapping them using polarized light. It has been demonstrated how single laser beams cannot only trap but also induce rotation of birefringent materials (dielectric) or those with anisotropic polarizability, when it is trapped by a circularly polarized trapping light [38, 41]. The lag in the rotation due to the viscosity of the medium where the nanoparticles are dispersed may be a good indicator of changes in temperature, if the temperature dependence of the viscosity of the medium is known [42–45]. Thus, the rotational dynamics of the particle can be used to determine the temperatureinduced change through the viscosity of the local environment, which decreases in a non-linear way as the temperature rises. This method has been reported to be able to determine the temperature change with sub-degree accuracy, which is more than one order of magnitude more accurate than measurements done in the same experimental conditions with luminescence-based thermometric techniques by using microparticles [38].

7 Thermometry with Optically Trapped Metal Nanoparticles 7.1 Optical Trapping of Gold Nanoparticles The optical trapping forces experienced by metallic nanoparticles are usually larger than for dielectric nanoparticles of the same size due to their larger polarizability. It is due to the coupling of the electric field of light and the surface plasmons, i.e. the

324

L. Labrador-Páez and P. Haro-González

collective resonant oscillations of the electrons in the conduction band of the metal [42]. Incident light matching the plasmon resonance wavelength may lead to enhanced light absorption (and subsequent light-to-heat conversion), to Rayleigh light scattering (determining the apparent color of the metallic nanoparticle and generating large radiation pressure), and to intense field intensity enhancement near de surface (enabling antenna effect) [42]. Metal nanoparticles tend to be much smaller than the wavelength of the optical trap, so they can be treated as a point dipole (see Section Fundamentals of Optical Trapping). Metal nanoparticles show a large and wavelength-dependent polarizability, which also depends on the particle shape, volume, frequency-dependent dielectric response of the metal, and the refractive index of the environment. Therefore, the optical gradient force results large and attractive when the metallic particle is trapped by light with lower frequency than the plasmon resonance and repulsive if it has a higher frequency. The radiation pressure increases strongly with the frequency of the light and around the wavelength of the plasmon resonance [42]. Most works dealing with temperature measuring and plasmonic nanoparticles characterize the temperature around the plasmonic particle using temperaturesensitive molecules or other nanoparticles [46]. However, in this section we focus on temperature measuring methods where the temperature reading is performed by the plasmonic nanoparticle itself. When employing plasmonic nanoparticles as nanothermometers, there is the risk that their large light-to-heat conversion may alter the temperature reading by artifactually increasing the temperature of the sensor itself. Moreover, the temperature of a plasmonic nanoparticle deposited over a substrate may be different depending on the material of the substrate, which may act as a heat dissipator and have a particular variation of refractive index with temperature. For that reason, optically trapping and isolating plasmonic nanoparticles increases the reliability of temperature measurements [47].

7.2 Spinning of Metal Nanoparticles for Temperature Sensing In the case of the spinning of plasmonic nanomaterials, the angular momentum is in general maximized for trapping wavelengths close to the plasmon resonance. Furthermore, anisotropic plasmonic nanomaterials rotate faster than spherical plasmonic materials of the same volume, as they experience an additional torque due to the large scattering at wavelengths close to the trapping wavelength [43]. However, the heating of plasmonic nanomaterials due to their absorption of the trapping laser beam may alter the viscosity of the fluid around it, and therefore its rotation and translation speeds may experience artefacts in temperature estimation [48, 49]. Spinning of gold nanorods is typically measured by autocorrelation, where the rotation of

Optical Trapping of Luminescent Nanothermometers

325

particles can be estimated by the analysis by autocorrelation of the time trace of the light scattered by the particle under linearly polarized illumination [43, 48–50].

7.3 Dark Field Spectroscopy for Temperature Sensing The heat produced by an optically trapped gold nanoparticle can be measured by its scattering spectrum obtained by dark field spectroscopy, as the heating localized around the plasmonic nanoparticle will reduce the refractive index of the surrounding medium, blue-shifting the wavelength of the plasmon resonance [9, 43, 47, 49]. Dark field spectroscopy or imaging setups can be easily incorporated into an optical tweezer by adding illumination with a ring of light and spectrally analyzing the backscattered light or performing hyperspectral imaging of the optical trap, respectively [9, 51].

7.4 Anti-Stokes Spectroscopy for Temperature Sensing Anti-Stokes emission from metal nanoparticles (see Fig. 3a) cannot be associated with a typical anti-Stokes Raman scattering process (i.e. it does not provide information on the nature of the metal). The mechanism causing the anti-Stokes emission may be electronic Raman scattering or Purcell effect enhanced hot carrier photoluminescence, depending on the plasmonic system. It is short-lived, so it is weak. One of the advantages of the anti-Stokes emission of metals is that it only depends on temperature and laser excitation power, so calibration is not essential and artifacts in temperature estimation can be easily prevented. It has an exponential-like shape, and it can be fitted to obtain an estimation of the temperature inside the metal nanoparticle (see Fig. 3) [52–55]. This temperature measurement process can be performed by single plasmonic particle analysis. However, it has not been combined with optical trapping to isolate and manipulate the plasmonic thermometer yet.

8 Perspectives High temperature sets a challenge for stable optical trapping of particles, as Brownian motion increases, making it easier for the particle to escape from the optical trap [56]. Nanothermometers must be optimized to maximize optical forces (see Section Optimization of the Optical Trapping of Luminescent Nanothermometers). Moreover, optical tweezers can be improved to achieve larger optical forces by, for instance, decreasing the size trap [11, 12]. The reliability of current methods in luminescent nanothermometry has been questioned recently [11, 57, 58]. Apart from the already mentioned issues such as

326

L. Labrador-Páez and P. Haro-González

Fig. 3 Anti-Stokes thermometry. a Typical experimental emission spectrum from a single gold bowtie excited at 633 nm. b Semilog plot of Anti-Stokes spectra, normalized to the emission intensity at a Stokes shift of −500 cm−1 , demonstrating the increasing portion of the signal at higher photon energies as the temperature is increased (the temperature legend is given in c). c Ratio of each spectrum in b to a reference spectrum obtained at 300 K. Dashed black lines are numerical fits to the ratio of two thermal population factors. d Extracted temperatures from numerical fitting in c plotted against the set temperature. Violin plots are used to show the distribution of extracted temperatures from several measurements. Reproduced with permission [55]. Copyright 2018 American Chemistry Society

light-to-heat conversion in metallic nanoparticles, there are other sources of inaccuracy. For instance, the difficulty to achieve reproducibility and reduce size distribution on the synthesis of nanoparticles affects the reliability of temperature measurements with single nanothermometers. The variability of nanoparticles size, doping, shell thickness, etc. makes it challenging to reliably calibrate single particle nanothermometers, as the temperature calibration may change slightly for each single particle [59]. Efforts should be directed towards the development of reproducible synthesis routes to achieve monodisperse and homogeneous nanothermometers [60]. Mostly lanthanide doped and plasmonic particles have been used for thermometry in optical trapping. Other materials, such as silicon nanorods, also work as thermometers when optically trapped [61]. The optical trapping of temperature-sensitive luminescent nanomaterials such as quantum dots or dye-loaded microspheres should be explored.

References 1. I.C. Lenton, E.K. Scott, H. Rubinsztein-Dunlop, I.A. Favre-Bulle, Optical tweezers exploring neuroscience. Front. Bioeng. Biotechnol. 1360 (2020) 2. R. Zhu, T. Avsievich, A. Popov, I. Meglinski, Optical tweezers in studies of red blood cells. Cells 9(3), 545 (2020) 3. P. Rodríguez-Sevilla, L. Labrador-Páez, D. Jaque, P. Haro-González, Optical trapping for biosensing: materials and applications. J. Mater. Chem. B. 5(46), 9085–9101 (2017) 4. K.C. Neuman, S.M. Block, Optical trapping. Rev. Sci. Instrum. 75(9), 2787 (2004). https://doi. org/10.1063/1.1785844 5. A. Ashkin, Forces of a single-beam gradient laser trap on a dielectric sphere in the ray optics regime. Biophys. J. 61(2), 569–582 (1992). https://doi.org/10.1016/S0006-3495(92)81860-X

Optical Trapping of Luminescent Nanothermometers

327

6. P. Haro-González, L. Martínez Maestro, M. Trevisani, S. Polizzi, D. Jaque, J. García Sole et al., Evaluation of rare earth doped silica sub-micrometric spheres as optically controlled temperature sensors. J. Appl. Phys. 112(5), 054702 (2012) 7. P. Haro-González, W.T. Ramsay, L.M. Maestro, B. Del Rosal, K. Santacruz-Gomez, M. del Carmen Iglesias-de la Cruz, et al., Quantum dot-based thermal spectroscopy and imaging of optically trapped microspheres and single Cells. Small 9(12), 2162–70 (2013) 8. P.B. Roder, B.E. Smith, X. Zhou, M.J. Crane, P.J. Pauzauskie, Laser refrigeration of hydrothermal nanocrystals in physiological media. Proc. Natl. Acad. Sci. 112(49), 15024– 15029 (2015) 9. A. Andres-Arroyo, F. Wang, W.J. Toe, P. Reece, Intrinsic heating in optically trapped Au nanoparticles measured by dark-field spectroscopy. Biomed. Opt. Expr. 6(9), 3646–3654 (2015) 10. P. Rodríguez-Sevilla, Y. Zhang, P. Haro-González, F. Sanz-Rodríguez, F. Jaque, J.G. Solé et al., Thermal scanning at the cellular level by an optically trapped upconverting fluorescent particle. Adv. Mater. 28(12), 2421–2426 (2016) 11. D. Lu, M. Pedroni, L. Labrador-Páez, M.I. Marqués, D. Jaque, P. Haro-González, Nanojet trapping of a single sub-10 nm upconverting nanoparticle in the full liquid water temperature range. Small 17(7), 2006764 (2021) 12. D. Lu, F. Gámez, P. Haro-González, Temperature effects on optical trapping stability. Micromachines 12(8), 954 (2021) 13. H. Rodríguez-Rodríguez, P. Rodriguez Sevilla, E. Martin Rodriguez, D.H. Ortgies, M. Pedroni, A. Speghini et al., Enhancing optical forces on fluorescent up-converting nanoparticles by surface charge tailoring. Small 11(13), 1555–1561 (2015) 14. P. Rodríguez-Sevilla, K. Prorok, A. Bednarkiewicz, M.I. Marqués, A. García-Martín, J. García Solé et al., Optical forces at the nanoscale: size and electrostatic effects. Nano Lett. 18(1), 602 (2018). https://doi.org/10.1021/acs.nanolett.7b04804 15. M. Wang, G. Abbineni, A. Clevenger, C. Mao, S. Xu, Upconversion nanoparticles: synthesis, surface modification and biological applications. Nanomed. Nanotechnol. Biol. Med. 7(6), 710–29 (2011). https://doi.org/10.1016/j.nano.2011.02.013 16. T. Hirsch, Surface Modification of Near Infrared-Emitting Nanoparticles for Biomedical Applications, in Near Infrared-Emitting Nanoparticles for Biomedical Applications. ed. by A. Benayas, E. Hemmer, G. Hong, D. Jaque (Springer International Publishing, Cham, 2020), pp.49–61 17. A.N. Grigorenko, N.W. Roberts, M.R. Dickinson, Y. Zhang, Nanometric optical tweezers based on nanostructured substrates. Nat. Photon. 2(6), 365–370 (2008). https://doi.org/10.1038/nph oton.2008.78 18. Y.-F. Chen, X. Serey, R. Sarkar, P. Chen, D. Erickson, Controlled photonic manipulation of proteins and other nanomaterials. Nano Lett. 12(3), 1633–1637 (2012). https://doi.org/10.1021/ nl204561r 19. M.S. Yousafzai, F. Ndoye, G. Coceano, J. Niemela, S. Bonin, G. Scoles et al., Substratedependent cell elasticity measured by optical tweezers indentation. Opt. Lasers Eng. 76, 27–33 (2016) 20. J.E. Molloy, M.J. Padgett, Lights, action: optical tweezers. Contemp. Phys. 43(4), 241–258 (2002). https://doi.org/10.1080/00107510110116051 21. A. Rohrbach, Stiffness of optical traps: quantitative agreement between experiment and electromagnetic theory. Phys. Rev. Lett. 95(16), (2005). https://doi.org/10.1103/PhysRevLett.95. 168102 22. P. Rodriguez-Sevilla, Y. Zhang, P. Haro-Gonzalez, F. Sanz-Rodriguez, F. Jaque, J.G. Sole et al., Thermal scanning at the cellular level by an optically trapped upconverting fluorescent particle. Adv. Mater. 28(12), 2421–2426 (2016). https://doi.org/10.1002/adma.201505020 23. P. Haro-Gonzalez, W.T. Ramsay, L. Martinez Maestro, B. del Rosal, K. Santacruz-Gomez, M. del Carmen Iglesias-de la Cruz, et al., Quantum dot-based thermal spectroscopy and imaging of optically trapped microspheres and single cells. Small. 9(12), 2162–70 (2013). https://doi. org/10.1002/smll.201201740

328

L. Labrador-Páez and P. Haro-González

24. H. Liang, K.T. Vu, P. Krishnan, T.C. Trang, D. Shin, S. Kimel et al., Wavelength dependence of cell cloning efficiency after optical trapping. Biophys. J. 70(3), 1529–1533 (1996) 25. K.C. Neuman, E.H. Chadd, G.F. Liou, K. Bergman, S.M. Block, Characterization of photodamage to Escherichia coli in optical traps. Biophys. J . 77(5), 2856–2863 (1999). https://doi. org/10.1016/s0006-3495(99)77117-1 26. S. Drobczy´nski, K. Prorok, K. Tamarov, K. Dus-Szachniewicz, V.-P. Lehto, A. Bednarkiewicz, Toward controlled photothermal treatment of single cell: optically induced heating and remote temperature monitoring in vitro through double wavelength optical tweezers. ACS Photon. 4(8), 1993–2002 (2017) 27. I. Verdeny, A. Farré, J. Mas, C. López-Quesada, E. Martín-Badosa, M. Montes-Usategui, Optical trapping: a review of essential concepts. Opt. Pura. Apl. 44(3), 527–551 (2011) 28. R.W. Bowman, M.J. Padgett, Optical trapping and binding. Rep. Progr. Phys. 76(2), (2013). https://doi.org/10.1088/0034-4885/76/2/026401 29. H. Zhang, K.-K. Liu, Optical tweezers for single cells. J. R. Soc. Interface 5(24), 671–690 (2008). https://doi.org/10.1098/rsif.2008.0052 30. K.C. Neuman, A. Nagy, Single-molecule force spectroscopy: optical tweezers, magnetic tweezers and atomic force microscopy. Nat. Methods 5(6), 491–505 (2008). https://doi.org/ 10.1038/nmeth.1218 31. C. Veigel, M.L. Bartoo, D.C.S. White, J.C. Sparrow, J.E. Molloy, The stiffness of rabbit skeletal actomyosin cross-bridges determined with an optical tweezers transducer. Biophys. J. 75(3), 1424–1438 (1998) 32. M.T. Valentine, N.R. Guydosh, B. Gutierrez-Medina, A.N. Fehr, J.O. Andreasson, S.M. Block, Precision steering of an optical trap by electro-optic deflection. Opt. Lett. 33(6), 599–601 (2008). https://doi.org/10.1364/ol.33.000599 33. F. Gittes, C.F. Schmidt, Interference model for back-focal-plane displacement detection in optical tweezers. Opt. Lett. 23(1), 7–9 (1998). https://doi.org/10.1364/ol.23.000007 34. D. Jaque, B.D. Rosal, E.M. Rodríguez, L.M. Maestro, P. Haro-González, J.G. Solé, Fluorescent nanothermometers for intracellular thermal sensing. Nanomedicine 9(7), 1047–62 (2014) 35. D. Jaque, F. Vetrone, Luminescence nanothermometry. Nanoscale 4(15), 4301–4326 (2012). https://doi.org/10.1039/C2nr30764b 36. A. Siai, P. Haro-Gonzalez, K.H. Naifer, M. Ferid, Optical temperature sensing of Er3+ /Yb3+ doped LaGdO3 based on fluorescence intensity ratio and lifetime thermometry. Opt. Mater. 76, 34–41 (2018). https://doi.org/10.1016/j.optmat.2017.12.018 37. S. Baral, S.C. Johnson, A.A. Alaulamie, H.H. Richardson, Nanothermometry using optically trapped erbium oxide nanoparticle. Appl. Phys. A 122(4), 340 (2016). https://doi.org/10.1007/ s00339-016-9886-0 38. P. Rodríguez-Sevilla, Y. Arita, X. Liu, D. Jaque, K. Dholakia, The temperature of an optically trapped, rotating microparticle. ACS Photon. 5(9), 3772–3778 (2018). https://doi.org/10.1021/ acsphotonics.8b00822 39. M.E.J. Friese, T.A. Nieminen, N.R. Heckenberg, H. Rubinsztein-Dunlop, Optical alignment and spinning of laser-trapped microscopic particles. Nature 394(6691), 348–350 (1998). https:// doi.org/10.1038/28566 40. L. Paterson, M.P. MacDonald, J. Arlt, W. Sibbett, P.E. Bryant, K. Dholakia, Controlled rotation of optically trapped microscopic particles. Science 292(5518), 912–914 (2001). https://doi.org/ 10.1126/science.1058591 41. A.I. Bishop, T.A. Nieminen, N.R. Heckenberg, H. Rubinsztein-Dunlop, Optical microrheology using rotating laser-trapped particles. Phys. Rev. Lett. 92(19), 198104 (2004). https://doi.org/ 10.1103/PhysRevLett.92.198104 42. A. Lehmuskero, P. Johansson, H. Rubinsztein-Dunlop, L. Tong, M. Kall, Laser trapping of colloidal metal nanoparticles. ACS Nano 9(4), 3453–3469 (2015) 43. L. Shao, M. Käll, Light-driven rotation of plasmonic nanomotors. Adv. Func. Mater. 28(25), 1706272 (2018) 44. E. Ortiz-Rivero, K. Prorok, I.R. Martín, R. Lisiecki, P. Haro-González, A. Bednarkiewicz et al., Laser refrigeration by an ytterbium-doped NaYF4 microspinner. Small 17(46), 2103122 (2021). https://doi.org/10.1002/smll.202103122

Optical Trapping of Luminescent Nanothermometers

329

45. E. Ortiz-Rivero, K. Prorok, M. Skowickł, D. Lu, A. Bednarkiewicz, D. Jaque et al., Single-cell biodetection by upconverting microspinners. Small 15(46), 1904154 (2019). https://doi.org/ 10.1002/smll.201904154 46. L. Jauffred, A. Samadi, H. Klingberg, P.M. Bendix, L.B. Oddershede, Plasmonic heating of nanostructures. Chem. Rev. 119(13), 8087–8130 (2019) 47. K. Setoura, Y. Okada, D. Werner, S. Hashimoto, Observation of nanoscale cooling effects by substrates and the surrounding media for single gold nanoparticles under CW-laser illumination. ACS Nano 7(9), 7874–7885 (2013) 48. F. Hajizadeh, L. Shao, D. Andrén, P. Johansson, H. Rubinsztein-Dunlop, M. Käll, Brownian fluctuations of an optically rotated nanorod. Optica. 4(7), 746–751 (2017) 49. D. Andrén, L. Shao, N. Odebo Länk, S.S. A´cimovi´c, P. Johansson, M. Käll, Probing photothermal effects on optically trapped gold nanorods by simultaneous plasmon spectroscopy and brownian dynamics analysis. ACS Nano 11(10), 10053–61 (2017). 50. H. Šípová, L. Shao, N. Odebo Länk, D. Andrén, M. Käll, Photothermal DNA release from laser-tweezed individual gold nanomotors driven by photon angular momentum. ACS Photon. 5(6), 2168–2175 (2018) 51. J.A. Rodrigo, T. Alieva, Light-driven transport of plasmonic nanoparticles on demand. Sci. Rep. 6(1), 1–6 (2016) 52. G. Baffou, Anti-Stokes thermometry in nanoplasmonics. ACS Nano 15(4), 5785–5792 (2021) 53. A. Carattino, M. Caldarola, M. Orrit, Gold nanoparticles as absolute nanothermometers. Nano Lett. 18(2), 874–880 (2018) 54. Y.-Y. Cai, L.J. Tauzin, B. Ostovar, S. Lee, S. Link, Light emission from plasmonic nanostructures. J. Chem. Phys. 155(6), 060901 (2021) 55. S. Jones, D. Andren, P. Karpinski, M. Kall, Photothermal heating of plasmonic nanoantennas: Influence on trapped particle dynamics and colloid distribution. ACS Photon. 5(7), 2878–2887 (2018) 56. D. Lu, L. Labrador-Páez, E. Ortiz-Rivero, P. Frades, M.A. Antoniak, D. Wawrzynczyk et al., Exploring single-nanoparticle dynamics at high temperature by optical tweezers. Nano Lett. 20(11), 8024–8031 (2020) 57. L. Labrador-Páez, M. Pedroni, A. Speghini, J. García-Solé, P. Haro-González, D. Jaque, Reliability of rare-earth-doped infrared luminescent nanothermometers. Nanoscale 10(47), 22319–22328 (2018) 58. Y. Shen, H.D. Santos, E.C. Ximendes, J. Lifante, A. Sanz-Portilla, L. Monge et al., Ag2S nanoheaters with multiparameter sensing for reliable thermal feedback during in vivo tumor therapy. Adv. Func. Mater. 30(49), 2002730 (2020) 59. J.R. Casar, C.A. McLellan, C. Siefe, J.A. Dionne, Lanthanide-based nanosensors: refining nanoparticle responsiveness for single particle imaging of stimuli. ACS Photon. 8(1), 3–17 (2020) 60. G. González-Rubio, P. Díaz-Núñez, A. Rivera, A. Prada, G. Tardajos, J. González-Izquierdo et al., Femtosecond laser reshaping yields gold nanorods with ultranarrow surface plasmon resonances. Science 358(6363), 640–644 (2017) 61. P. Karpinski, S. Jones, H. Sipova-Jungova, R. Verre, M. Käll, Optical rotation and thermometry of laser tweezed silicon nanorods. Nano Lett. 20(9), 6494–6501 (2020)

Critical Analysis of the Recent Advances, Applications and Uses on Luminescence Thermometry Maria Cinta Pujol Baiges and Joan Josep Carvajal Martí

Abstract Luminescence thermometry provides for numerous applications ranging from biosciences to engineering. The research in this field has dominated the luminescence temperature dependence in numerous materials and has got the comprehension of the mechanisms behind the thermal sensitivity. In this chapter, it is pretended to highlight the last novel approaches in the research field of luminescence thermometry, such as novel materials, and future directions pointed by the prominent researchers in the field. It is given an overview of the remaining challenges for the luminescent thermometry. Keywords Approaches for luminescence thermometry · Multiparametric · Diffraction limit · Resolution · Sensitivity · Uncertainty · Persistent luminescence · Pumping regime · Internet of Things

1 Introduction Presently, temperature sensors account for around 80% of the worldwide sensor market, which was valued at USD 6412.5 million in 2020 and is projected to be worth USD 10,028.5 million by 2026, according to Mordor Intelligence Inc. [1]. Temperature is one of the most demanded sensing targets, in science and in the global economy [2, 3]. Optical thermometers are gaining market share recently, particularly in industrial contexts of strong electromagnetic fields and moving objects. Recently, an increase in the contribution of luminescent materials displaying temperaturedependent properties (also called thermographic phosphors) stand out. Luminescence thermometry (also known as phosphor thermometry), as largely discussed along previous chapters, is a spectroscopic technique for remote detection M. C. Pujol Baiges (B) · J. J. Carvajal Martí Departament Química Física i Inorgànica, Universitat Rovira i Virgili, Campus Sescelades, Marcel·lí Domingo, 1, 43007 Tarragona, Spain e-mail: [email protected] J. J. Carvajal Martí e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 J. J. Carvajal Martí and M. C. Pujol Baiges (eds.), Luminescent Thermometry, https://doi.org/10.1007/978-3-031-28516-5_11

331

332

M. C. Pujol Baiges and J. J. Carvajal Martí

of temperature based on the temperature dependence of the luminescence properties of phosphors with numerous applications ranging from biosciences to engineering. The research related to this field, after an initial step dominated essentially by the demonstration of luminescence temperature dependence in numerous materials, now is moving to a more mature level requiring the comprehension of the mechanisms behind the thermal sensitivity, a complete description of the thermometer performance, and the development of specific applications. At the nanoscale, luminescence nanothermometry is a research field aiming at measuring the temperature at nanoscopic scales where conventional methods are unpractical. Thus, nowadays nanomaterials with temperature-dependent luminescent properties are one of the most versatile thermometers on the microscopic scale with applications in biology, electronics and catalysis, among others. Different phosphors have been examined for the material in which are based luminescent thermometers, such as polymers, DNA or protein conjugated systems, organic dyes, quantum dots, transition metal- and lanthanide-doped materials providing a remote (and not contactless, as erroneously mentioned in the literature) thermal detection through their light emission properties. Individual thermal probes and more complex structures, including core@shell nanoarchitectures, heaterthermometer nanoplatforms, and mixtures of thermographic phosphors encapsulated into polymers and organic–inorganic hybrids have been reported in the literature, acting as luminescent thermometers. All this made to arise interest for luminescence thermometry in a wide variety of fields, such as microelectronics, microoptics, photonics, micro and nanofluidics, nanomedicine, theragnostic, and many other conceivable applications, such as thermally-induced drug release, phonon-, plasmonic-, and magnetic-induced hyperthermia and wherever exothermal chemical or enzymatic reactions occur at submicrometer levels. In this chapter, it is pretended to highlight the last novel approaches in the research field of luminescence thermometry, such as novel materials, and future directions pointed by the prominent researchers in the field.

2 Novel Approaches in Luminescence Thermometry It is very important to understand the processes responsible for thermal changes in spectroscopic or other physical parameters. In the recent years, these processes have been meticulously studied, identified and theoretically described. This allowed that today, luminescence thermometry has entered an advanced level of development.

2.1 Luminescent Primary Thermometers A more detailed description of this novel approach can be found in Chap. 3. Up to now, many of the reported luminescent thermometers require calibration against

Critical Analysis of the Recent Advances, Applications and Uses …

333

a primary thermometer, in which the temperature is calculated by a well-defined equation of state. Recurrent calibrations are, thus, mandatory, particularly when the thermometers are used in different media, limiting the present technology. The determination of the temperature based on well-defined physical principles by luminescent primary thermometers is one of the ways to overcome this limitation. In luminescent primary thermometers (from now on simply primary thermometers), the temperature is calculated by a well-known equation of state, with all parameters or constants determined a priori. Temperature is, thus, calculated without any previous calibration, meaning that it can be applied as a reference (standard) thermal readout. Secondary thermometers, instead, require calibration at least once, if not multiple times, namely if they are used in different media. Despite the developments in luminescence thermometry, only a few primary luminescent thermometers have been developed up to now based on the emission and excitation features of thermographic phosphors, essentially Ln3+ -doped materials. Examples include the temperature dependence of the (i) peak energy of semiconductor nanoparticles [4]; (ii) emission intensities of two Ln3+ excited states in thermal equilibrium [5], and (iii) lifetime of lanthanide emissions. There are four types of primary thermometers whose thermometric parameters are based on emission features: 1st type: Those based on the Varshni’s law that describes the variation of a semiconductor energy gap (E g ) with temperature [6]: Eg (T ) = E0 −

aT 2 T +β

(1)

where E 0 represents the energy gap when the temperature tends to 0 K, β is a constant related to the material Debye temperature, and a is a phenomenological constant. This type of primary luminescent thermometers has been only demonstrated up to now in silicon nanoparticles functionalized with 1-dodecene with an average diameter of 2.0 ± 0.2 nm in solution, suspended in toluene, and in films, prepared by dropcasting into a Si wafer. The thermometric performance of this system is lower than that for other primary thermometers also described in Chapter 3. Nevertheless, these primary thermometers are independent of the sample processing and the working environment. 2nd Type: The second type of primary luminescent thermometers based on emission features are those based on the Boltzmann’s law. Photoluminescence features are used to determine the relative electronic population of the emitting levels using the radiative intensity of their emissions. Taking this into account, the intensity ratio between two emissions arising from thermally coupled energy levels of an ion can have expressed as [7]: ( ) 1 Δ 1 kB = ln − T T0 ΔE Δ0

(2)

334

M. C. Pujol Baiges and J. J. Carvajal Martí

where Δ0 is the intensity ratio when the temperature tends to 0 K. This equation supposes that the thermal equilibrium is effective. This means that the thermal exchange rate between two excited states is high enough such that thermodynamic equilibrium is always sustained, and all emissive ions can be considered distinguishable and independent units. This implies that the radiative decay rates from the excited states need to be much lower than the respective non-radiative rates to establish a sufficiently quick thermal equilibrium between the populations of the states. Moreover, the assumptions for the validity of the Boltzmann equilibrium consider exclusively no interacting ions, thus, excluding naturally the so-called energy transfer driven luminescent thermometers, involving ligand-to-ion and ion-ion interactions [8–10]. These kinds of primary luminescent thermometers have been used to calibrate a secondary luminescent thermometer based on the emissions of another lanthanide ion [2, 3]. This can be very useful, for instance, to predict the temperature in vivo; determined up to now only by secondary thermometers that could be difficult to calibrate, in some circumstances. Also, they can be used during the determination of the emission quantum yields [11], demonstrating that it is possible to simultaneously determine the emission quantum yield and the local temperature of the sample. They can also be used to correct the thermometric parameter from the interference of intruding emissions, envisaging reliable temperature measurements, correcting the integrated areas of their emissions, or even those effects arising from absorption of media (water for example) at particular wavelengths, allowing the development of primary luminescent thermometers operating under two distinct excitation wavelengths [12]. 3rd type: Another type of primary luminescent thermometer based on emission features is that based on the Mott-Seitz model, that considers the competition between radiative and non-radiative transitions of an emitting level with a lifetime (τ ): τ (T ) =

τ0

) ( 1 + αr exp − kΔE BT

(3)

where τ0 is the decay time at the limit when the temperature tends to 0 K (also known as the radiative decay time), αr is the ratio between the radiative and the non-radiative transfer rates and ΔE is the activation energy for the depopulation of the emitting level [13]. This type of thermometers, based on Eu3+ and Tb3+ in organic–inorganic hybrids, were used to sense temperature in real-time using a photography taken with the CCD camera of a smartphone [14], since the temperature dependence of the intensity ratio between the transitions of Eu3+ and Tb3+ is analogous to that found for their lifetimes. This constitutes the first example in which smartphones are used as an effective alternative to portable spectrometers to calculate the temperature using the induced colour temperature change.

Critical Analysis of the Recent Advances, Applications and Uses …

335

They have also been used, when based on Dy3+ single-molecule magnets, to work as a dual and synchronous thermometric/magnetic optical sensor in large ranges (10– 180 K and up to 45 T) [15], constituting also the first example of a molecular primary luminescent thermometer. 4th type: The last type of primary thermometers based on emission features are those based on an excited state absorption approach. The process of energy absorption from an excited state can be also used to produce a thermally induced change in the material’s luminescence. The general procedure consists of monitoring the emission lines with the temperature when the population of an excited state is pumped to the emitting or a higher level. The primary thermometer combines both ground state absorption (GSA) and excited state absorption (ESA), defining single-band ratiometric (SBR) thermometric parameters [16–19]: ΔSBR =

IESA IGSA

(4)

In such thermometers, the two electronic levels involved in the thermometer are not thermally coupled. However, this type of thermometers involve excitation at two different wavelengths. Another class of primary luminescent thermometers are those based on excitation features. Although such thermometers require more elaborate equipment and longer acquisition times to record the excitation spectra, these drawbacks can be minimised by using few excitation bands recorded in a narrower spectral range or using the excitation intensities at selected wavelengths instead of scanning a given spectral range. One of the advantages of these type of thermometers is that they allow for the definition of multiple thermometric parameters. This increases the chances of avoiding or eliminating interferences from the medium (e.g., competing absorption or wavelength-dependent scattering), and important feature in biological media, for instance. In this class of primary luminescent thermometers the thermometric parameters Δi are independent of the mode of acquisition (rate of energy or photons per second) of the excitation spectrum being under continuous excitation given by [20]: Δi =

) ( Sα, →β , −ΔEi = Ai exp Sα→β kB T

(5)

where Si is the area of the i-th transition in the excitation spectrum and ΔEi = Eα, − Eα is the energy difference between the barycentres of the low-lying α and α , levels. The pre-exponential factor Ai depends on the ratio of the absorption transition rates, the degeneracies of the α and α , levels, and the dielectric factor due to the local-field effects of a medium (or matrix) with refractive index n. This dielectric factor is 1 when both α → β and α , → β , transitions have the same mechanism, either (forced) electric-dipole or magnetic dipole, and then, Ai is independent of the temperature. Otherwise, the dielectric factor depends on the refractive index, which varies with the temperature. However, it has been shown that this dielectric factor is

336

M. C. Pujol Baiges and J. J. Carvajal Martí

practically constant in a wide range of temperatures, and, thus, Ai can be considered as constant. At the end, the equation of state is the same than the one describing the temperature prediction of any thermometer based on two thermally coupled electronic levels using the Boltzmann statistics. Also, if α = α , , self-referencing thermometers can be defined, and the thermometric parameter becomes constant (or temperature independent). Thus, having the same measuring physical–chemical system, there are distinct temperature-independent thermometric parameters that can be used, which is a remarkable feature allowing to distinguish between artificial temperature variations from the real ones of the surrounding medium. Finally, there is a last example of primary luminescent thermometers, that are those based simultaneously on excitation and emission features. When the transitions in the excitation spectrum of some lanthanide ions, such as Eu3+ , involve the same final energy level, the pre-exponential factor Ai defined in Eq. 6 can be expressed as the ratio of the areas in the emission spectrum and the temperature predicted by this class of primary thermometers becomes [20]: T=

1 να, β Sβ→α, ΔEi , Ai = f (n) kB ln(Ai /Δi ) ναβ Sβ→α

(6)

where Sβ→α and Sβ→α, are the integrated areas in the emission spectrum corresponding to the corresponding transitions with wavenumbers ναβ and να, β , respectively, and the local-field correction χL for absorption is estimated as χL (n) = ) ]2 {( 2 n + 2 /3 . Again, the dielectric factor is 1 when the two transitions have the same mechanism, whereas f (n) = χL (n)/n2 if they have different mechanisms. Again, several different intensity ratios can be defined for this kind of thermometers as thermometric parameters, which allows avoiding or eliminating interferences from the medium in the measurements. Despite all these different successful approaches, the performance of luminescent primary thermometers is still low when compared to that of non-luminescent primary thermometers, demonstrating the long way to go. Furthermore, the accuracy of the luminescent thermometers is still scarcely estimated, as most of the publications only report a single temperature readout and, thus, consistency analysis is not possible. In the few cases where quantification of the accuracy has been reported, they are still orders of magnitude below those of the well-established primary non-luminescent thermometers.

2.2 Multiparametric Approach Classical approach in luminescence thermometry reports a unique thermometric ratio. Another approach to luminescence thermometry is based on the multiparametric sensing. In this case, instead of evaluating only one thermometric parameter,

Critical Analysis of the Recent Advances, Applications and Uses …

337

several thermometric parameters are calculated, trying to cover different ranges of temperature, or in most cases, gaining reliability. As mentioned in the literature, acquiring in a synchronised way multiple thermal readings can permit checking the accuracy of luminescence thermometry. Thus, if the measurements are artefact-free, then all the different thermal readings should allow determining the same temperature. However, if the different thermal readouts diverge, this fact indicates that the luminescent nanothermometer is not working properly, so experimental artefacts affect the measurements. This approach can lead to design multiparametric self–reference luminescence temperature sensors, which makes the temperature measurement more accurate. Therefore, it is beneficial to develop multiparametric luminescent thermometers, and recently, several multiparametric luminescent thermometers have been reported. Some of them are summarized in Table 1. Different approaches have been used to design these luminescent thermometers: • To combine FIR signal with lifetime (LF)-based luminescence thermometry signal in a single luminescent centre in one host. Here, mainly lanthanide ions have been used as single luminescent centers. • Combination of FIR signal with LF-based luminescence thermometry from two luminescent centres in one host. These two centres can be two different lanthanide ions or the combination of a transition metal with lanthanide ions. • Use of luminescent band from different transition metals. The luminescent properties of inorganic materials doped with TM ions are attracting attention with the aim of developing highly sensitive luminescent thermometers. Cr3+ , Mn2+ /3+ /4+ , Ti3+ /4+ , V3+ ,4+5+ , Fe3+ , and Ni2+ ions have been thoroughly analysed for this purpose. Their advantages are the unique spectroscopic properties of TM ions, including the high thermal susceptibility of their emission intensity, as well as the ability to modify their spectroscopic properties by modifying the crystal field strength. In addition, many of these ions are very abundant on earth, which implies an additional economic benefit. • To combine FIR signal with LF-based luminescence thermometry from two luminescent centres, in this case in two separated hosts. Using a double luminescent thermometer working in two different spectral ranges allows for self-referenced temperature determination, for which, every single readout from one part of the sensor can be compared with the respective measurements using the second temperature sensor, as discussed previously for primary luminescent thermometers. As shown, in the examples summarized in Table 1, in most of the cases, when FIR and other thermometric parameters are used in the same thermometer, the highest thermal sensitivity is achieved when FIR is used as one of the thermometric parameters to analyse.

338

M. C. Pujol Baiges and J. J. Carvajal Martí

Table 1 Luminescent thermometers based on multiparametric approaches Material

Active centres

Thermometric parameters

Relative thermal sensitivity

References

Ca2 MgWO6

Er3+

FIR of H and S levels of Er3+ FIR red/green LF

0.92% K−1 (303 K, FIR H/S) 0.47% K−1 (303 K, FIR R/G) 0.11% K−1 (573 K, LF)

Jiang [21]

YVO4

Er3+

FIR Spectral line position bandwidth

2.61% K−1 (FIR)

Kolesnikov et al. [22]

SrB4 O7

Sm2+

FIR 580 nm/684 nm LF of the visible emissions

3.36%·K−1 at 500 K (LF) Cao et al. [23]

CaWO4

Ho3+

Intensity of Ho3+ emission at 1190 nm Spectral position of the W–O charge transfer band LF of 5 S2 level

0.0333 nm K−1 (spectral band position)

Zhou et al. [24]

Sr2 CeO4 NCs

Ln3+ (Sm, Ho, Yb, Nd)

FIR LF

2.80% K−1 (228 K, 8%Yb3+ : Sr2 CeO4)

Marciniak et al. [25]

CaAl2 O4

Co3+ , Nd3+

FIR LF

1.86% K−1 (263 K, FIR) 2.36% K−1 (349 K, LF)

Kniec et al. [26]

Y3 Al5 O12

Cr, Dy

FIR of 4 T2 → 4 A2 /2 E → 4 A2 FIR of 6 H15/2 → 4 F9/2 /6 H13/2 → 4 F9/2

SrTiO3

Mn4+ , Ln3+ (Eu, Dy)

Intensity FIR LF

5.64% K−1 (303 K, FIR)

Trejgis et al. [28]

Ag2 S NPs

None

MLR

50% K−1

Maturi et al. [29]

Coumarin 7 Ln3+ MOF

Eu3+

FIR of Eu3+ to dye molecules Band position of Coumarin 7 emission

0.50% K−1 (363 K, FIR)

Liu et al. [30]

Hexagonal NaYF4 core shell

Yb3+ ,Er3+ ; Yb3+ ,Nd3+

FIR(1) 1060 nm/980 nm FIR(2) 520 nm/540 nm

0.019% K−1 (FIR 1) 0.015% K−1 (FIR 2)

Marciniak et al. [31]

Ristíc et al. [27]

FIR: fluorescence intense ratio; LF: lifetime luminescence thermometry; NCs: nanocrystals. Ln3+ : trivalent lanthanide ions; MLR: multiple linear regression

Multiple regression based on multiparametric sensing Despite it helps to improve the reliability, the so-called multiparameter sensing is still unable to improve significantly the relative thermal sensitivity of the nanothermometers and the reported values are far below the most sensitive luminescent thermometers reported so far. The temperature resolution, δT , represents the smallest temperature change that can be detected in each measurement. δT depends on the type of the material and the experimental setup used. Depending on the experimental detection setup used, the acquisition conditions applied and the signal-to-noise ratio in the experiment, δT

Critical Analysis of the Recent Advances, Applications and Uses …

339

might change. For several of the applications in which luminescent thermometers are used, there is a need to improve this parameter. Maturi et al. [29] aimed to improve the temperature uncertainty of the luminescent thermometers by using a new definition of the thermometric parameter, based on the multiple linear regression (MLR). The MLR is a statistical mathematic approach that expresses the relationship between several independent variables and a dependent variable, representing the linear relationship as a single functional formula. The MLR model can be applied in this context assuming the temperature as the dependent variable (T) and the several independent variables being the different thermometric parameters, Δi . Then, temperature can be expressed as: T = C + β1 Δ1 + β2 Δ2 + …, where βi are the regression coefficients for the independent variable Δi, T is the dependent variable, in this case, the temperature, and C is the constant of the regression equation. The multiple linear regression model is based on a mathematical assumption that a linear relationship exists between both the independent and dependent variables. For this model to work, it is assumed that there is no significant correlation between the multiple independent variables. By applying the MLR process, the relative thermal sensitivity can be described by: √Σ n

Sr =

(Δt βt )−2

(7)

t=1

Maturi et al. [29] provided experimental evidence of how the synergy between MLR and multiparametric thermal sensing leads to a tenfold improvement in the performance of multiparametric luminescent nanothermometers establishing worldrecord values for S r and δT. They tested this novel approach using a luminescent thermometer based on a modified green fluorescent protein. This chromophore shows a broad emission centred in the green, covering the wavelength range from 480 to 600 nm, when excited at 408 nm. The luminescence shows thermal quenching when increasing from room temperature to 323 K. Five different thermometric parameters were used for this approach: the intensity ratio of the two peaks, the peak energy of the two peaks, and their respective full width at half maximum. Each of these parameters can be used as independent Δ values for multiparametric thermal reading and their temperature dependencies were described by a single linear fit with a positive correlation. By combining all the parameters, Sr reached 3.0% K–1 , which represents a tenfold increase compared to the highest Sr obtained in single parametric sensing in the same experiment (0.33% K–1 for one full width). The same authors applied this approach to Ag2 S NPs as luminescent thermometers. They operate in the 2nd BW for real sub-tissue thermal sensing. Ag2 S nanoparticles possess an excellent in vivo biocompatibility due to their high physic-chemical stability. Ag2 S nanoparticles emit a single band centred at 1200 nm with a strong temperature dependence. In this case, they use two different thermometric parameters: the position of the emission band and the ratio between two intensities in the band emitted by the NPs. Applying the MLR approach, they reached a maximum thermal sensitivity as height as 50% K−1 at 298 K. The improved performance of MLR is

340

M. C. Pujol Baiges and J. J. Carvajal Martí

manifested in both Sr and δT values calculated from each thermometric parameter. The intensity ratio between the two emitted intensities has a thermal uncertainty value ranging from 1.1 to 2.1 K, and the position of the emission band reaches the value δT ≈ 0.15 K, achieving an improvement by one order of magnitude. Headedly, the MLR approach gives temperature uncertainties between 0.05 and 0.10 K with an improvement of more than 20 times with respect to the one obtained with the intensity ratio approach using a single thermometric parameter.

2.3 Luminescence Nanothermometry Below the Diffraction Limit Temperature measurement with high spatial resolution, Δx, is of great interest in various fields of research, like in monitoring microelectronic circuits or in biomedical applications. Substantial efforts have been made to improve Δx ranging from contact measurements like atomic force microscopy to remote luminescence nanothermometry. Scanning thermal microscopy, for instance, can resolve features below 50 nm with a temperature uncertainty of around 10–3 K. However, it can used only in surface measurements. Thiem et al. [32, 33] numerically investigated the improvement in spatial resolution of nanothermometers with a super-resolution microscopy technique, the stimulated emission depletion microscopy (STED) approach. This technique is based on creating super-resolution images by selectively deactivating fluorophores, minimizing the area of illumination at the focal point. This fact is what provides the super-resolution below the diffraction limit [34]. It is based on luminescence confocal images which are acquired by scanning a light point focused on a region of interest and collecting the luminescence sequentially, pixel by pixel. STED takes advantage of the non-linear response of fluorophores commonly used to label biological samples in order to achieve this high resolution. By applying a numerical beam propagation method, it was investigated the improvement that can be achieved in the spatial resolution of a 3D nanothermometric model. Thus, the results indicate the possibility to measure temperatures spatially resolved below the diffraction limit in a remote way with single degree temperature precision using currently available luminescent nanothermometers. The model was based on the combination of wide-angle beam propagation with laser rate equations using the scalar Helmholtz equation, in 2D simulations assuming a cylindrical symmetric system. In this approach, stimulated emission is used for spatially resolved luminescence suppression. By using specific depletion modes, Δx could be strongly increased below the limit of confocal microscopy. The corresponding Δx can be calculated by the beam divergence θ , the refractive index and the excitation wavelength λexc to be approximately Δx = 0.35λexc /n sin(θ ). The only specific material parameters necessary for the calculation are the absorption and emission cross sections and the luminescence lifetime. To demonstrate the

Critical Analysis of the Recent Advances, Applications and Uses …

341

possibilities of this technique, and as an example, Pr3+ :YLiF4 nanocrystals were selected, since they combine high thermal sensitivity (1.1% K−1 ) with being excited at a short wavelength (444 nm). If a Gaussian beam geometry is considered for the excitation laser, it can provide a waist of 500 nm, which enables tight focusing of the excitation beam with a fluence of 1200 MW cm−2 to ensure sufficient luminescence emission. The wavelength input for the depletion beam was chosen to be 607 nm, to assure a large spectral separation of the two signal emission peaks (520 and 639 nm) used to calculate the intensity ratio from which the temperature was derived. The results obtained showed that the improvement in spatial resolution is accompanied by a poorer temperature resolution, as can be seen in Fig. 1. This effect is explained by the change in the signal-to-noise ratio, which occurs when the shape of the population distribution changes from a Gaussian to a steeper distribution, leading to a compromise between both resolutions. Through this, the FWHM of the population strongly decreases, which consequently leads to an increase of Δx. On the other hand, when the signal-to-noise ratio decreases, the measurement error in the intensity ratio increases, which leads to the proportional increase of ΔT. The inverse product of Δx and ΔT was used as a figure of merit to quantify the performance of STED nanothermometry. The overall performance was optimised by gradually increasing the power ratio between the power of the depletion beam and the saturation power (that is calculated from the saturation intensity), and the beam waist ratio, that is the ratio between the waist of the depletion beam and the excitation beam. The combination of a large diameter for the excitation beam and a strong excitation power compensates for the steeper distribution of the population by assuring sufficient noise suppression, though at the cost of a reduced Δx. A dashed line in Fig. 1 marks the spatial resolution limit of the corresponding confocal microscope, defining the diffraction limit. The results show that temperature measurements below the diffraction limit are possible, while maintaining a value for the error in the intensity ratio below 8%. These results allowed also to study the influence of the sample position, relative to the excitation beam, corresponding to an axial scan through the sample, to better understand the relation between the temperature and the spatial resolutions. Figure 2 shows that inside the focal region, precise temperature measurements are possible. When the sample thickness increases, both Δx and ΔT decrease, while the output power (Pout ) of the luminescence increases due to the higher number of active ions generating the emission. These results show the potential of adapting the resolution achieved by the STED technique to specific applications, following the correlation between Δx and ΔT . This correlation, according to the authors, would also show the feasibility of 3D volumetric temperature measurements with high spatial resolution and ways to address the required compromises, by, for example, shifting the sample position. This model can be applied to all different luminescent nanothermometers that use energy level populations.

342

M. C. Pujol Baiges and J. J. Carvajal Martí

Fig. 1 a Influence of the ratio between the power of the depletion beam and the saturation power (left), and the ratio between the beam waist of the depletion and the excitation beam (right) on the spatial and temperature resolutions for two different definitions of the spatial resolution distinguished from the acceptable noise level, corresponding to 50% (definition in optical microscopy) and 5% (in agreement with luminescence nanothermometry measurements). b Combined resolution value to optimize Δx and ΔT . Reproduced with permission from Thiem et al. [32, 33]

3 Engineering Novel Materials 3.1 Heavily Doped Lanthanide Nanoparticles In lanthanide-doped materials, new strategies have been used to increase the brightness of the emissions of these materials, such as by increasing the concentration of doping ions (heavily doped lanthanide nanoparticles). Chapter 4 provides a detailed

Critical Analysis of the Recent Advances, Applications and Uses …

343

Fig. 2 Dependence of ΔT , Δx, and Pout a on the position of the sample relative to the excitation beam, and b on the sample thickness. Reproduced with permission from Thiem et al. [32, 33]

overview on how the use of heavily doped lanthanide nanoparticles influences the performance of luminescent thermometers. For instance, inspired on core–shell structures, that induced a luminescence enhancement by reducing the interaction between the active emitting ion and the surface quenchers, an active shell with sensitizer ions was developed [35–37]. These core-active shell nanoparticles could enhance the luminescence intensity several times in comparison with that of the core-inert shell counterparts with the same particle size. Other works focused on developing successive layer by layer homogeneous doping to synthesise onion-like structures. The periodical dopant distribution increased the quantum yield to nearly 2 times than that of the heterogeneous doping nanocrystals. The advantages of engineering the local dopant distribution overcome the underlying problems of Yb3+ heavily doped luminescent nanoparticles, i.e., the inability of most systems containing high Yb3+ concentrations to preserve the excitation energy. For example, in an orthorhombic crystallographic structure in which the lanthanide ions are distributed in arrays of tetrad clusters, this unique arrangement enables the preservation of the excitation energy within the sublattice domain and effectively minimises the migration of excitation energy to defects, even using a Yb3+ content as high as 98 mol % [38, 39]. To release the luminescence potential of Ln3+ heavily doped nanoparticles, another option is to use high power density excitations. The Ln3+ heavily doped nanoparticles are particularly beneficial for stronger excitations because of their much higher

344

M. C. Pujol Baiges and J. J. Carvajal Martí

upper limit for luminescence as compared to the low doped nanoparticles. The luminescence intensity of Ln3+ ions gradually increase with the excitation intensity, or the power density. However, the luminescence of low doped nanoparticles tends to be saturated upon strong excitation due to the limited number of luminescent ions encountered in the material. In contrast, the heavy doping greatly increases the ion numbers and thus, does not become easily saturated at high excitation intensities. The mechanisms generated by high excitation powers create more activators. Weak excitations can only activate a small number of activators in the Ln3+ heavily doped nanoparticles, and part of these active activators can be further deactivated by the inner or surface quenchers and the detrimental interactions, leading to a limited luminescence intensity. However, working at high excitation powers, a larger number of active Ln3+ ion can overwhelm the energy loss processes, enabling a significantly enhanced luminescence [11]. A third strategy is that related with high temperature treatments. The concentration quenching effect is often related to the size-dependent luminescence, as particle size can mainly determine the specific surface area and thus affect to the surface quenching processes. As a common phenomenon used during the nanoparticles synthesis, heat treatments at higher temperatures could induce an aggregation of the nanoparticles, a process in which they combine to form larger particles. The luminescence intensity of Ln3+ ions heavily doped nanoparticles can be greatly enhanced at a higher synthesis temperature, however, due to the reduction of surface quenching. In addition, besides the suppression of the surface quenching, high temperature treatments may also help to reduce the inner defects by the improvement of crystallinity, which also contributes to increase the luminescence intensity [15]. Finally, another effect influenced in heavily doped lanthanide nanoparticles is that spectroscopic parameters of Ln3+ ions depend on their surrounding environment. The Ln3+ dopants act as additives and generate slight local structural changes. Therefore, the different doping levels could alter these spectroscopic parameters. This variation might not be obvious in low doped samples, but should be quite evident in heavily doping conditions, affecting to the radiative transition rates probabilities. For example, when doping with high levels of Er3+ in the NaGdF4 host, it was observed that all the radiative transition rates increased gradually with the doping concentration [40, 41]. Luminescent heavily lanthanide doped thermometers have been evaluated either based on thermally- and non-thermally-coupled energy levels, or in LF thermometers. Luminescent thermometers based on thermally-coupled energy levels have been demonstrated on Er3+ and Nd3+ heavily doped systems using the intensity ratio approach. Also, for instance, by developing core-multishell nanostructures heavily doped with Tm3+ /Yb3+ and Er3+ /Yb3+ distributed in different regions, allowed to develop different temperature sensing channels. The Er3+ /Yb3+ codoped layer showed a good sensing performance in the high temperature range, while the Stark sublevels of Tm3+ allowed to develop a highly sensitive thermometer in the low temperature range [42], due to the tiny energy difference between these Tm3+ sublevels (around 300 cm−1 ).

Critical Analysis of the Recent Advances, Applications and Uses …

345

These strategies have effects on the thermometers based on them. For instance, the enhanced luminescence intensity induced by Ln3+ heavily doping can increase the signal-to-noise ratio of the sensing systems, and thus, it can generally decrease the uncertainty of the thermometer. However, it is noteworthy that the increase of luminescence intensity does not necessarily improve all the sensing performance. The high luminescence potential of the Ln3+ heavily doped nanoparticles can be released by a strong laser irradiance, but accordingly, the laser induced heating effect becomes considerable. This laser induced heating effect may lead to temperature overestimations, which in turn impedes their reliability for applications which require high levels of accuracy. To avoid this laser heating effect, an obvious option is using excitation powers as low as possible. However, the weak luminescence signal generated by the weak excitation is always a severe issue. From another side, by even using a low excitation power, the self-heating effect of heavily doped materials can still be obvious. In fact, for thermometers based on the fluorescence intensity ratio approach, a criterion for the sensing validity is that the incident laser power falls into the region in which the luminescent thermometer parameter hardly changes [43]. But this might lead to large sensing errors. Unlike in the luminescent materials in powder form, it has been shown that laser heating can be largely suppressed by using diluted colloidal solutions or vitreous hosts [40, 41]. Another solution to reduce the laser heating effect on Ln3+ heavily doped nanothermometers is applying a compensation method, based on the experimental observation and the estimated temperature gradient of the laser heating on the surface of the sample. A linear dependence of temperature increase with the excitation power density might be introduced into conventional luminescent thermometers based on thermally coupled levels, operating under the fluorescence intensity ratio approach [44, 45]. This correction is only valid for small laser spots, what is not a serious concern as a small laser spot is a precondition for high resolution temperature mapping. The restricted energy spacing between the thermally coupled levels (200– 2000 cm−1 ) is a limitation for the further improvement of the sensitivity of luminescent thermometers based on this scheme. Another limitation is the difficulty to be applied in the low temperature region, due to the low population of the upper level by thermal means. To improve this situation, the use of non-thermally coupled levels of two different lanthanide ions or two different luminescent centres, including two valence states from one type of lanthanide ions [46, 47], has been proposed. Because not being limited by the energy gap, this strategy can allow achieving very high relative thermal sensitivities [46, 48].

346

M. C. Pujol Baiges and J. J. Carvajal Martí

3.2 Metal–Organic Frameworks for Luminescence Thermometry Metal–organic frameworks (MOFs) are crystalline porous materials built up from metallic nodes and organic linkers. The modularity and tunability of MOFs make them very attractive for a wide variety of potential applications including gas storage and release, chemical separations, catalysis, drug delivery, light harvesting and energy conversion, sensing, conductivity, ion-exchange, removal of toxic substances from air and water, degradation of chemical warfare agents, etc. Different parts of the MOFs may be responsible for generating luminescence in MOFs. From one side the inorganic cluster, from another side the organic ligand and even also the incorporation of a host–guest molecule in the pores of the structure might be luminophores. When compared to organic dyes or inorganic luminescent materials, the luminescence of MOFs is more diverse because the metal nodes, organic linkers, ligand–metal charge transfer, and guest species within framework can potentially generate emissions. The application of MOFs in the field of luminescence thermometry is discussed in detail in Chap. 5. Up to now, mainly, MOFs designed to develop luminescent thermometers are mainly micrometre sized powders. MOFs built up from all kinds of inorganic metal cations and/or cluster nodes with various organic ligand linkers, turn out to be rational candidates for luminescent thermometers owing to the abundance of metal cations and the infinity of organic connectors with diverse geometries and luminophore functional groups. MOFs contain organic ligands but also coordinate solvent molecules, so their thermal stability is limited. Then, MOFs are not suitable for very high-temperature applications (T > 500 °C). A thorough investigation of the crystal structure of several MOFs evidences the role of the inorganic network topology on the range of temperatures in which the thermometer can operate. A high connectivity of the [LnOx ] polyhedra in the form of layers or chains, favours a thermal sensitivity in the cryogenic range, probably due to the low distance between the Ln3+ cations. By the contrary, when the connectivity decreases, the thermal sensitivity of the material is shifted towards higher temperatures. NIR luminescent MOFs for temperature sensing in the physiological range have already been developed [49], while the particularities of Eu-Tb MOFs make them also interesting candidates for luminescence thermometry [50]. Most of the MOFs used as luminescent thermometers are based on emission intensity measurements, and normally the temperature is evaluated from the measurement of the intensities of two energy levels (or two Stark components of an excited state) in thermal equilibrium. Because of the diversity of emissions arising from MOFs, many combinations are possible to elaborate ratiometric MOF thermometers and three different behaviours have been deduced for an effective design of dual-emitting MOFs: i. Luminescence generated by two independent emitting centres which are electronically independent. In these systems any energy transfer occurring between both emitters and the thermometric properties arise from separate thermal quenching

Critical Analysis of the Recent Advances, Applications and Uses …

347

mechanisms of each luminescent emitter. With this strategy, highly sensitive luminescent thermometers could be designed if the emitters have very distinct thermal quenching. ii. Luminescence generated by two closely related luminescent centres with one centre acting as sensitizer (or donor) via an energy transfer process. In that case, emitters must be spatially close to favour the energy transfer process, which regularly occurs between the organic ligand and the lanthanide ions (also called antenna effect) or between two lanthanide ions. The energy transfer process will compete with the donor emission and can be thermally dependent, which results in a very different temperature-dependence of the two emissions. As an example, Eu3+ and Tb3+ present a tremendous potential for applications as ratiometric luminescence thermometry based in MOFs [51]. In such materials, the organic ligand acts as structuring agent but also as sensitizer to favour the emission of the lanthanide ions. The organic ligand, with a higher absorption cross-section than the lanthanide ions, absorbs energy in the UV range, and via an energy transfer process from its triplet excited state, transfers the energy to the emitting energy states of Eu3+ and Tb3+ that generate the characteristic emission lines of the two emitters. In Eu-Tb MOFs luminescent thermometers operating in the cryogenic range (T < 100 K), a high connectivity of the [LnOx ] polyhedra seem to be favourable to the design of such luminescent thermometers. Instead, for luminescent thermometers operating in the medium range of temperatures (100 K < T < 300 K) a structure composed by isolated [LnOx ] polyhedra or dimers seems to be more favourable. The distance between isolated [LnOx ] polyhedra or dimers seems to be a crucial parameter that impacts the maximal temperature of operation of the thermometers. Thus, an increase of the distance between the [LnOx ] polyhedra tends to increase the temperature where the material is more sensitive. However, the topology of the inorganic network as well as the distance between the monomers or dimers are not the only parameters governing the temperature domain. The impact of the organic ligand should not be neglected, and in particular the energy position of its triplet level. For instance, an increase of 50 K in the operating temperature range has been observed for ligands of the imidazole family [52]. This has been attributed to the occurrence of different deactivation pathways in these ligands due to the presence of N-donor terminal or N/O-donor chelating aromatic ancillary ligands. Finally, the concentration of lanthanide ions, and in particular that of Eu3+ , seems to have a crucial role in the optimization of the relative thermal sensitivity of the MOFs luminescent thermometers [9, 10]. In fact, a continuous decrease in the relative thermal sensitivity was observed with the increase of the Eu3+ concentration while a reduction of the maximal temperature at which this sensitivity is produced was also observed. Thus, the Eu3+ /Tb3+ molar ratio appears to be a crucial parameter to improve the thermal performance of these materials as luminescent thermometers. iii. Luminescence generated by two thermally coupled energy levels of a single luminescent centre. The Boltzmann distribution law governs the electron populations of the two involved levels, and the temperature changes will induce an

348

M. C. Pujol Baiges and J. J. Carvajal Martí

electronic population re-distribution between energy levels, modifying the luminescence intensity ratio. This approach has not been explored too much in MOFs so far, with examples involving Dy3+ [53], in which the low energy difference (about 991 cm−1 ) between the 4 I15/2 and 4 F9/2 levels seems to be advantageous to design temperature sensors to operate in a relatively high temperature range. For the design of MOFs with emission in the NIR, apart from the selection of the suitable lanthanide ion with emission in this region, the organic ligand needs to have a much higher triplet energy level than the acceptor level of the lanthanide ions to limit the energy transfer or interaction between the linker and the lanthanide ions. Furthermore, it is also recommended to reduce the presence of some highenergy chemical bonds, such as C–H, N–H, and O–H, which can act as oscillators and quench the NIR emissions. For instance, to limit the presence of C–H bonds, fluorinated organic ligands can be used [54–56]. Normally in this type of MOFs, Nd3+ and Yb3+ have been chosen as NIR emitters for ratiometric thermometers, excited at 808 nm. Ratiometric luminescent thermometers can be also designed in MOFs with a single lanthanide ion, when the difference between the ligand excited state and the Ln3+ lowest emitting level is lower than 1500 cm−1 , increasing the probability of ionto-ligand back energy transfer [57]. In those cases, a strong thermal quenching of the emission from the organic ligand is observed, while the lanthanide emission (Eu3+ ) is almost invariant with the temperature [58]. This kind of thermometers can achieve a very high thermal sensitivity, up to 7.14%K−1 , although at low temperatures (65 K) [59]. In other cases, however, the contrary behaviours were observed, where the decrease of the temperature prevents the ion-to-ligand back energy transfer, resulting in a stronger Eu3+ emission. Doping lanthanide containing MOFs (with Gd3+ as the central metal, for instance) with a small amount of Eu3+ is also an alternative to get a single lanthanide containing metal organic frameworks working as ratiometric thermometers. The choice of the organic ligand is primordial to design sensitive thermometers [60–62], being those containing methoxyl and methyl groups the ones exhibiting the highest sensitivity, of the order of 7.78% K−1 at 313 K. In this kind of thermometers, since the temperature readout is obtained from the emissions of the host and the lanthanide ion, it is possible to incorporate a second lanthanide ion in the MOF structure to develop extra functionalities. MOFs are very well-known for their porous character, and the content of the porosity can be modified by exchange reactions, allowing, for instance, the incorporation of luminescent guests. In such systems, one emitter is contained in the framework while the other is allocated c the cavities. Luminescent thermometers of this kind have been demonstrated incorporating perylene dye into a porous Eu3+ MOF, in which by increasing the temperature from 20 to 80 ºC, the luminescence intensity at 473 nm of perylene dye substantially decreases due to the thermal activation of nonradiative pathways, while the intensity of the emission of Eu3+ , centred at 615 nm, increases as the temperature increases [63], Another example is the impregnation of a Tb3+ MOF with a coumarin derivative [53], that achieved a thermal sensitivity of

Critical Analysis of the Recent Advances, Applications and Uses …

349

4.48% K−1 at 300 K. These examples, however, had to be excited in the UV. An alternative is the use of two-photon luminescent dyes to afford the location of the excitation and emission bands in the biological windows to use them in biological applications. For that, Wan et al. [64] used the luminescent composite formed by the (Me2 NH2 )3 [In3 (BTB)4 ]·12DMF·22H2 O MOF, and the 4-[4-(diphenylamino)styryl]1-dodecylpyridinium) dye, that can be excited at 1064 nm (in the II-BW, see Chap. 6), with emission at 650 nm (in the I-BW, see Chap. 6). As all luminescent thermometers based on a single emitter, the performance of this type of thermometers depend on the emitter concentration and an additional two-photon dye could be introduced within the pores to design a ratiometric luminescent thermometer. The richness of luminescent organic dyes and porous MOFs offers wide choices for developing a great variety of MOF@dye composites, whose emission range can be designed and expanded to tune the energy transfer between the MOF and the dye, and thus to optimize the properties of the dual-emitting MOF@dye composites. Polyoxometalates, that can be described as anionic soluble metal oxides of early transition metals (usually W, Mo and V) and which can also contain lanthanide cations can also be introduced in the cavities of MOFs. For instance, [EuW10 O39 ]9− was incorporated into the cavities of a MOF to develop a luminescent composite operating in the 200–300 K range [47]. In another example, perovskite CsPbBr3 quantum dots were also encapsulated into a Eu-MOF cavity, generating a dual emitting luminescent thermometer [30]. In that case, the opposite variation of the temperature-dependence of each emission conducted to a high relative thermal sensitivity (3.9% K−1 at 373 K).

3.3 Persistent Luminescence Materials Novel approaches in luminescence thermometry are designed to overcome some of its already identified limitations. One of the drawbacks already described in the literature is related to the excitation light inevitably resulting in an undesired generation of heat that affects the accuracy of the thermometer. Therefore, by developing a type of temperature sensing method that does not require a continuous excitation would avoid this problem. This drawback can be overcome via a non-real-time photo-excitation scheme using a persistent phosphor. Persistent luminescence is also named afterglow. Persistent luminescence is a light-emission phenomenon of a persistent phosphor that can store external excitation energy and then persistently emit light for minutes to hours after the removal of the excitation [65]. As it is well known, traps (i.e., defects in the crystals) play a quite important role for generating persistent luminescence. The persistent luminescence phosphor to be used in a luminescence thermometry technique requires pre-irradiation for charging the irradiation energy that can be thermally released to produce the afterglow with a temperature-dependent spectral distribution. Lately, several studies in which the luminescence thermometry is based on persistent luminescence phosphors have been carried out. Some of these works are

350

M. C. Pujol Baiges and J. J. Carvajal Martí

Table 2 Selected examples of luminescence thermometry based on persistent luminescence phosphors Host

Doping ions Dy3+ ,

Tb3+

Sr max (% K−1 ) Ts (ºC) T range (ºC)

References

0.75

Martín et al. [66]

Sr4 Al14 O25

Eu2+ ,

10

10–60

SrAl2 O4

Eu2+ , Dy3+ , Tb3+ 7

−73

−213 to −33 Zhao et al. [67]

NaYF4

Tb3+

30

30–210

NaYF4

Tb3+

1.06

30

30–150

Wu et al. [69]

Y3 Al2 Ga3 O12 Tb3+

4.12

170

30–170

Liao et al. [65]

SrSnO3

Pr3+

4.03

25

298–373

Wei et al. [70]

Y2 Ti2 O7

Pr3+

5.25

25

289–573

Lei et al. [71]

Gd2 Ti2 O7

Pr3+

4.62

25

289–573

Lei et al. [71]

SnO2

Eu3+

1.83

523 K

293–623

Das et al. [72]

TiO2

Eu3+

2.43

533

307–533

Nikolíc et al. [73]

0.65

Li et al. [68]

summarised in Table 2. In the first four rows of Table 2, luminescence thermometry has been based on the Boltzmann equilibrium mechanism. A non ratiometric approach has been used in Mn2+ -activated Zn2 GeO4 persistent green-emitting phosphors [74]. As the temperature increased, the emission shape of Mn2+ did almost not change, whereas the emission intensity gradually decreased because of the thermal quenching effect. The maximum Sr value of 4.90% K−1 at 373 K. The persistence luminescence in this case has a lifetime in the range of seconds. Other recent works have used the ratiometric approach, and mainly the FIR. For example, [65] reported s highly sensitive FIR thermometer based on a thermal quenching (TQ) mechanism using the Y3 Al2 Ga3 O12 :Pr3 + persistent phosphor, which can generate an intense afterglow that originates from the 4f1 5d1 → 4f2 and 3 P0 → 3 H4 emissions of Pr3+ after ultraviolet excitation. The FIR of the two emissions shows a strong temperature dependence in the range of 30–170 ºC, giving a maximum Sr of up to 4.12% ºC−1 at 170 ºC. The FIR is defined as the integrated intensity ratio of the 4f1 5d1 → 4f2 emission (I1 , 280–450 nm) to the 3 P0 → 3 H4 emission (I2 , 480–490 nm). The 3 P0 state is populated mainly by TQ relaxation. The FIR was fitted using the TQ model. As it is known, Er3+ doped NaYF4 is one of the most efficient upconverting phosphors. Wu et al. [69] reported also its persistent luminescence when excited by X-rays. With this excitation, NaYF4 :Er3+ exhibits long persistent luminescence over almost one day. The persistence luminescence mechanism of NaYF4 :Er3+ probably stems from the fluorine vacancy related defects. The FIR between the two green bands associated with the 2 H11/2 → 4 I15/2 and 4 S3/2 → 4 I15/2 electronic transitions of Er3+ ions in NaYF4 :Er3+ was used to determine the temperature under the X-ray excitation mechanism. Once the X-ray radiation was ceased, NaYF4 :0.5%Er3+ emitted clear persistent green luminescence, with a shape for the luminescence spectrum similar to the shape of the upconversion spectrum obtained upon excitation at 980 nm. The

Critical Analysis of the Recent Advances, Applications and Uses …

351

calculated Boltzmann equation parameters B and ΔE/k equal to be 6.27, and 976 K, respectively, corresponding to the expected ones. The Sr obtained was 976/T 2 and the maximum Sr is 1.06% K−1 at 303 K. The mechanism proposed to explain the whole process of persistent luminescence of NaYF4 :Er3+ is because the excited electrons enter to the conduction band, leading to holes in the valence band. The electrons are then captured by the fluorine vacancyrelated defects, causing the well-known F centres. This is known as the ‘charging’ process. After ceasing X-ray excitation, the electrons will be released gradually from the fluorine vacancy-related defects to the conduction band when heated. The combinations of these released electrons and holes will generate energy and then a transference of this energy to the adjacent Er3+ occurs, which triggers the persistent luminescence. Figure 3 shows the persistent luminescence in Er3+ doped NaYF4 , and the FIR (LIR in the figure) after several times of ceasing the activity of the excitation source.

Fig. 3 a Normalized persistent luminescence spectra of NaYF4 :0.5%Er3+ at different temperatures after exposure to X-ray excitation for 5 min, 10 min and 20 min. b FIR (LIR in the figure) between the 527 and 540 nm bands shown in (a) as a function of temperature. c Logarithm of FIR as a function of the reciprocal of temperature. Reproduced with permission from Wu et al. [69]

352

M. C. Pujol Baiges and J. J. Carvajal Martí

Another example of FIR based on persistent luminescence, is the use of SrSnO3 : Pr3+ phosphors with red long afterglow. Temperature sensors were prepared by the albumin assisted sol–gel method. These phosphors have an average decay time of 74.25 s. Pr3+ ions not only act as activators, emitting the red long afterglow, but also act as trap providers. Moreover, the temperature sensing ability was studied based on the non-thermally coupled intensity ratio of the trap emission and the 3 P0 → 3 H4 transition emission. The absolute and relative thermal sensitivities of SrSnO3 : 0.5% Pr3+ were 0.0374 K−1 and 4.03% K−1 , respectively. The mechanism of the red long afterglow of SrSnO3 : Pr3+ phosphors is based on structural defects acting as hole traps, while cations can act as electron traps. Electrons were excited by 254 nm UV light from the valence band to the conduction band. Because of the existence of electrons or hole traps in these phosphors, electrons or holes are easily captured. When the UV light source was turned off, the electrons or holes in these traps were released again by thermal activation, and then the electrons are transferred to the 1 D2 levels of the Pr3+ emitting centers, forming the long red afterglow (1 D2 → 3 H4 ). Thus, in this case Pr3+ not only acted as an activator, but also acted as a trap provider.

4 New Strategies to Improve Thermal Sensitivity and Temperature Resolution in Luminescent Thermometers 4.1 Optical Trapping of Luminescent Nanothermometers The isolation of a single luminescent nanothermometer by an optical trap would allow to measure temperature locally, in a less invasive way, and reducing the luminescence background; so, this can be done by the optical trapping phenomenon; described in more detail in Chap. 10. The application of optical trapping for luminescence thermometry began in 2012 when [75] demonstrated that the luminescence of a lanthanide-doped microparticle was temperature-sensitive, and that the particle could be manipulated using an optical trap. The application of optical trapping for luminescence thermometry made also possible to assess the heating produced by an optical trap in the solvent and in an optically trapped cell by measuring the temperature-sensitive luminescence of non-trapped quantum dots dispersed in the medium [76]. Later, in 2015, the first applications of optically trapped luminescent thermometers were demonstrated, like the characterization of the heating and cooling of an optically trapped microcrystal doped with lanthanides by its own temperature-dependent luminescence. It has been possible, for instance, to characterize the heating and optical refrigeration of

Critical Analysis of the Recent Advances, Applications and Uses …

353

a lanthanide-doped optically trapped microcrystal by its own temperature-dependent luminescence [77]. It is also worth to mention that it was possible to estimate the heat produced by an optically trapped gold nanoparticle by measuring its scattering spectrum combining dark field spectroscopy and optical trapping techniques [78]. Finally, another interesting achievement was to measure the heat profile next to a cell undergoing a photothermal treatment through the emission of an optically trapped lanthanide-doped upconverting nanoparticle [79, 80]. Any particle with a refractive index greater than that of the surrounding medium and a non-negligible polarizability can be optically trapped. There are two strategies to optimize the optical trapping and manipulation of nanoparticles: (i) the reduction of the optical trap volume: near-field techniques, like plasmonic tweezers, slot waveguides, photonic crystal resonators or photonic nanojets, offer a significant increase in trapping efficiency and improve the thermal stability of nanoparticles. Photonic nanojets can be easily implemented in optical trapping, using microspheres acting as microlenses, that under certain conditions can produce a narrow beam with a waist under the Abbe diffraction limit. (ii) the modification of the surface of the nanoparticle: the nanoparticle surface can be modified to increase the optical forces, since the magnitude of the optical forces applied on a nanoparticle depends on the ratio of the refractive index of the material that forms the nanoparticle and the refractive index of the medium in which this nanoparticle is encountered. It also depends on the polarizability of the nanoparticle, which is related to its surface charge. In fact, the optical forces acting on a nanoparticle depend more strongly on the electrostatic properties than on their volume. Most temperature sensing experiments can be performed by using a single optical trap that will manipulate the thermometer and excite its luminescence. However, multiple traps, that might be created by merging different laser sources in one objective, by using time-shared traps, by splitting a laser beam into two beams with orthogonal polarization, or by modulating a laser beam by means of interferometric patterns or holographic liquid crystals, would make more complex experiments possible. Among dielectric nanoparticles, only lanthanide-doped luminescent particles have been successfully used for optical trapping of luminescent nanothermometers [80]. However, the relative weak forces caused by the optical trapping of dielectric nanoparticles results in some limitations, since any strong interaction will usually override the weak forces in the optical trap. This can be solved by trapping bigger clusters of nanoparticles, and using microspheres containing nanoparticles inside [81, 82]. However, since the spatial resolution of the temperature measurements depends on the size of the trapped particle, it is desirable to trap the smallest possible particles. One of the smallest single nanoparticles trapped to measure temperature was a Er2 O3 nanoparticle with a size of ~ 150 nm, that was used to thermal imaging a cluster of gold nanorods on a glass substrate [81], with a spatial resolution similar to the size of the trapped nanoparticle.

354

M. C. Pujol Baiges and J. J. Carvajal Martí

While optical trapping relies on the transfer of linear momentum from light to the particle, rotation may also be engineered by transferring angular momentum from light to the particle, which is transformed into mechanical angular momentum. This can be either spin angular momentum (associated with the light’s circular polarization) or orbital angular momentum (associated with inclined wavefronts of light). Particles tend to align with their more polarizable axis parallel to the electric field of light to minimize their potential energy, so their orientation can be manipulated by trapping them using polarized light. In this context, the lag in the rotation due to the viscosity of the medium where the nanoparticles are dispersed may be a good indicator of changes in temperature, if it is known how the viscosity of the medium changes with temperature [83–86]. Thus, the rotational dynamics of the trapped particle can be used to determine the temperature-induced change through the viscosity of the local environment. This method allowed to determine temperature changes with sub-degree accuracy, which is more than one order of magnitude more accurate than measurements done in the same experimental conditions with luminescence-based thermometric techniques by using microparticles [87]. The optical trapping forces experienced by metallic nanoparticles are usually larger than those experienced by dielectric nanoparticles of the same size due to their larger polarizability. It is due to the coupling of the electric field of light and the surface plasmons. When employing plasmonic nanoparticles as nanothermometers, there is a risk that their large light-to-heat conversion may alter the temperature reading by increasing the temperature of the sensor itself. Moreover, the temperature of a plasmonic nanoparticle deposited on a substrate may be different depending on the material that constitutes the substrate, which may act as a heat dissipator and have a particular variation of the refractive index with temperature. For that reason, optical trapping and the isolation of plasmonic nanoparticles increases the reliability of temperature measurements [88]. Anisotropic plasmonic nanomaterials rotate faster than spherical plasmonic materials of the same volume, as they experience an additional torque due to the large scattering at wavelengths close to the trapping wavelength. However, the heating of plasmonic nanomaterials due to the absorption of the trapping laser beam may alter the viscosity of the fluid around them, and therefore their rotation and translation speeds may experience artefacts in temperature estimation [89, 90]. The heat produced by an optically trapped gold nanoparticle can be measured by its scattering spectrum obtained by dark field spectroscopy, as the heating localized around the plasmonic nanoparticle will reduce the refractive index of the surrounding medium, blue shifting the wavelength of the plasmon resonance [78, 86, 88, 89]. The mechanism for the anti-Stokes emission from metal nanoparticles may be electronic Raman scattering or Purcell effect enhanced hot carrier photoluminescence, depending on the plasmonic system. It is short-lived, so it is weak. However, one of the advantages of the anti-Stokes emission of metals is that it only depends on temperature and laser excitation power, so calibration is not essential and artefacts in temperature estimation can be easily prevented. It has an exponential-like shape, and it can be fitted to obtain an estimation of the temperature inside the metal nanoparticle Baffou [91], Carattino et al. [92], Cai et al. [93–95], Jones et [al. 96]. It has

Critical Analysis of the Recent Advances, Applications and Uses …

355

not been combined yet with optical trapping to isolate and manipulate the plasmonic thermometer, however. Also, high temperatures set a challenge for stable optical trapping of particles, as Brownian motion increases, making it easier for the particle to escape from the optical trap. To solve these inconveniences, luminescent nanothermometers must be optimized to maximize optical forces. Moreover, optical tweezers can be improved to achieve larger optical forces by, for instance, decreasing the size trap. Mostly, lanthanide-doped and plasmonic particles have been used for luminescence thermometry in optical trapping. Other materials, such as silicon nanorods, also work as thermometers when optically trapped [97].

4.2 Influence of the Pumping Regime on the Temperature Resolution for Luminescence Thermometry The fluorescence intensity ratio between two emissions arising from thermally coupled electronic levels, and their thermal sensitivity and temperature resolution lead to the material selection for luminescence thermometry straightforward, as one should correctly choose the relation between the energy gap between the emitting levels and the product between the Boltzmann constant and the temperature. However, other effects need to be considered to optimize the temperature resolution and the complexity of the measurement setup. For instance, the absorption of the light by the host material in which the luminescent centres are embedded may cause measurement errors, or non-radiative relaxations may cause measurement errors if the transition rate is comparable to the thermal relaxation rate. Another potential source of error is the pumping regime since it strongly affects the population densities of the lower energy levels during the measurements [32, 33]. Currently, the thermal sensitivity, and consequently the temperature resolution of a luminescent nanothermometer is characterized either under pulsed or continuous excitation. However, no analysis has been undertaken to know if calibration data recorded with either of these pumping configurations are interchangeable. The role of re-absorption caused by a population of lower energy levels has been neglected as a potential bottleneck for high temperature resolution in luminescence thermometry, as well as for material selection for this application. Thiem et al. [32, 33] conducted a study of the time dependent evolution of the population densities in different luminescent thermometers under a pulsed excitation scheme. They found out that the population of the lower energy electronic levels changes with the fluctuation of power and pulse duration of the excitation source, common in currently used pump sources for this kind of applications, such as OPOs or Q-switched lasers, which affects negatively to the temperature resolution. In fact, a population of lower energy levels must be considered as the origin of deviations in the fluorescence intensity ratio, even in the absence of temperature variations, therefore causing errors. To address this question, Thiem et al. used time dependent laser rate

356

M. C. Pujol Baiges and J. J. Carvajal Martí

equations, deriving the corresponding relationships for the population of the lower energy electronic levels. They defined different calibration regimes for luminescence nanothermometry, depending on the pump duration and the luminescence lifetime. They reached the conclusion that either short pulses (< 1 ns) or continuous excitation should be the preferred pumping mechanisms for luminescence thermometry, since both help to improve the temperature resolution, which at the end also influences the sampling rate and the complexity of the measurement setup. The populations of the thermally coupled high energy emission levels reach their respective maxima closely following the excitation pulse, whereas the population of the lower energy levels are shifted in time due to the lifetime of the upper levels. This induces a temperature measurement error caused by the different absorptions of luminescence emission over time. This error does not directly correlate to the actual temperature resolution achievable in a measurement since additional noise sources should also be considered. Instead, this value should rather be considered as a minimum error value. Thus, it becomes apparent that pump rate fluctuations need to be minimized to perform high precision temperature measurements, as can be seen in Fig. 4. Apart from the pump noise, the overall error is also influenced by the strength of the emitted luminescence. Consequently, switching the pumping regime may result in the need to also adjust the pump energy or power, respectively. Short pulse pumping, of the order of 100 ps, can be used for high precision measurements, even without knowing exactly the emission lifetime of the material used. However, especially for cost effective measurements, a continuous excitation is favourable. On the other hand, the integration time necessary to achieve a reliable signal strength using a continuous emitting laser might be not applicable for high sampling rates, with the additional drawback of the heating of the sample generated by this continuous excitation. Thus, both pumping regimes offer advantages and disadvantages, and their usage should be carefully chosen depending on the desired applications. For example, the measurement of rapidly changing temperatures should be conducted with a high sampling rate and therefore pulsed excitation. For monitoring more stable temperatures, continuous excitation may be preferred due to cost effectiveness. Following this analysis, a calibration formalism can be derived to maintain already existing baseline measurements, enabling a change of pumping regime without requiring additional calibration measurements.

4.3 Impact of Noise and Background on Temperature Uncertainty in Luminescence Thermometry Although the comparison of luminescent thermometers based on their relative thermal sensitivity is convenient, this parameter gives an incomplete description of the potential performance of the material for certain applications. The precision of the temperature measurement that can be achieved with a particular luminescent thermometer depends on the relative thermal sensitivity of the material, but also on

Critical Analysis of the Recent Advances, Applications and Uses …

357

Fig. 4 a Temporal evolution of electronic population density calculated for the different energy electronic levels involved in the fluorescence intensity ratio (FIR) for luminescence thermometry measurements when Pr3+ is used as the active ion. b Temperature error reading (ΔT) for FIR measurements as a function of the pump energy fluctuations (ΔP). c ΔT for FIR measurements as a function of the temporal fluctuations of the excitation pulse width (FHWMEXC ) and d achievable temperature resolution depending on pulse duration. Reproduced with permission from Thiem et al. [32, 33]

their luminescence strength compared to measurement noise and background signal. In this context, van Swieten et al. [98] reached the conclusion that by determining the noise characteristics of the instrumentation used for the measurement, the uncertainty of a temperature measurement (δT) can be predicted quantitatively. Up to now δT has been determined by experimentally recording a series of luminescence spectra and calculating the standard deviation of the extracted temperatures [99, 100], as can be seen in Fig. 5. The noise level on a single spectrum can be estimated from fluctuations in the baseline. However, with this procedure, the temperature uncertainty is underestimated because it does not consider the noise on detected photons. These methods, however, are often used in idealized circumstances where background signal is minimal, many particles that act as luminescent thermometers are measured, luminescence is efficiently collected, and long measurement times are used. However, in fact, δT depends

358

M. C. Pujol Baiges and J. J. Carvajal Martí

Fig. 5 Uncertainty of a temperature measurement determined experimentally for core–shell Er,Yb:GdVO4 @SiO2 nanoparticles, that determines the minimum temperature change that these luminescent thermometers can discriminate, calculated after recording 60 consecutive emission spectra at a fixed temperature (313 K), and determining the FWHM of the resulting Gaussian distribution of temperatures determined using the calibration curve. Reproduced with permission from Savchuk et al. [99, 100]

strongly on these circumstances, since it is not an intrinsic property of the luminescent thermometer analysed, in contrast with the relative thermal sensitivity. Consequently, different δT have been reported in the literature for the same luminescent thermometer, while some measurement conditions, such as the environment of the thermometer, for instance, were similar. Normally, in the reported works background fluorescence, blackbody radiation, emissions from higher-excited levels of the thermometer itself, that can interfere with temperature measurements, are not considered. Thus, although subtracting a reference spectrum of any background signal removes the systematic error, its influence on δT remains. This makes a fair comparison of potential luminescent thermometric materials impossible. By using the statistics of photon detection, van Swieten et al. [98] quantified how noise and background signal affect the temperature uncertainty in a measurement of temperature using a particular luminescent thermometer, considering not only the properties of the material itself, but also the characteristics of the detector and the sample. For instance, in a thermometer based on two emitting levels, an increase in temperature affects the relative intensities emitted by these levels, resulting in a change in the intensity ratio in the emission spectrum. For the measurement of these emissions, they normally are spectrally separated by a grating and captured by a CCD, a photomultiplier tube or a photodiode array. The next step in the detection process is the translation of photoelectrons to digital counts for each pixel, which enables the construction of the emission spectrum. The measurement error produced on the determination of this intensity ratio follows from error propagation of the expected counts for each emission with their corresponding variances, and the signal-to-noise

Critical Analysis of the Recent Advances, Applications and Uses …

359

Fig. 6 Temperature uncertainty achieved with a model thermometer consisting of the emissions arising from two emitting levels, as represented in (a). b Simulated luminescence spectrum comprising two Gaussian emission bands spectrally separated with Poissonian detection noise. The inset shows a histogram of the temperatures that are extracted from 10,000 simulated spectra using the ratio of integrated counts, a real temperature of 298 K, and a relative sensitivity of 1% K−1 . The black line is a normal distribution with a mean value of 298 K and a standard deviation that is calculated considering that the random error on the measured temperature only depends on the probability distribution function of the measured intensity ratio and the relative thermal sensitivity of the thermometer. c Same as in (b) but for a total luminescence intensity that is 10 times higher, showing how the temperature uncertainty decreases significantly. Reproduced with permission from van Swieten et al. [98]

ratio. In principle, this error should be lower for a higher number of counts recorded. From another side, since the conversion of the intensity ratio to a temperature value depends on the relative thermal sensitivity, any error on the calibrated relative thermal sensitivity will lead to a systematic difference between the measured and the real temperature with the random error on the measured temperature depending only on the probability function of the measured intensity ratio and the relative thermal sensitivity of the thermometer. Taking all this into account, it can be concluded that the temperature uncertainty decreases as the signal-to-noise ratio increases or the number of counts increases, as it can be seen in Fig. 6, illustrating how experiments performed with a higher number of counts, for example by using longer acquisition times or brighter emissions, have lower temperature uncertainties. The same reasoning can be concluded for temperature measurements based on a spectral shift. CCD cameras conveniently records an entire spectrum within one capture. However, the acquisition of a spectrum with a scanning monochromator and a single-point detector, such as a photomultiplier tube, might lead to additional temperature errors if the excitation intensity fluctuates during the measurement, for instance. To understand the magnitude of the variations in the measured temperature it is necessary to consider the noise generated by the detector used. The main noise sources in a measurement using a CCD camera arise from counting noise due to the statistics of incident photons and readout noise caused by the translation of photoelectrons to digital counts by the analogue-to-digital converter. This kind of error can be accounted by acquiring many reference images, collected when illuminating the CCD camera with white light and summing and plotting the digital counts of pixels with the same mean value, that follows a distribution that approximates to a normal distribution. However, the value that can be extracted is particular for the CCD camera characterised in the particular setting configuration used and cannot

360

M. C. Pujol Baiges and J. J. Carvajal Martí

be extrapolated to all CCD cameras. When accounting for this kind of error to the one considered previously, the decrease of the temperature uncertainty is favoured as the temperature increases. In more sensitive detectors, as electron multiplication CCD cameras, the electron multiplication process is an additional source of error, commonly referred to as the excess noise factor. In that case, electron multiplication gain can only improve a temperature measurement if the number of output electrons is small when compared to other noise sources. In practice, electron multiplication is useful if and only if the signal is weak compared to readout noise. An additional source of error to be considered is the heating of the sample, as normally is observed when experiments are performed at higher laser powers, or also as variations in laser intensity would result in higher count rates coinciding with more laser heating. Strategies as an efficient coating of the active cores of the luminescent thermometers might supress this effect, as can be seen in Fig. 7 for Er,Yb:GdVO4 nanoparticles coated with a SiO2 layer [99, 100]. Here it should be taken in consideration that fluctuations in excitation intensity below the heating threshold do not affect the intensity ratio nor the signal-to-noise ratio. Background emission arising from the surroundings of the luminescent thermometer might be another source of uncertainty which is relevant when the thermometer is used in realistic experimental conditions. Such a distortion of the spectrum affects the temperature uncertainty even after subtraction of the background, since the noise on the background signal cannot be removed. The corrected spectrum thus, contains more noise compared to the background-free spectrum, which is translated into an increased temperature uncertainty. Additional counts from dark current in the photodetector have an equivalent impact on the temperature uncertainty. In practice, however, low levels of background are challenging to avoid completely, especially at high temperatures where blackbody radiation becomes a problem. Thus, by using error propagation, van Swieten et al. [98] quantitatively explained the temperature uncertainties determined by recording a series of spectra. According Fig. 7 Time evolution of the intensity ratio for Er,Yb:GdVO4 and Er,Yb:GdVO4 @SiO2 nanoparticles, showing how bare nanoparticles suffer from laser heating when continuously illuminated with the excitation laser source, while this effect is dissipated by the inert coating. Reproduced with permission from Savchuk et al. [99, 100]

Critical Analysis of the Recent Advances, Applications and Uses …

361

to their results, the precision of a temperature measurement depends not only on the relative thermal sensitivity of the luminescent thermometer considered, but also on the measurement conditions, which makes the comparison of thermometers even more challenging. In fact, temperature uncertainties measured under idealized experimental conditions are difficult to compare among thermometers and may not be relevant for different applications since different experimental setting configurations can yield to different contributions of diverse noise sources. However, in this context, what is always true is that the temperature uncertainty is minimal always for a high relative thermal sensitivity and for a strong luminescence signal. If signals strength is essential for precise temperature measurement, then makes it possible to select other relevant parameters to define a good luminescent thermometer, such as the absorption cross section of the luminescent centre, the photoluminescence quantum yield of the material selected as luminescent thermometer, the detection efficiency of the detector selected, the excitation power, and the integration time of the signal, among others. The values of these parameters, especially those referring to the measurement system, are not intrinsic thermometer properties, but depend strongly on the application for which the luminescent thermometer has been designed. In conclusion, one can convey that experimentally determined temperature uncertainty depends strongly on measurement conditions and is therefore a poor parameter to compare the potential of thermometric materials. In this way, van Swieten et al. [98] were able to propose a guideline on how to compare different luminescent thermometers relevant irrespective of the spectroscopic equipment used or the sample under consideration. This would allow developing and choosing the ideal luminescent thermometer for the desired application.

4.4 Novel Approaches to Achieve Better Thermal Sensitivities Although the work devoted to the use of Ln3+ ions in luminescence thermometry has been carried out for decades, recent years continue to abound with new developments and strategies to improve their thermometric properties. Intensive work is being carried out to increase the sensitivity of luminescent thermometers, especially those based on lanthanide ions through increasing their emission brightness. For that, five different approaches have been explored in the last years, as described in detail in Chap. 2: • The use of transition metal ions together with lanthanide ions: lanthanide ions based luminescent thermometers suffer from a relatively low emission intensity, resulting from the low absorption cross-section of the 4f–4f transitions. Sensitization of Ln3+ luminescence through the energy transfer from transition metals solves these two aspects. The use of transition metals as sensitizers brings several advantages, such as their significantly stronger absorption cross section owing to which they can absorb radiation much more efficiently under the same excitation power density conditions. Moreover, the absorption bands of transition

362

M. C. Pujol Baiges and J. J. Carvajal Martí

metals ions, associated with the d-d electronic transitions are much broader spectrally than the narrow 4f–4f absorption bands of Ln3+ ions, which not only facilitates optical excitation of the phosphor, but also significantly reduces the cost of purchasing an optical excitation source, since there is no need to use a wavelength that matches very precisely the narrow band, as occurs for Ln3+ ions. Finally, the lack of resonance between the excited states of transition metals and Ln3+ allows for an enhancement of the relative thermal sensitivity since the probability of phonon-assisted energy transfer processes between these elements reveals strong thermal dependence. A certain limitation for these systems is the spectral overlap between the Ln3+ emission bands and the transition metals absorption bands. The co-doping of Ln3+ based luminescent thermometers with transition metal ions may beneficially influence the thermometric properties of this type of temperature sensors by enhancing the emission brightness and relative sensitivity. Additionally, this strategy extends the spectral range of optical excitation. However, to match the appropriate transition metal ions with the corresponding lanthanide ions, it is necessary to determine in which spectral region the absorption bands of the transition metals are located, since they are highly susceptible to the crystal field of the host material. In this way, the energy mismatch between the energy levels of the transition metals and the Ln3+ ions can be determined to predict whether energy transfer will be efficient. Finally, it is necessary to optimise the concentration of doping ions for which the energy transfer processes between transition metals and Ln3+ ions are most favourable and dominant over the back energy transfer processes from the Ln3+ ions to the transition metals. Moreover, transition metals have special requirements for the crystallographic positions they can occupy. Therefore, not all host materials are appropriate for transition metal ions in this context. This approach not only leads to the enhancement of the emission intensity and a broadening of the spectral range of optical excitation, but also to an increase of the relative thermal sensitivity. • the use of multilevel thermally coupled systems: a limitation of luminescence thermometers based on the luminescence intensity ratio from two thermally coupled levels is the low maximal achievable value of relative sensitivity, not exceeding 1.5% K−1 . To circumvent this limitation, thermal coupling of three (or even more) energy levels has been proposed. The relative sensitivity of such a system is the sum of the relative sensitivities of the thermometers based on the individual level pairs. The introduction of an additional intermediate energy level between the initial two thermally coupled levels facilitates an increase in the population probability of the upper level through the thermalization process. Consequently, the theoretical maximum ΔE between the initial two thermally coupled levels can be doubled, resulting in a significant increase in sensitivity for such a luminescent thermometer. Relative sensitivities as high as 3.8% K−1 have been reported when applying this approach, well above those predicted for a conventional two thermally coupled levels thermometer. An additional benefit of this kind of thermometers is that their sensitivity increases as the temperature increases, with a better thermometric performance at higher temperatures, resulting from the

Critical Analysis of the Recent Advances, Applications and Uses …

363

higher intensity of the emission generated from the upper energy level. However, in some cases, the low intensity luminescence generated from the upper level might result in a relatively large uncertainty in the determination of the intensity ratio value. This approach can be extrapolated to a multiple array of energy states. Furthermore, by modulating the energy separation, the limits of operating temperatures can be tuned: the larger the energy separation between the extreme electronic levels, the higher the operating temperature can be. Thus, luminescence thermometry based on multilevel thermally coupled systems allows not only to increase the relative sensitivity, but also to extend the useful temperature range over which Boltzmann thermometers can operate. Such optimization of the thermal ranges performed by selecting appropriate energy levels allows to read the temperature over the entire analysed range with a temperature uncertainty below 0.1 K. Another important advantage of the multilevel thermally coupled systems approach is that the emission bands used for temperature determination are, due to the larger energy distance between the analysed electronic levels, spectrally better separated from each other. This facilitates separation of the analysed signals during temperature readout. However, we must keep in mind that the higher the energy separation from the metastable level, the lower the emission intensity from this level. This carries the risk of low accuracy of the temperature readout. • the use of materials that exhibit a thermally induced first-order structural phase transition: In most host materials, parameters such as phonon energy, crystal field strength, crystal symmetry, or local symmetry, i.e., the immediate environment in which the dopant is located, among others, do not depend on temperature. However, in some cases, broad structural polymorphism and temperature-induced first-order structural phase transitions can lead to changes in these parameters that therefore would significantly affect the luminescence properties of Ln3+ ions. When such a structural phase transition takes place, a symmetry breaking process often occurs, which directly affects the spectroscopic properties of materials doped with Ln3+ ions, such as the number of Stark sublevels determined by the crystal field, the probabilities of radiative transitions, and the strength of manifold splitting. The temperature of structural transition between phases, as the particle size of the crystal decreases, is reduced. Despite the high values of relative thermal sensitivities obtained when using such approach, attention should be drawn to the narrow temperature range in which such phase transition based luminescent thermometers can be applied. It is strongly dictated by the phase transition temperature of the material in which the luminescence center is located. Despite the impressive thermal sensitivities and temperature uncertainties, a hysteresis loop occurring for the intensity ratio values within heating and cooling cycles for phase transition-driven luminescent thermometers is nevertheless an important issue to pay attention to. It results from the difference between the endothermic and exothermic reactions occurring during heating and cooling routes, respectively. Therefore, when using such materials for high-sensitivity temperature imaging the monotonic change of temperature in the system (only if temperature rises or only if temperature falls) must be maintained. The main parameter defining the

364

M. C. Pujol Baiges and J. J. Carvajal Martí

application of this type of materials is the structural phase transition temperature, which can be controlled not only by changing the host material itself, but also by the crystal size determined by the synthesis process, and appropriate doping. • the use of an absorption process from the excited state in single band ratiometric based thermometers: the limited number of available pairs of thermally coupled levels limits the number of luminescent thermometers based on two neighbouring thermally coupled energy levels of Ln3+ ions that can be developed, and at the same time narrows the spectral ranges over which they can be developed. Additionally, the close proximity of the position of the emission bands in this type of thermometers can significantly hinder their spectral separation and thus limit the accuracy of the temperature readout. Another limitation for the ratiometric luminescent thermometers is that the dispersive dependence of light scattering and absorption by the medium in which the thermometer is located can also modify the shape of the emission spectrum and thus affect the value of the intensity ratio calculated. These limitations can be addressed by using the excited state absorption process for luminescence thermometry, also known as single band ratiometric approach. This technique takes advantage of the luminescence intensity ratio of a single band of Ln3+ ions upon two different excitation wavelengths (matched absorption from the excited state and from the ground state). This approach brings two benefits: (i) the requirement of having spectrally separated bands for temperature readout is not needed, and (ii) the influence of the interaction of the radiation with the medium is eliminated. An increase in temperature results in an increase in the occupancy of the higher energy level, causing a rise in the luminescence intensity obtained with optical excitation matching the absorption from that level. Consequently, an increase in temperature leads to an enhancement of the signal-to-noise ratio, improving the accuracy of temperature readout in the higher temperature range. The same thermalization process simultaneously reduces the population of the ground level, resulting in a gradually decreasing intensity excited by the ground state absorption. Furthermore, the possibility of analysing changes in luminescence intensity excited from two thermally coupled manifolds of ground level widens the range of ions suitable for luminescence thermometry applications compared to conventional thermometers based only on the coupling of excited levels. By preserving the energy distance between the ground and excited levels at a level comparable to that of conventional thermometers using thermally coupled energy levels, much higher sensitivities can be obtained. There are other parameters affecting the thermal sensitivity that should be considered to select the right host materials for luminescence thermometry using this approach. One of these parameters is the cross-relaxation. Another one is the nonradiative processes depopulating the level from which the excited state absorption occurs. These include both processes related to host phonon energy and processes related to surface effects. The low energy of the matrix phonons favours an extension of the useful temperature range because, for low phonon energies, the probability of nonradiative depopulation of the level from which absorption occurs is reduced, making it sufficiently populated to allow the observation of luminescence

Critical Analysis of the Recent Advances, Applications and Uses …

365

upon excited state absorption excitation already at lower temperatures. Additionally, significantly higher thermal sensitivities are achieved in this case when compared to those obtained in materials with higher phonon energies. Therefore, host materials with low phonon energies, such as fluorides, are preferred to obtain high thermal sensitivities. Surface effects concern surface processes interacting with luminescent ions located nearby the surface layer of the particle. This layer is characterised by a lower degree of crystallinity and a high level of defects. Therefore, by decreasing the nanoparticle size, a decrease in the relative thermal sensitivity values and an increase in the threshold temperature above which excited state absorption induced luminescence is achieved is observed. From another side, to enhance the thermal variation of the intensity ratio, one possibility is increasing the thermal quenching processes of ground state absorption induced luminescence. For that, host matrices for which photoionization or thermal quenching of luminescence via the cross-over transition to charge transfer state occur must be considered. Although in a certain temperature range, this solution allows for an increase in relative thermal sensitivity, since by increasing the rate of temperature quenching of emission with ground state absorption excitation simultaneously reduces the rise in emission intensity with excited state absorption excitation. It is therefore necessary to find a balance between these processes. It has to be also considered when using a system of two pairs of thermally coupled levels in which the thermometric parameter is the intensity ratio between the emission from the lower lying excited level under ground state absorption excitation and the emission from the higher lying excited level under excited state absorption excitation, the predicted thermal sensitivity value is the sum of the sensitivity value of the conventional ratiometric approach of thermally coupled levels and the sensitivity value obtained for the single band ratiometric approach. Nevertheless, the application of the approaches based on excited state absorption is not trivial since it requires keeping the excitation power density of both beams at a constant value. Thus, the main limitation of this approach is that the density of optical excitation reaching the luminescent thermometer has a significant effect on the temperature readout. The fact that such a solution exploits two excitation wavelengths, for which the light extinction coefficient in many media can be significantly different, can make the temperature readout unreliable. However, in media such as air, or using excited state absorption in conventional ratiometric thermometers, very high accuracy can be obtained not only for sensing but also for thermal imaging. The significant influence of the host material parameters on the probability of excited state absorption processes enables the optimization of the thermometric parameters of a luminescent thermometer over a fairly large extent. Importantly, the luminescence obtained with excited state absorption excitation exhibits a positive thermal coefficient, i.e., an increase in the emission intensity with increasing temperature is observed, which is an important advantage to obtain a high signal-to-noise ratio even at high temperatures. • the use of materials with a negative thermal expansion coefficient: host lattice shrinkage at high temperature may induce a decrease in the distance between

366

M. C. Pujol Baiges and J. J. Carvajal Martí

the sensitizer and the activator, enhancing the energy transfer efficiency, which is especially favourable in the case of up-converting phosphors. For that, negative thermal expansion materials are a very promising group of materials from the luminescence thermometric perspective. As the temperature increases, the signal is amplified, preventing the high temperature uncertainty associated with the low noise-to-signal ratio. To take full advantage of the benefits of these materials, the dopant ions should be properly selected and currently the best identified pair of ions is Yb3+ –Ln3+ . To obtain satisfactory thermal sensitivities, a large variation in signal intensity is necessary. However, negative thermal expansion materials tend to be hygroscopic. The absorption of water molecules occurs in the crystallographic voids, leading to an easy loss of the negative thermal expansion properties. When these crystals are hydrated, water molecules can hinder rocking motion and thus inhibit the contraction of the crystal lattice upon heating. From the luminescence point of view, water molecules are high-energy oscillators, which significantly quench luminescence. The presence of water molecules is very undesirable in luminescence thermometry because it not only causes a weakening of the signal intensity, but it can also lead to an erroneous temperature readout due to the removal of OH− groups during the drying of the materials upon heating.

5 Novel Applications 5.1 Biological Applications for Luminescence Thermometry Luminescence thermometry provides very promising results in the field of biomedical applications which require high temperature, spatial and temporal resolutions and high thermal sensitivity. Indeed, temperature is known to play a crucial role in determining dynamics and properties of biosystems such as cell division rates, denaturation processes, etc. High-resolution thermal sensing is required in cancer therapeutic processes as hyperthermia, which is based on externally inducing an increase of tumour’s temperature up to cytotoxic levels (43–45 ºC). From a physiological point of view, living organisms should maintain a reasonable and stable temperature, in which biochemical processes can be hastened. Luminescent nanothermometers with excitation and emission bands in the visible have very limited applicability in vivo due to the large absorption and scattering of visible light by biological tissues. That is why systems operating in the biological windows (BWs) are needed, i.e., in the near infrared (NIR) range. The BWs correspond to the wavelength ranges in which the absorption and scattering of biological tissues are minimum (see Chap. 6), allowing for in vivo fluorescence imaging and sensing at tissue depths up to a few centimetres. When dealing with temperature measurements under in vivo or in vitro conditions, most of the commonly used thermometers are not capable of providing a thermal reading due to either their relatively big size or to the possible invasive and harmful contact that they could have on a living

Critical Analysis of the Recent Advances, Applications and Uses …

367

system. Luminescence thermometry has been recognised as a minimally invasive thermal reading technique suitable for biological applications. Luminescence thermometry has the advantages of remote observation, short acquisition periods and the possibility of having real-time information. In fact, as has been highlighted all along this book, luminescence thermometry seems to be under a rather constant expansion of its applicability in the biological domain. During last years, luminescence nanothermometry has established itself as an unrivalled technique for in vivo remote temperature sensing, a research area where alternative methods (such as magnetic resonance, infrared thermal imaging or ultrasound thermal imaging) have failed to provide accurate, real-time and cost-effective thermal readouts [101]. Chapter 7 provided a good overview of the achievements of luminescence thermometry in this field. At the early infancy of luminescence thermometry, its applicability in living biological systems was restricted to small semi-transparent organisms, such as fly larvae and the nematode Caenorhabditis elegans, using emissions in the visible range. For that, the measurement of the fluorescence polarization anisotropy of the green fluorescent protein was used, allowing mapping the in vivo intracellular temperature due to local heat generation [102]. Rhodamine 800 as reference and a polymer incorporating Eu-tris(dinaphthoylmethane)-bis-trioctylphosphone oxide as the temperature sensing unit were used in a luminescent thermometer designed for the determination of heat produced by the muscle of a living beetle when activated during pre-flight preparation [103] and the endogenous heat production transfer between muscle fibres [104]. Hybrid nanoparticles composed of a core of lanthanide-doped nanoparticles surrounded by a passive shell of the same material and an external carbon shell, capable of self-monitored heating were used to monitor the intracellular temperature in living cells when it was increased up to 60 ºC, inducing cell death [105]. The core of these nanoparticles acted as the temperature sensor based on the spectral analysis of the lanthanide ion emissions, while the carbon shell acted as a heating unit when excited at the right wavelength. In that case, the core and the shell were excited at two different wavelengths. After that, cancer cells were incubated with these nanoparticles into nude mice to induce the growth of malignant tumours and, subsequently, treat them by hyperthermia. This experiment demonstrated that it is possible to have precise control over photothermal treatments by profiting from the multifunctionality of the luminescent nanoparticles. However, the main limitations of this work are the need of two different excitation wavelengths to excite separately the luminescent thermometer and the heater, and the use of a luminescent probe emitting in the visible with a limited penetration depth. A different scenario emerged when emissions in the biological windows were used. As such, the penetration depth of both excitation and emission could be maximised, and, in principle, applications going from the subcutaneous to even deeper levels could be possible. In this context PbS/CdS/ZnS quantum dots with therapeutic and temperature sensing capabilities were used as multifunctional agents. It was possible to use the photothermal conversion efficiency of the quantum dots in such a way that with only one excitation laser beam, the quantum dots could emit light (from which the thermal dynamics of the system could be evaluated) and

368

M. C. Pujol Baiges and J. J. Carvajal Martí

generate heat (which would ablate the tumours) [106]. Merging these two functionalities is also possible in lanthanide-doped nanoparticles [107–109]. In this way, in vivo photothermal therapies could highly benefit from the feedback provided by luminescent thermometers, and the discrepancies between the internal and surface temperatures were evidenced, showing the importance of an accurate temperature sensing at the injection site for the minimization of side effects caused by extra heating, for instance. However, to be more general in terms of application (to deeper tumours, for instance), then compelling strategies need to be developed for targeting the nanostructures to tumours. For instance, nanocapsules composed of magnetic and quantum dot nanoparticles might overcome the general limitations of control over magnetic hyperthermia [110]. These nanocapsules act as multimodal contrast agents under several imaging techniques (MRI, CT, photoacoustic and NIR fluorescence imaging). Here, the NIR fluorescent images of a CD1 mouse in which the nanocapsules were intradermally injected, recorded before and 5 min after the application of an AC magnetic field allowed to evidence a local heating with an increase of temperature close to 2.5 ºC at the site of the intradermal injection (see Chap. 7). Using the emission of Nd3+ , Yb3+ double-doped core–shell nanoparticles it has been possible to record the in vivo subcutaneous thermal transients [111]. When a biological tissue undergoes a thermal relaxation, the cooling dynamics strongly depends on the intrinsic properties of the tissue. Thus, a reliable measurement of the thermal relaxation can provide information about the state of the tissue and be used for the detection of anomalies caused by incipient malfunctions such as dehydration, inflammation, or, even, tumour growth. In this context, while there are other methods to calculate the thermal diffusivity of biological tissues like thermal needles and micro-fabricated thermal sensors, luminescence thermometry shows the benefits of its fast response time and minimal invasiveness. Inspired by this work, the lifetime of the emission of Ag2 S quantum dots was used to detect an inflammation at deeper organs through the determination of the thermal dynamics of a murine liver during inflammation [112]. With this approach, it was possible to get the time course of liver temperature.

5.2 Contactless Luminescence Nanothermometry in the Brain Another area of interest for the application of luminescence thermometry in biological systems is the possibility to measure the temperature in the brain, as it is reviewed in Chap. 9. Brain temperature distribution and thermal dynamics are associated to changes in brain physiology and neural activity, which can be modulated by localised, selective heating or cooling. As pointed out above, luminescence nanothermometry allows for remote temperature mapping at high spatiotemporal resolutions in living organisms, cells and tissues, including the nervous system [113]. So that, it would be an excellent approach to determine the temperature in the brain.

Critical Analysis of the Recent Advances, Applications and Uses …

369

In general, the brain is slightly warmer than the body temperature and has a heterogeneous thermal distribution, with a thermal gradient of up to 1 K between different brain regions [114]. The brain temperature fluctuates in a relatively large range (±3 K) due to changes in wakefulness, behaviours and external stimuli, while blood temperature remains constant [115]. Temperature modulates the capacitance of the neuronal cell membrane and regulates the permeability of certain ionic channels, making it possible to use localised heating to modulate neural activity, as a contactless alternative to electrical stimulation methods to treat neurological disorders [116]. Brain temperature measurement could be very useful in the diagnosis of neurological disorders since it is directly related to the neural activity, and changes in brain thermal dynamics could reflect disease-induced alterations [38, 39, 117]. Beyond diagnosis, local modulation of the temperature of the nervous system has therapeutic applications, ranging from well-established neuroprotective cooling to emerging magneto- and photothermal approaches to modulate neural activity [118]. Brain temperature measurement has attracted increasing interest as a relevant clinical parameter. Harnessing the power of brain temperature to develop new diagnostic and therapeutic strategies for neurological disorders requires a remote, minimally invasive technique capable of real-time sensing with high spatial resolution. Nanoparticle-mediated thermal stimulation (both photothermal and magnetothermal) can increase the selectivity of remote stimulation techniques and minimise off-target heating and potential tissue damage. Luminescence nanothermometry in the second NIR window is a leading candidate for this task, with a capacity of recording subdegree brain temperature changes through the scalp and skull in animal models. Furthermore, luminescence nanothermometry is not limited to point measurements and can allow for real-time sensing at a high spatial resolution. However, the application of luminescence nanothermometry to complex organisms has been limited, partly due to the challenges associated with the delivery of the nanothermometers to the target tissues. In fact, delivering nanothermometers to a particular target area in the brain is technically complex and only small volumes (< 30 μl for mice) can be injected [119]. This complicates imaging at acceptable signal-to-noise ratios for accurate sensing, especially in deep brain locations. The brain has a very large scattering coefficient with a very pronounced wavelength dependence. While transcranial fluorescence imaging is possible across the NIR, fluorescent contrast agents emitting in the II-BW allow higher resolution imaging and sensing since scattering decreases for longer wavelengths. This is essential for brain luminescence thermometry, since the excitation light and the fluorescence emitted by the nanothermometers need to penetrate through the scalp and skull. These nanothermometers also allow sensing with superior signal-to-noise ratios due to the minimal autofluorescence from biological tissues in this spectral range [120]. Conventional preclinical fluorescence imaging instruments are not sensitive to IIBW radiation. This has slowed the progress in preclinical bioimaging and sensing (including nanothermometry) in this spectral range. While many luminescent nanoparticles, including lanthanide-doped nanoparticles and quantum dots of various compositions have a II-BW temperature sensitive

370

M. C. Pujol Baiges and J. J. Carvajal Martí

emission, only Ag2 S quantum dots have been used for brain luminescence thermometry so far operating as intensity-based luminescent thermometers [121]. Some of the findings observed are that when inducing a whole-body cooling of almost 20 ºC, the temperature of the brain dropped only by 3 ºC during the cooling process and returned to its baseline value 30 min after active cooling stopped, while the body temperature (measured with a rectal probe) and the skin temperature (measured with a thermographic camera) dropped by almost 18 ºC and needed about 90 min to recover to baseline values. Also, it was observed that the brain had a much slower cooling rate whereas the skin cooled down up to three times faster.

5.3 Benefits of Lifetime-Based Luminescence Thermometry for Biological Applications Compared with luminescence intensity, luminescence lifetime (LF) is more stable and reliable, independent of probe concentration, excitation light power density and measurement method [75, 122, 123]. Time-resolved imaging technology can distinguish different luminescent probes even if they have the same emission band generated under excitation at the same wavelength, so the time accuracy of the instrument is higher than the spectral accuracy. When the luminescent thermometer is placed in a high temperature environment, the temperature quenching on the luminescence intensity is much greater than that on the lifetime measurement. Therefore, when the temperature quenching of the sample luminescence is strong and the signal-to-noise ratio is very low for the luminescence intensity-based method, the luminescence lifetime signal can still be measured accurately. The luminescence lifetime varies dramatically with temperature within a relatively narrow range of temperatures. Therefore, the applicable temperature range for lifetime measurements is smaller than that for ratiometric luminescence temperature measurements, but it generally has higher sensitivity and is therefore widely welcomed in the field of biological research. The benefits of the use of lifetime luminescence thermometry in biological systems are extendedly reviewed in Chap. 8. Here a short summary of the main aspects analysed in Chap. 8 is offered. Accurate determination of temperature in the body of living systems is nontrivial, as optical transmission of tissues is temperature-dependent and varies from type to type, thus distorting the emitted luminescence spectral shape before it reaches the detector. LF luminescence nanothermometry provides a unique solution to this problem, which is independent of probe concentration, excitation light power density, and biological tissue optical properties. As pointed out above, light in the NIR can solve the problem of rapid attenuation of visible light in biological tissues, thus contributing to the realization of deep tissue imaging. However, the imaging quality and the accuracy of sensing are inevitably affected by the absorption and scattering

Critical Analysis of the Recent Advances, Applications and Uses …

371

of luminescence anisotropy in biological tissues. The intrinsic characteristics of the LF provide an effective means to solve this problem. At present, there is little research on the construction of LF nanoprobes and the advantages of LF imaging. For instance, the luminescence lifetime of N, S co-doped carbon dots, that is temperature-dependent, has been used for NIR luminescence lifetime imaging in cells [90]. The LF of the carbon dots was independent of pH, concentration of carbon dots and ambient ion strength. Apart, carbon dots exhibited an excellent biocompatibility and low cytotoxicity, having also a high thermal stability and reusability. However, the LF of this type of materials is in the nanosecond range, making it difficult to completely distinguish from the auto luminescence arising from the biological tissues. Ultra-long-lived (above 1 s) luminescent nanocapsules, composed of a photosensitiser (PbPc(OBu)8 ), a photoenergy cache unit (PCU), and an emitter (perylene) were developed for LF luminescence thermometry. The photosensitiser absorbs light and converts it to oxygen, producing singlet oxygen. The photoenergy cache unit can react with singlet oxygen and store energy by chemical bonds. The emitter receives energy from the photoenergy cache unit and releases it through radiative transitions [124]. The lifetime of the nanocapsules did not change when the concentration of any of the three units changed, neither when the concentration of nanocapsules changed. This represents a great advantage of LF luminescent nanothermometers, being independent of their own concentration. It was not affected even by the oxygen concentration, the presence of active oxygen species, the excitation power density, and the tissue penetration depth. With the increase of the temperature, the luminescence lifetime of the nanocapsules decreased significantly, which could be interpreted as the acceleration of the photochemical reaction at high temperature. Thus, this ultra-long-lived luminescent nanoprobes have a great potential to be used as luminescence lifetimebased nanothermometers to measure temperature in vivo. In fact, they were used to measure the temperature of a Kunming mouse inflammation model. The luminescence lifetime technique allowed to demonstrate that the average temperature in the arthritis area was 2.6 ºC higher than that in the healthy. Also, because of the ultra-long-lived luminescence process, in vivo LF imaging could be performed using relatively cheap consumer-grade cameras. In addition, a relatively inexpensive low frequency pulsed laser or LED could be used as the excitation source. Such approach would allow attaining simple and inexpensive luminescence lifetime thermometry imaging in vivo systems. There is still a lack of long-lifetime lanthanide ion doped lifetime nanoprobes with luminescence bands located in the II-BW for in vivo temperature sensing research. For this purpose, an inert core/active shell/inert shell structure of tiny nanoparticles (13.5 nm) was developed, in which thermosensitive lanthanide pairs (Ytterbium and Neodymium) were spatially confined in the thin middle shell of sodium yttrium fluoride (1 nm thick), ensuring being homogeneously close to the surrounding environment while protected by the outmost calcium fluoride shell (2.5 nm thick) that shielded external bioactive milieu interferences [125]. Neodymium was used as sensitizer, absorbing the excitation light at 800 nm, and transmitting the energy to Ytterbium, that generated luminescence at 980 nm. This ternary structure enabled

372

M. C. Pujol Baiges and J. J. Carvajal Martí

the luminescent nanothermometers to consistently resolve temperature changes at depths of up to 4 mm in biological tissues in the physiological temperature range. These lifetime-based heat sensitive nanoprobes could be used for in vivo diagnosis of mouse inflammation and drawing the accurate temperature distribution curve of the positioning area of the nanoprobes. The lifetime of these probes was stable under strong laser irradiation, different nanoprobes concentrations, different pH, and had reliable repeatability. The nanoprobes could detect accurate temperature at different tissue depths and performing accurate thermography of the temperature distribution in vivo.

5.4 Luminescence Thermometry for the Internet of Things The Internet of Things (IoT) is beginning to be incorporated into our day-to-day life and extending to all sorts of branches of society, including industry 4.0, healthcare 4.0, and smart cities. IoT can be defined as a network of entities connected to each other’s exchanging information. In it, several technologies play a key role and work together combining information and communication technologies, such as cloud, edge and fog computing, big data, artificial intelligence, and machine learning, cybersecurity and access control technologies, or data transmission with technologies as efficient energy-harvesting and sensors, resulting in cyberphysical systems [126]. One of the main data sources for IoT are sensors, that are present almost everywhere, monitoring different parameters within the network or on its surroundings, either in continuous mode or occasionally after the interaction. Among the distinct sensing technologies, optical sensors appear as real alternatives to electronic ones due to their inherent features such as contactless/remoteness, large-scale measures, faster response times, and immunity to electromagnetic fields, which in some scenarios are more advantageous. An optical temperature sensor for IoT can be taken as a technology whose sensing process generates information that can be processed or transmitted to a commercially available device with an interface connected to a communication network. Smartphones seem to be a key element in the development of mobile optical (mOptica) sensors towards IoT, and in particular, for temperature sensors, as luminescent thermometers, because they are a ubiquitous technology with the necessary processing capacity and, most importantly, they can convert luminescent signals into electronics due to the CCD camera they incorporate. Thus, in this scenario, for what concerns to the application of luminescence thermometry as temperature sensors for IoT, the thermometric parameter can be based on images acquired by a CCD camera. Despite most of the examples reported up to now in this context do not include an IoT connection, the methodology can be easily transposed to optical sensing, with relative thermal sensitivity values comparable to those reported for spectrometer-based luminescence thermometry. Several examples of emission intensity-based luminescent thermometers that correlate the Commission International d’Éclairage (CIE) coordinates or the RGB coordinates to temperature have been reported. Despite not using a smartphone,

Critical Analysis of the Recent Advances, Applications and Uses …

373

Savchuk et al. [99, 100] developed a compact, low-cost and non-invasive temperature sensor using a commercial digital colour sensor, like the ones that can be found nowadays in the smartphone’s CCD cameras. Figure 3 shows the scheme of the temperature sensor proposed. The digital colour sensor, build in a mosaic structure, contain different filters that allows the simultaneous detection of signals in the blue, green and red regions of the electromagnetic spectrum. This temperature sensor was probed for its use in luminescence thermometry in the biological range of temperatures using Er,Yb:NaYF4 , and up to 673 K for microelectronic applications using Tm,Yb:GdVO4 upconverting nanoparticles. In that case, with a microprocessor like the ones that can also be encountered in current technology smartphones, it was possible to calculate the luminescence intensity ratio corresponding to the signals of two of these channels (see Fig. 8 to visualize the working principle of this temperature sensor), and after comparting it to a previously determined calibration curve, the temperature was visualized in the LCD display. This reduced substantially the measurement procedures, times and processing of the luminescent signals used to determine the temperature. The relative thermal sensitivity obtained in both cases was like that previously reported for the same kind of nanoparticles using spectrometer-based techniques. It was also tested in a flexible and transparent polymer composite containing Er,Yb:NaYF4 nanoparticles, based on PDMS, a standard polymer used for the fabrication of microfluidic devices used for biomedical studies, allowing for precise and fast temperature measurements. Othong et al. [127] developed a dual function sensor based on Cd2 (2,5-tpt)(4,5idc)(H2 O)4 to detect the water percentage using the emission colour changes as the temperature changed. Under excitation in the UV, a change of colour from yellow to green, visible at naked eye, was generated. However, in this work, the authors only used the intensity ratio between the R and G coordinates associated with colour change through a smartphone to sense the water percentage. Thus, a thermometric parameter could also have been defined based on these coordinates to quantify temperature using a smartphone. From their side, Kumbhakar et al. [128] used Mn2+ -doped ZnS quantum dots encapsulated into a polymer thin film, with emission bands in the blue and the green to determine the temperature. A temperature change in these materials induced a change in colour perception that could be detected in a photographic record. The authors determined the temperature by the difference between the intensity of the whole emission at room temperature and the intensity at the different temperatures through a custom-made App able to capture photography and measure the colour intensity. Lee and Kim [129] developed a polyethylene glycol-functionalized graphene oxide based colorimetric thermosensor using guanine rich DNAzyme and peptide nucleic acid that changed from colourless to green as the temperature increased from 277 to 353 K. This colour change could be perceived by naked eye and quantified through the images taken with a smartphone together with an image analyser software to determine the green colour coordinate intensity.

374

M. C. Pujol Baiges and J. J. Carvajal Martí

Fig. 8 a Scheme of a temperature sensor for luminescence thermometry incorporating a Hamamatsu S9706 digital colour sensor. b Working principle for this type of temperature sensor, through the calculation of the intensity ratio between the emissions in two of the three detection regions of the digital colour sensor. Adapted with permission from Savchuk et al. [99, 100]

Critical Analysis of the Recent Advances, Applications and Uses …

375

Some examples of ratiometric luminescent thermometers are those developed by Shi et al. [130], that used dihydrophenazine in the temperature range 278–408 K, in which the ratio between the intensities of the orange-red and blue bands were used as the thermometric parameter. The colour change of the emission produced when the temperature increased was visible at naked eye, from the blue at low temperature to orange-red at high temperature. Although the colour quantification was only carried out using spectral data, the broad colour variation in this range of temperatures is remarkable. This would allow developing IoT temperature sensor labels since these luminescent thermometers can be coated onto any object surfaces for simple and fast large-area temperature detection. Ma et al. [131] incorporated temperaturesensitive Zn2+ and Co2+ metal–ligand complexes into polyethylene glycol, in which the Zn2+ complex displays a colour response with temperature changing from blue to yellow as the temperature increases. Piotrowski et al. [132] prepared themochromic luminescent materials codoped with Mn4+ and Tb3+ for temperature sensing through the colour variation from red to green as the temperature increased from 298 to 425 K under UV excitation, using the CIE colour coordinates. The colour variation occurs because the red component of the emission spectra, generated by Mn4+ decreases as the temperature increases, while the emission of Tb3+ is independent of temperature. The colour variation was quantified using photographic records from a digital camera in the RGB colour system. This is a representative example that quantifies emission colour through remote captured images with a smartphone, illustrating its possibility of use in extreme scenarios, outside the expected working regime for a smartphone. Ramalho et al. [133] reported an example in which IoT was explicitly included through recording smartphone pictures of QR codes printed using luminescent ink formed by an organic–inorganic hybrid codoped with Eu3+ and Tb3+ ions. This approach allowed to sense, in real time, the absolute temperature in the 283–317 K range with a temperature resolution of 0.194 K. Processing the material in the form of QR codes gave the opportunity to easily interact with the smartphone, using a custom developed App, providing temperature information in real time, and being able to be inserted into a sensor network. Katumo et al. measured temperature using a phosphor whose lifetime was temperature dependent and was also long enough to be recorded by a 30 fps CCD camera (i.e., one frame every 30 ms). In fact, this is the frequency allowed by recording a video with a smartphone, and its subsequent analysis in a computer that evaluated the intensity variation of the red channel as a function of time [134] allowed to determine temperature. So that, lifetime-based measurements were made using a cheap alternative to thermal cameras, that also present disadvantages of lacking good spatial resolution, calibration, and temperature measurements dependence on the object and the knowledge of their properties (e.g., emissivity). The end-to-end operation idea involving the use of smartphones to quantify temperature without the necessity of having additional equipment might be easily translated to those works using luminescence or thermochromism. The main input for all these systems is photographic records used to analyse the intensity of the pixels to calculate an intensity ratio or a lifetime value. These parameters are traditionally calculated from data acquired with expensive laboratory equipment, and it has

376

M. C. Pujol Baiges and J. J. Carvajal Martí

been demonstrated that it can be done with inexpensive and easily accessible equipment. In a step forward towards the popularisation of luminescence thermometry, photographs of luminescent materials recorded by a smartphone could be used to sense, in real-time, the absolute temperature. The decoding tool and the temperature sensing require only a CCD camera, thus smartphones and closed-circuit television, also known as video surveillance, may be used in their original configuration to sense temperature forming a distributed network of temperature sensors. This would prove that the well-established methods in luminescence thermometry can be used by acquiring an image not compromising the results. This represents an important advance in the field of luminescence temperature sensors. However, mOptical sensing requires developing phosphors with improved optical features suitable to operate in combination with a smartphone or similar devices in the future, acting as a sensing probe by increasing the emission colour change or the lifetime value. In parallel, the excitation and illumination sources are a challenge requiring high-quality LEDs to be used externally or to be incorporated into the smartphone. In addition, despite the high processing capacity of the current smartphones, an optical to electronic signal converter with a quality equivalent to traditional equipment still fails to perform as high quality and reliable excitation and illumination sources. This need will benefit from the noticeable enhancement of the external quantum efficiency of compact and higher output near-UV-emitting LEDs pushing and suit well engineering needs in a bordering field. Thus, luminescent thermometers reveal serious arguments to be an alternative to electronic thermometers, which so far are the dominant technology in the field of mOptical sensing at present and are already moving towards the IoT. When compared with waveguide optical signal systems, despite these systems have the advantage of including several types of sensors with high sensitivity that can be spread across a vast distance, they are mainly stationary because they require the attachment of the optical fibre to the detection and the illumination device as an extra piece of the equipment. For these kinds of thermometers breakthrough into the Internet of Things and our daily life, the use of smartphones seems to be the most logical route, since this technology is an on-growing trend in many different sensing areas. This is due to their widespread availability and, more importantly, their ability to perform the transitions from optics to electronics, which is, until now, one key base topic. This would allow luminescent thermometers to be part of large and complex sensor networks, leaving specific equipment for specific niche applications. Up to now, progress has been made in introducing mOptical sensors on the IoT, even though none of the examples reported up to now is robust enough for a real scenario application. The strengths of the examples reported up to now are a broad colour variation to reduce readout errors, a broad temperature range to maximize the number of applications, the use of primary thermometers that allow recording the measurements without any prior calibration, and the ability to excite the luminescent thermometers without any external illumination source but the smartphone LED itself. Having a temperaturesensing probe with these features and working in combination with a user-friendly App able to process the data collected (image or video) to quantify temperature would create a robust thermometer capable to be operated by both individuals and

Critical Analysis of the Recent Advances, Applications and Uses …

377

industries in a multitude of different areas, from industry 4.0, healthcare 4.0 to smart cities.

5.5 Self-assessed Photothermal Therapy Control of the temperature in ultrasmall areas is a fundamental precondition for understanding and using microscopic thermal processes and is especially useful in some biological applications such as hyperthermia therapy, i.e., photothermal therapy. Photothermal therapy releases heat through the process of photothermal transformation, achieves local hyperthermia for the tumour tissue within the 41– 48 ºC range to ensure the lethal temperature for cancer cell and the safety for healthy tissue. Therefore, a valuable photothermal therapy probe should satisfy the following basic characteristics: 1. Deep penetration enables treatment inside the human body rather than in the skin surface; 2. Efficient light-to-heat transformation validates the effective photothermal therapy by using low laser power and thus suppressing the overheating of the healthy tissues; 3. Small particle sizes accelerate the effective excretion of the nanoprobes after photothermal therapy to avoid in vivo toxicity and are also useful for intracellular studies and applications; 4. Temperature monitoring provides real-time feedback of the photothermal therapy. Safety issues and the lack of self-temperature-feedback restrain the use of Au and semiconductor nanoheaters from practical photothermal therapy applications. Ln3+ heavily doped nanoparticles, alternatively, show great potential as effective photothermal therapy probes, as has been already mentioned in Chapter 4. Using the strong self-heating and the temperature sensing abilities, Ln3+ heavily doped nanoparticles can be therefore perfect self-monitored photothermal therapy probes. Other advantages of their use are their low toxicity, and the adjustable emission wavelength within the biological window regions by rationally engineering the doping composition and distribution. Self-assessed photothermal agents are a sort of optically excited nanoparticles capable of simultaneous heat generation and remote thermal sensing [135]. They constitute a unique platform for developing minimally invasive, cost-effective and fully controlled hyperthermia therapies. Particularly, when dealing with the in vivo thermal treatment of tumours, self-assessed photothermal agents make possible realtime control over the intratumoral temperature that is required to be found in the therapeutic range. This, in turn, avoids the extremes of having either insufficient or excessive heating, which could lead to either inefficient therapy or collateral damage, respectively. When using self-assessed photothermal agents both heating and luminescence-based temperature monitoring are triggered by a single laser beam,

378

M. C. Pujol Baiges and J. J. Carvajal Martí

so that intratumoral heating and temperature sensing could be achieved straightforwardly. There exist several examples of self-assessed photothermal agents based on either lanthanide-doped nanoparticles or infrared-emitting quantum dots. These systems combine high radiation-to-heat conversion efficiencies with appreciable relative thermal sensitivities.

5.6 Luminescence Thermometry in Engineering Applications Temperature measurements are needed in a wide variety of mechanical processes and devices, including batteries, photovoltaic cells, gas turbines, lubrication, catalysts, microreactors, etc. The rate of many of these processes are linked to their real working temperature, that influences, among other things, the physical properties of elastic or viscoelastic solids and fluids, the rate of phase changes and chemical reactions. Thus, luminescence thermometry has been already demonstrated to be useful in engineering for applications such as temperature monitoring to prevent corrosion on stored nuclear waste or in catalyst deactivation in packed bed reactor. It has also been used for the measurement of surface temperature after fuel injection which affects the evaporation of fuel films or the production of pollutants, to name a few. In other situations, in which it is interesting to know the rate of an exothermic or endothermic process, the heat dissipation, the phase change enthalpy, or localise heat sources or sinks, luminescence thermometry can also play an important role to determine the variation of temperature with time, although the temperature information in these cases must be combined with some modelling to extract the desired parameters. The thermal properties of a particular material can also be extracted from temperature measurements, like thermal conductivity, that can be used later to detect defects in materials, for instance, that represent singularities in thermal properties. These measurements can also be used to design thermal protection or active cooling in aerodynamics research, for instance, through thermal and fluid-mechanical design of a wide range of systems. Thus, in general, when talking about the uses of luminescence thermometry in engineering applications we are referring to obtain temperature information in mechanical or chemical processes relevant to energy, transport and manufacturing. Microfluidics is one of these fields in which luminescence thermometry has demonstrated to be a valuable tool. Microfluidic devices have been used as research tools and chemical devices due to their portability, the use of small volumes of chemicals, and the possibility of obtaining flow conditions that are not easy to get in large scale devices due to flow perturbations. In such systems, the small size of the flow channels and their complex geometry difficult the implementation of conventional temperature measurements with thermocouples. Thus, luminescence-based measurements are particularly useful. Organic dyes dissolved in fluid have been used to study heat propagation in electroosmotic flows [136], although a limited number of fluids can be characterised with such luminophores. By using nanoparticles dispersed into the fluid flows, constituted by a gold core a polysiloxane shell containing fluorescein

Critical Analysis of the Recent Advances, Applications and Uses …

379

and a hydrophilic shell, the fluid temperature distribution in cavitating microflows could be determined [137]. In another example, Eu3+ containing nanoparticles were used to determine the temperature distribution inside the anode liquid flow of a micro methanol fuel cell [138], a crucial parameter to understand heat transfer and be able to design systems that maintain the cell operation temperature at an optimum level of efficiency. The same nanoparticles were also used to study acoustic tweezing in microparticles three dimensionally [139], an important analysis for biomedical applications, but in which the heat generated by the acoustothermal effect might alter living cells. Er3+ and Yb3+ containing particles were used to determine the temperature generated in a microreactor by the mix of an acid and a base, a highly exothermic reaction [140], showing the versatility of the use of luminescence thermometry in microfluidic devices. In tribology applications, the flow and deformation of a lubricant in bearings facilitates very high rotation speeds between two parts in close vicinity, in which local viscous heating leads to temperature changes that might affect the rheological properties of the lubricant. Core shell CdSe quantum dots dispersed in squalene used as lubricant in a ball-on-disc tribometer have been used to determine the temperature distribution in bearings with a spatial resolution of 10 μm [141]. In that case, however, it was necessary to know first the effect of pressure that would also affect the luminescence signal. For that, first the pressure distribution was measured at a constant temperature, and then, since the pressure field was considered to be temperature independent, the temperature distribution could be measured under non-isothermal conditions. That allowed to determine that the temperature increased by 20 K in the contact area due to shear heating. Monitoring of structural materials allows mitigating the risk of catastrophic failure. Here, luminescence thermometry can be used to identify defects in a sample [142], by detecting variations in its thermal properties. Defects can be associated, for instance, with locally lower thermal conductivities due to the inclusion of air gaps. To probe this potentiality of luminescence thermometry phosphor targets doped with lanthanide and transition metal ions were inserted on the two opposite sides of a semitransparent elastomer and aerogel samples containing air gaps to mimic defects that were later heated. Such thermal probes allowed to determined temperature through lifetime measurements in the range −50 to 200 ºC using different phosphors to cover the full range of temperature. It was observed that the temperature difference between the two sides of the samples scaled up with the defect size. In a similar context, the combination of luminescence thermometry and digital image correlation allowed to simultaneously measure strain field and temperature during thermomechanical testing [143]. Long-term monitoring of nuclear waste containers is necessary to know their mechanical integrity. Temperature monitoring is a good indicator for this purpose as it points to potential troublesome locations and reactive containers in the storage facilities, and at the same time, provides information to perform corrosion modelling. Luminescence thermometry fprobed to be a better alternative than IR thermometry,

380

M. C. Pujol Baiges and J. J. Carvajal Martí

because the corrosion of the containers changed their emissivity making the calibration of the IR thermometer to be not adequate, and providing errors in the determination of temperature of around 10 K. Especially bright transition metal doped luminescent nanoparticles were used to determine the temperature in NaOH fuel ponds and dry stored containers [144] with temperature uncertainties below 1 K. This probes the possibility to use luminescent thermometers in challenging and optically inaccessible industrial environments, independent of reflected radiation, surface emissivity, contact issues, electrical interference, attenuation by intervening material, and ionizing radiation. Also, luminescence thermometry was probed to be useful to measure temperature within or directly at the surface of a material undergoing a chemical reaction. Luminescent particles pressed onto the surface of fibreboards and polymethylmethacrylate were used to determine temperature up to 600 ºC as these materials were burnt [145, 146]. They have also been used to characterise the combustion of coke particles [93, 94] since it is strongly influenced by the ash layer formed surrounding the coke particle that remains intact during passage through the furnaces used in high-temperature material processing industries and alters oxygen diffusion and heat transfer to the reacting coke particle core. Since the thermal conductivity of the ash layer depends on ash porosity, sintering, temperature, etc., the objective of the study was to determine the temperature at the surface of this ash layer during combustion using lifetime-based luminescence thermometry. In-situ temperatures up to 950 ºC were measured with an accuracy of 12 ºC, identifying the core-surface temperature difference and the direction of the heat transfer. Luminescent thermometers have also proved to be useful in the understanding of energy localisation within explosive materials subjected to mechanical excitation such as impact or ultrasound [147]. Such studies are necessary to improve the safety and performance of explosive materials. The objective of this study was to visualise the initiation and growth of hot spots in space and time to understand the role of heat transfer and discontinuities in these materials. For that ZnO:Ga luminescent particles were sandwiched inside a binder block of octahydro 1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) crystals. The whole block composite polymer-bonded explosive was coupled to an ultrasound transducer to trigger ignition. Images of evolving temperature fields during the formation of hot spots allowed to determine that temperature raised over 100 K after ignition specifically at the location of the HMX crystals. Luminescence thermometry has also been used in the monitoring of the thermal processing of metals by measuring the temperature of aluminium alloy billets during preheating before forging [148] and induction-heated superalloys [149], since heating to an accurate temperature is critical for producing specific alloys for aircraft applications. Again, in this context, the poor contact, conduction heat loss and point measurements are drawbacks for thermocouples, and stray radiation and unknown surface emissivity are drawbacks for IR thermometry, that luminescence thermometry can overcome. In fact, luminescence thermometry can be used to calibrate emissivity for IR thermometers, including how this parameter changes as the temperature increases since metallic surfaces might result altered by heat treatment. In catalysis, in which the temperature of the reactor must be closely controlled to favour the right chemical route preventing at the same time the deactivation of the catalyst particles, temperature mapping is important,

Critical Analysis of the Recent Advances, Applications and Uses …

381

which can also serve to observe the reaction front dynamics to design better dynamic control. Luminescence thermometry has been used both to verify the temperature of the heating spiral [140] and to track the reaction from dynamics [150] by mixing the luminescent thermometers with commercial Zeolite catalyst particles. Temperature is also a critical parameter to determine the mechanical properties of additively manufactured components, like in laser sintering and 3D printing. Laser sintering is a complex process involving transient and cyclic heat transfer, phase change processes, and multiple materials and processing parameters (laser power, writing speed, particle size, thermal and mechanical properties and geometry). It is necessary to know these parameters to manufacture components with novel geometries and appropriate structural integrity in a reliable way. For that Zn-based phosphates and aluminates doped with Mn have been used as luminescent thermometers to determine the local temperatures reached in the part that is being fabricated [151]. Luminescence thermometry has also been applied to photovoltaic cells, in which temperature directly impacts on their efficiency. In that case, the encapsulation applied to the photovoltaic cells limits the use of contact sensors like thermocouples, and causes an attenuated transmission of IR radiation, limiting also the use of IR thermometry. Instead, luminescent thermometers can be applied directly on the backside of the photovoltaic element and the external encapsulation [152]. As a result, it has been observed that there is a significant thermal gradient between the inside and outside parts of the photovoltaic cell of around 10 K, owing to the low conductivity of the PDMS encapsulation layer. These results indicate the need for internal temperature measurements to correctly and quickly identify faults inside the device, and the potential presence of hot spots due to localised high current densities. In gas turbines the temperature of the gases and walls are critical since it strongly affects to the combustion stability, generation of NOx , the overall engine efficiency and the operational lifetime of components exposed to extraordinary thermal loads. Luminescence thermometry here has been used to understand the fluid mechanics and heat transfer and use the information to perform numerical simulations and to evaluate film cooling geometries under realistic conditions of main flow turbulence and combustion to improve their effectiveness using less air [153, 154]. Precise (± 18 K) single shot measurements of the cooling jet temperature were obtained, and combined with velocity measurements, allowed calculation of the two-dimensional turbulent heat flux, a nearby unique capability provided by this thermometric technique. Different film cooling geometries could also be studied [155, 156], and flame-cooling air interactions in model gas turbine combustors [157]. Thus, luminescent thermometers provided valuable information in many types of reactors, heat exchangers, fluid pipelines, engines, etc. where there is a strong coupling between the turbulent flow, mixing and heat transfer adjacent to the wall. In reciprocating engines, where gases reach high pressures and temperatures during the compression stroke, these conditions promote fast rates of exothermic reactions triggered by the addition of fuel in diesel engines or by the ignition spark in gasoline engines. However, these engines suffer from significant heat losses to the walls due to the reduced reactor volume, that contribute to degrade the engine efficiency and promote the production of pollutants. In this context, luminescence thermometry has been used

382

M. C. Pujol Baiges and J. J. Carvajal Martí

to understand the temperature distribution of the gas phase inside the cylinders [158] helping to control the combustion timing and minimising pollutant emission. It was possible to observe how the temperature advanced towards the cylinder head during the compression stroke, but with a relatively uniform temperature increase during the compression process. Owing to their high thermomechanical stability, inorganic luminescent nanoparticles like YAG:Dy, YAG:Pr and BAM:Eu, that support temperatures above 1500 K, can be used in such applications, and they even allow to measure post-reaction events. Luminescence thermometry has proved to be also useful in aerodynamics, in which fluid flow exchanges momentum with the solid body results, among other phenomena, in heat transfer due to the flow-induced friction. For these specific applications, temperature sensitive paints have been traditionally used [61], just by analysing the change of the luminescence intensity of the band emission of the active centres and calculating a ratio between the luminescence image acquired in the presence of the supersonic flow with that of the image obtained under no-flow conditions. Thermal history paints, made of luminescent materials that modify and retain their luminescence properties when exposed to high temperatures, can be used to measure temperature when there is no optical access to the part of interest in an industrial system. They have benefits when compared to thermocouples due to the possibility of measuring temperature on rotating parts with a spatial distribution, for instance. Lifetime based thermal history paints have been used to measure the temperature distribution on the flame tube of a gas turbine combustor [159] and on the turbine casing of an automotive turbocharger [160], operated for tens of hours before examining the luminescence of the paint. In conclusion, luminescence thermometry proved to be very useful in analysing a diversity of chemical and mechanical processes where conventional thermometers do not provide the desired functionality. The low cross-sensitivity of the measured temperature to the composition of the surrounding fluid or solid is a major advantage of this technique allowing accurate measurements even in systems where the chemical composition changes. However, owing to a wide variety of luminescence properties, the choice of the luminescent particles plays a decisive role in the thermal relative sensitivity and temperature uncertainty that can be reached with a specific temperature measurement technique in a specific application.

6 Final Reflections: Current Limitations of Luminescence Thermometry and Future Challenges Along all this chapter, several new approaches, novel materials and innovative applications for luminescence thermometry have been summarized. Here we discuss some reflections about the current limitations of this relatively new thermometric technique, and also about their future challenges.

Critical Analysis of the Recent Advances, Applications and Uses …

383

6.1 Current Limitations Lanthanide-doped luminescent nanothermometers have been identified as one of the most sensitive and versatile systems for remote local temperature sensing in a great variety of fields, but especially in nanomedicine. This is mainly due to the possibility of tailoring their absorption and emission bands by conveniently selecting the lanthanide dopants, so that they would both lie within the NIR biological windows (allowing subcutaneous thermal sensing), the low dosage needed (even at the level of a single nanoparticle), low cytotoxicity, easy surface functionalisation, and high thermal sensitivity [161]. By developing core–shell structures, their brightness and thermal sensitivities can be improved, enabling also developing multifunctional platforms, using them for activating chemical therapies or drug delivery and multiplexing sensing processes, and even for achieving treat-and-see therapies [162]. However, the emission of the luminescent ions is susceptible to alteration by some experimental or external factors, whose effects have been traditionally underestimated by the scientific community, generating some sources of artefacts during the sensing process, which might finally result in a false reading from the sensor. Artefacts may arise from system inhomogeneities, from the limitations of the experimental procedure or detection system, from the interference of contaminants or external signals, but also from an unexpected dependence of the system’s response under the experimental conditions [162]. Among, those, there are three aspects that have been detailedly analysed, concluding on which effects they might cause on the temperature readings: • The dependence of the lanthanide ion emission (band shape and/or branching ratios) on the applied excitation power: consequently, special precautions need to be taken when using lanthanide-doped luminescent thermometers to maintain constant the applied excitation power. Nevertheless, significant differences may arise between the applied and the on-target excitation power densities, especially in nanomedicine due to the complicated propagation of light through tissues and fluids. A great variety of processes can lead to significant alterations in the ontarget excitation power density, such as the absorption of the emitted luminescence by the environment, the deformation of the focal volume due to changes of the refractive index and optical aberrations, etc. These modifications may lead to an erroneous measured value of temperature since a fixed applied excitation power density may result in a quite different on-target excitation power density. For instance, if two up-conversion emission bands are used for ratiometric luminescence thermometry arising from two different mechanisms, involving a different number of absorbed photons to generate each emission band, their intensities and consequently the intensity ratio might change as a function of the applied excitation power density. Thus, a reduction in the excitation power density and a temperature increase might have a similar effect on the emission spectra of the lanthanide-doped luminescent thermometers. This drawback could be easily overcome by performing all the measurements using a constant excitation power density. Nevertheless, even

384

M. C. Pujol Baiges and J. J. Carvajal Martí

in the case when the applied excitation power is kept constant, the actual excitation power density reaching the lanthanide-doped luminescent thermometers in a real sample could become altered by experimental parameters typically disregarded in luminescence nanothermometry experiments. For instance, different focus depths in a liquid medium could lead to relevant changes in the local excitation power density due to scattering or medium absorption. These spectral changes produced by artefacts can be falsely attributed to a change in temperature. These effects are so intense that a significant thermal equivalent noise (TEN) appears, that can be defined as the variation of the intensity ratio with the parameter producing the artefact divided by the relative thermal sensitivity of the luminescent thermometer. This TEN can be as large as 28.7 K mm−1 , although it can be reduced for high S R values. The lower the applied excitation power, the more sensitive the spectral shape to the variation of the excitation power is. Thus, the tendency to reduce the applied excitation power in nanomedicine to minimise the side effects of radiation (like extra heating, for instance) would lead to an increased dependence of the spectral shape on depth. The actual power density in the focal volume at a certain depth will mainly depend on the deformation of the focal volume with depth due to optical aberrations. This, in turn, would be mainly determined by the numerical aperture (NA) of the optical system used in the experimental setup. The NA is a parameter that accounts for the maximum deviation angle of the collected light with respect to the optical axis. The larger the objective lens NA, the larger the solid angle of the collected luminescence and the smaller the focus diameter. However, also, the larger the NA, the more pronounced the induced spectral changes and, therefore, the larger the induced TEN. For instance, for intracellular thermal sensing experiments the use of high NA is mandatory as high resolution is required. Thus, such measurements would be more affected by the uncontrolled changes in the on-target power density. In order to prevent lanthanide-doped luminescent thermometers from being affected by this unexpected variation of the emission band shape with the on-target (real) excitation power density, experimental factors enhancing the variation of the excitation power density with the depth in the sample volume must be avoided. This includes avoiding the use of high NA objective lenses and the use of excitation at wavelengths overlapping with solvent absorption bands. Special care should also be taken when interpreting spectra from temperature measurements in media with a relevant concentration of scattering agents (such as inhomogeneous tissues). Another source of artefacts could be the heating generated by the luminescent thermometers themselves due to non-radiative de-excitation mechanisms. If this happens, there would be an apparent heating with increasing the excitation power. • The partial self-absorption of the luminescence by the luminescent thermometers themselves: this factor alters the shape of the spectra of lanthanide-doped luminescent thermometers. This process takes place when the excitation and emission spectra overlap each other and is only fulfilled in the case of optical transitions involving the fundamental energy level, both in absorption and emission processes. It is also known as radiation trapping. This is the case, for instance, of the emission band associated with the 4 F3/2 → 4 I9/2 transition of Nd3+ ions, widely

Critical Analysis of the Recent Advances, Applications and Uses …

385

used for luminescence nanothermometry involving biological applications. If self-absorption takes place, the longer the path of the luminescence through the sample volume, the higher the probability of partial reabsorption, and so the more altered the resulting emission spectrum will be. Consequently, the magnitude of this artefact depends on the experimental conditions used in a particular measurement. The effect of self-absorption could lead to modifications in the emission band shape not related to temperature variations. This could happen when the number of lanthanide-doped luminescent thermometers located between the excitation beam focus (where they are emitting) and the detection system is modified. The magnitude of the spectral modifications caused by self-absorption does not only depend on the concentration of lanthanide-doped luminescent nanothermometers, but also on the doping level of the individual nanothermometers. If the doping level is reduced, then the absorption coefficient of each nanothermometer is also reduced and so the magnitude of self-absorption induced spectral changes. According to these data, for highly reliable temperature measurements, the use of lanthanide-doped luminescent thermometers with low doping levels is advisable; despite low doping levels usually leads to noisy emission spectra. Thus, the TEN due to self-absorption would be reduced at expenses of an increase in signal noise. For the case of the state-of-the-art lanthanide-doped luminescent thermometers employing the emission band of Nd3+ in the first biological window, self-absorption would lead to a significant TEN of 26 K mm−1 . Quantum dots do always present an overlap of their absorption and emission spectra, so they experience an alteration of the emission spectra leading to a red shift as their luminescence is self-absorbed. Similar effects are also expected to be present during luminescence thermometry measurements based on organic dyes. • The optical absorption of the luminescence by the surrounding media: if the absorption of the surrounding media (solvent, biological fluid or tissue) shows a certain overlap with the emission of the employed lanthanide-doped luminescent thermometers, then the resulting luminescence could show changes not related to temperature variations, but, instead, to this artefact. The scientific community is aware of the side effects of excitation of the luminescence at wavelengths at which water absorbs light. For this reason, great efforts have been made during the last few years to replace Yb3+ ions as sensitizers. Water absorption could also produce false reading from luminescent nanothermometers when there is an overlap between their emissions and the absorption of water. As it happened with the self-absorption of the luminescence, the longer the path of the luminescence through the dispersive volume, the more pronounced the optical absorption of the luminescence by the medium will be, and the spectral shape of the emitted light becomes strongly altered in its route from the emitting thermometers towards the detection system. For the state-of-the-art lanthanide-doped luminescent thermometers employing Nd3+ ions, the TEN would be as large as 78 K mm−1 , and for the case of nanothermometers employing Nd3+ and Yb3+ ions, it would be 46 K mm−1 . As an example, in in vitro measurements (such as intracellular temperature measurements) cells are normally placed in water-filled chambers

386

M. C. Pujol Baiges and J. J. Carvajal Martí

with a typical thickness of 0.2 mm. Therefore, if Nd-doped luminescent thermometers are used to measure the temperature at different locations within the chamber, the error induced in temperature estimation due to changes in depth could be as large as 15 K. Nd-doped luminescent thermometers are also widely used for in vivo applications in which the optical path of luminescence can be as large as several millimetres. In this case, the presence of water in the tissues could lead to an error in temperature estimations of the order of tens of degrees. Therefore, an adequate selection of the emission bands employed to avoid overlap with a high gradient of the environment absorption, performing in situ calibration of the luminescent thermometers, and a strict control of the experimental parameters (e.g., the depth of the nanothermometers in the specimen) are mandatory to avoid induced errors in the emission bands susceptible to be affected by this artefact. Figure 9 includes the emission bands of SrF2 luminescent nanoparticles doped with Nd3+ , Yb3+ , Yb3+ -Tm3+ , or Yb3+ -Er3+ , which are the lanthanide ions mostly used in luminescent nanothermometry in in vitro and in vivo experiments. The absorption of human skin (which in the infrared is mainly due to water) is overlapped to these emissions, to show the emission bands susceptible of being altered by the absorption of the environment in biological applications. Only the emission band of Nd3+ at around 1060 nm, among those lying in the windows of transparency of biological tissues, is expected to be free of suffering from any of the artefacts mentioned above.

Fig. 9 Normalized emissions of SrF2 nanoparticles doped with Yb3+ , Er3+ (green), Yb3+ , Tm3+ (red), Nd3+ (purple) and Yb3+ (black), where those emission bands susceptible to undergoing selfabsorption are pointed out with a grilled filling of the curve. Those emission bands whose shape depends on the applied excitation power are indicated by a circular yellow and red warning sign. The absorption of human skin (dashed blue line) is overlapped to show the main absorption bands that could alter the emission of the nanoparticles. Grey backgrounds denote the spectral ranges out of the biological windows of transparency of biological tissues. Reproduced with permission from Labrador-Páez et al. [162]

Critical Analysis of the Recent Advances, Applications and Uses …

387

6.2 Future Challenges As all these artefacts may affect the majority of the known luminescent nanosensors, it is necessary to assess the reliability of potential luminescent nanothermometers to avoid inducing false spectral changes, and thus, false results. For this purpose, simple tests on the luminescent nanothermometers can be performed to identify the sources of artefacts and having more control over the experimental conditions: (i) the emission at different excitation powers must be checked, (ii) the emission at different depths in the sample volume must be obtained (to rule out self-absorption), (iii) calibration procedures ex situ and in situ must be performed to become aware of the influence of the surrounding media on the luminescent thermometers. Additionally, more efficient and tailored luminescent thermometers, more sensitive detectors, and tailored excitation beam geometries [163] would enhance the penetration depth for in vivo applications and reduce noise. In the field of luminescence thermometry, although the research on luminescent primary thermometers is on the starting steps, its potential to predict the temperature in a remote detection is a crucial advantage in terms of avoiding calibration procedures. Future challenges will encompass, also, the use of multivariable parametric analysis [29], using artificial intelligence and machine learning methods [164] to enhance the performance of luminescent primary thermometers. This would permit the calculation of the precision and accuracy, almost absent in the reports published up to now. Moreover, the evaluation of the performance of luminescent primary thermometers should start to embrace intrinsic parameters of the materials, such as their brightness and quantum yield at a given excitation wavelength.

6.2.1

Future Challenges in Engineering Novel Materials for Luminescence Thermometry

In the field of MOF based luminescent thermometers, additional structural parameters must be identified to be able to rationally design new Ln-bearing MOFs acting as luminescent thermometers in a particular range of temperatures. Among these parameters, the choice of the organic ligand is important, and efforts must be focused on the tailoring of the ligand to design MOF luminescent thermometers with targeted performances. Furthermore, the porosity of MOFs must be more exploited to combine the measure of the temperature with other functionalities such as pH measurement, O2 sensing, drug delivery, or magnetic resonance imaging, among others. MOFs are interesting platforms to design multifunctional materials, especially for biomedical applications. The elaboration of such systems will require to control the particle size in the nano-submicron range, the surface chemistry, the stability in physiological media, and the toxicity of the materials. The versatility of MOFs offers a large field of research for the future, such for instance the research of NIR emitting MOFs for

388

M. C. Pujol Baiges and J. J. Carvajal Martí

ratiometric thermometry, that is still at its infancy, although the potentialities of this kind of platforms are very promising. NIR emitting MOFs are good candidates for theragnostic, i.e., combining therapy and diagnosis. Efforts must be made to rationalise the research on highly sensitive luminescent thermometers based on MOFs to predict better their operating ranges, their thermal sensitivities and other thermometric performances. Many parameters can be tuned, such as the ligand nature, the synthetic parameters (solvent, temperature, pH…) to synthesize MOFs which lead to a huge number of crystallographic structures. Thus, it is important to identify structural parameters governing their thermometric performances. Among them, the topology of the inorganic network seems to be one of the most relevant, since it governs most of the energy transfer processes responsible for the temperature dependence of the emissions. Brighter, more temperature sensitive luminescent particles and more sensitive detectors would allow to improve spatial, temporal and temperature resolutions in the future for luminescent thermometers. Also, it must be considered that luminescence thermometry in the engineering fields has not reached the same level of maturity in terms of thoroughness of the uncertainty estimations and the degree of utilisation of the temperature information to extract quantities such as heat generation rates, heat transfer coefficients and thermal conductivities. There are also different levels of maturity in the diverse application fields. So that, while luminescence thermometry in microfluidics or non-destructive testing is still in the demonstration stage, in fields like aerodynamic testing or temperature monitoring is near industrial use. In the future, the measurement of multiple parameters, like velocity, pressure, oxygen concentration, pH, state of oxidation and chemical composition, simultaneously to temperature will become more important in fields like emerging clean technologies for power generation, chemical synthesis, manufacturing, agriculture, water treatment and mobility, for which luminescence thermometry is very well situated. The reliability of current methods of luminescence nanothermometry has been questioned recently [101, 162, 165]. Apart from the light-to-heat conversion in metallic nanoparticles, there are other sources of inaccuracy, such as the difficulty to achieve reproducibility and reduce size distribution on the synthesis of the nanoparticles, that affect the reliability of the temperature measurements with single nanothermometers. The variability of nanoparticles size, doping, shell thickness, etc., makes it challenging to reliably calibrate single particle nanothermometers, as the temperature calibration may change slightly for each single particle [166].

6.2.2

Future Challenges in Biomedical Applications

A way to reach precise and reliable in vivo intratumoral thermal feedback during in vivo photothermal treatments is the use of luminescent nanoparticles capable of multiparametric thermal sensing [101]. The convergence of the different thermal readouts becomes a solid indicator of their reliability. One of the most common pitfalls when dealing with intratumoral thermal reading is the fact that self-assessed photothermal agents are surrounded by tumoral tissues

Critical Analysis of the Recent Advances, Applications and Uses …

389

and, therefore, the detected spectrum will always be modulated by their optical transmission. Reliable temperature reading inside the tumour would require exact knowledge of the optical properties of the tumoral tissue to separate the tissueinduced spectral distortions from those purely caused by temperature variations. Alternatively, reliable intratumoral temperature reading could be achieved by using multiparametric luminescent nanothermometers, capable of providing multiple temperature readouts based on the analysis of different spectroscopic parameters. Synchronised acquisition of multiple temperature readings unfolds a completely new way to check the accuracy of luminescent thermometers following the argument that if the measurements are artefact-free, then all the different temperature readings should converge. If the different temperature readouts diverge, however, it constitutes a clear indication that the luminescent nanothermometer is not working properly, namely that experimental artefacts are crucially affecting the measurements. For instance, by using the simultaneous thermal quenching and spectral shift suffered when the temperature increases in Ag2 S quantum dots, offer the possibility of multiparametric luminescence temperature reading in a single step [101], since it is possible to analyse the thermal dependence of the emitted intensity, the peak position and the ratio between the emitted intensities at the same time. Nevertheless, before extracting temperature information from the emission spectra, it is mandatory to evaluate at which extent they are distorted by the tissue extinction. In fact, depending on the thermometric parameter selected, the acquisition of intratumoral temperature readouts would be contaminated by tissue-induced spectral distortions. This will be more easily noticed in the emitted intensity and the ratio between the emitted intensities than in the peak position, since the first two parameters present numerous ways of being calculated (i.e., there are many options of defining the limits of integration of the emitted intensity, and there are several combinations of wavelengths to compute the intensity ratio). In fact, if this is not considered, discrepancies up to 7 K between the temperature readouts obtained from the analysis of the different spectral parameters have been noticed [101]. That large temperature divergences reduces the reliability of the temperature measurements, as such large uncertainty is not acceptable when dealing with biological systems, where few degrees can have a great impact on the physiological behaviour of the live system under study. To improve the reliability and robustness of the temperature measurements, the presence of tissue-induced spectral distortions should be avoided. This task, however, is not straightforward because of the impossibility of knowing a priori how to compute the emitted intensity and the intensity ratio in such a way that the disagreement of temperature readouts is minimised. Thus, when the measurement has this ambiguous nature, it is mandatory to thoroughly inspect all the possible combinations of thermometric parameters and look for the ones that minimise the discrepancy. Also, a standard should be defined. For instance, if there is only one way of calculating one of the thermometric parameters, such as the peak position, if it corresponds to a temperature readout that is indeed correct, it can be considered as standard to which all the other readouts must be compared. Despite this, even if the universality of the method is not guaranteed, when comparing with other complex approaches that consider all tissue-induced effects, it can still be considered as more practical

390

M. C. Pujol Baiges and J. J. Carvajal Martí

under some circumstances. Once the standard thermal readout has been defined, it is only needed to find the way of computing the other thermometric parameters that minimise the deviation from the standard. The rationale comes from the fact that if the temperature reading from different thermometric parameters agree, this can only mean that either the tissue-induced effects were completely avoided in the selection of the parameters, or the tissues are affecting the parameters in the same form. While the later would imply an improvement in accuracy only, the former would entail an enhancement in accuracy as well. The question concerning which of these two conditions is factual can only be answered by a thorough consideration of the thermal dependence of scattering and absorption of light in several biological tissues. This approach constitutes a safety step for the analysis of future in vivo luminescence thermometry studies. At this point, it is worth emphasizing that, due to stronger attenuation-induced effects, tumours that are more internalized than melanoma will most likely request more elaborated protocols for the analysis of the data. Also, a change in the skin colour might completely change the dependence observed. Thus, future studies might be necessary to completely account for the tissue-induced effects. On another side, luminescence lifetime thermometry has still many gaps. More research and exploration are urgently needed in this field. The NIR emitting lanthanide organic complex luminescence thermometric materials have the advantages of low biotoxicity, high biocompatibility and high temperature sensitivity, which can meet the requirements of luminescence thermometry in biomedical fields. Although temperature measurements based on the luminescence intensity ratio had become the mainstream way, sensitivity, resolution and high temperature detection need still to be improved. In addition, there are also some problems in the stability, reversibility and luminescence efficiency of temperature-sensitive materials. It has become a valuable and practical research direction to search for new luminescent temperature sensing materials with higher sensitivity, higher resolution, wider temperature measurement range and better stability. Commercial instruments capable of II-BW multispectral and hyperspectral small animal imaging entered the market recently. It is expected that this would speed up the progress in luminescence thermometry in the II-BW, which has so far been applied to proof-of-concept experiments that have shown its potential for monitoring therapies based on localised heating [107] and for disease diagnosis based on changes in thermal dynamics [111]. Further developments in imaging technology might expand the application of luminescence nanothermometry in the II-BW to freely moving animals. Luminescence thermometry can overcome the pitfalls of other thermometry techniques (invasiveness, low spatiotemporal resolution) to measure brain temperature. However, still, several challenges remain for it to enable further progress in neuroscience research. Currently, luminescence nanothermometry is unable to detect temperature changes of physiological significance in the brain with sufficient accuracy. Intensity-based measurements have very limited applicability and only allow for recording temperature changes over short periods, during which the nanothermometer concentration can be considered constant. Ratiometric luminescence thermometry could address this issue without requiring major changes to the imaging

Critical Analysis of the Recent Advances, Applications and Uses …

391

approach. However, the wavelength dependence of light-tissue interaction affects to the shape of the emission band, which complicates the use of this ratiometric approach to sense temperature in in vivo experiments. Since light-tissue interaction does not affect the radiative lifetime of luminescent nanothermometers, it might be the preferred luminescence thermometry technique to measure the brain temperature in living organisms, despite requiring a more complex setup for time-resolved detection. From another side, the temperature uncertainty is another issue that may limit the applicability of luminescence nanothermometry in neuroscience. Since physiologically relevant temperature changes in the brain are often smaller than 1 ºC, thermal accuracies equal or better than 0.1 ºC are needed. Brighter luminescent nanothermometers with emissions in the II-BW would allow to obtain improved thermal accuracies and better signal-to-noise ratios. Besides material development, mathematical analysis tools, like multiple linear regression and multiparametric temperature data as explained in Sect. 2.2, might boost also the temperature uncertainty and resolution. Another aspect that needs to be addressed is the delivery of the luminescent nanothermometers into the brain. Up to now, only intracerebral injection to deliver the nanothermometers to a target area in the brain has been used. This process implies drilling a hole in the skull of the animal and slowly injecting a small volume of nanoparticles dispersed in a saline solution. Nanoparticles injected intravenously are not only susceptible to be uptake by the reticuloendothelial system, but also incapable of cross an intact blood–brain barrier. To solve this issue, a method to disrupt blood–brain barrier permeability (like ultrasounds or laser-induced heating) would be required in conjunction to systemic injection. Specific antibodies or peptides conjugated to the nanoparticle’s surfaces can also allow nanoparticles to cross the blood–brain barrier. Intrathecal and intranasal routes, both of which bypass the blood–brain barrier, also allow nanoparticle delivery to the brain. The minimally invasive nature of intranasal administration makes it particularly attractive for nanothermometers delivery. When functional, together with the fundamental neuroscience knowledge gained with luminescence thermometry in the II-BW, could lead to new diagnostic and therapeutic strategies for neurological disorders. Temperature sensitive luminescent nanomaterials have the capability of being successfully applied under in vivo conditions, allowing to develop several proofof-concepts that, in the long term, could be crucial for the development of reliable and minimally invasive methods of therapies and/or diagnostics. Thus, despite much progress has been made, we are still living the early infancy of in vivo luminescence thermometry. To continue with the progress of the field, the correct functionalization of the nanostructures should be made in such a way that they could have specific affinity with the type of tissues where the temperature should be measured. The optical trapping of temperature-sensitive luminescent nanomaterials such as quantum dots or dye-loaded microspheres should also be explored for this purpose. In nanomedicine applications based on optical trapping effects the most competitive wavelength ranges are around 800, 830 and 1064 nm. A very novel approach is

392

M. C. Pujol Baiges and J. J. Carvajal Martí

based on the fact that some modifications to the surface of the luminescent nanoparticles simultaneously improve the optical trapping effect and the luminescent properties. Therefore, the development of reproducible synthesis routes to obtain luminescent nanothermometers monodispersing in size and homogeneity must be a work objective. The inclusion of a smartphone, widely available and easily accessible equipment for science and engineering to use is an important step towards mOptical sensing for the Internet of Things, even though the analysis and excitation are externally provided. Challenges for the future encompass materials science with the development of new materials suitable for mOptical sensing, computer science and engineering to develop new and specific software (mobile applications) to use smartphones as transducers, and the collection and analysis of all the information provided by the thermal probes. Only with the merge of all these technologies towards the same objective will it be possible to progress in this field, creating new alternatives to traditional methods and seizing the opportunities offered by today’s technology to develop systems with relevant societal impact.

References 1. Temperature sensors market-growth, trends, covid-19 impact, and forecasts (2022–2027). P. ID: 453564 (2022) 2. C.D.S. Brites, S. Balabhadra, L.D. Carlos, Adv. Opt. Mater. 7, 1801239 (2019) 3. C.D.S. Brites, E.D. Martínez, R.R. Urbano, C. Rettori, L.D. Carlos, Self-calibrated double luminescent thermometers through upconverting nanoparticles. Front. Chem. 7, 267 (2019). https://doi.org/10.3389/fchem.2019.00267 4. A.M.P. Botas, C.D.S. Brites, J. Wu, U. Kortshagen, R.N. Pereira, L.D. Carlos et al., A new generation of primary luminescent thermometers operating in different media based on silicon nanoparticles. Part. Part. Syst. Charact. 33, 740–748 (2016). https://doi.org/10.1002/ppsc.201 600198 5. A.R.N. Bastos, C.D.S. Brites, P.A. Rojas-Gutierrez, R.A.S. Ferreira, R.L. Longo, C. Dewolf et al., Thermal properties of lipid bilayers derived from the transient heating regime of upconverting nanoparticles. Nanoscale 12, 24169–24176 (2020). https://doi.org/10.1039/d0nr06 989b 6. Y.P. Varshni, Temperature dependence of energy gap in semiconductors. Physica 34, 149–154 (1967). https://doi.org/10.1016/0031-8914(67)90062-6 7. M. Suta, A. Meijerink, A theoretical framework for ratiometric single ion luminescent thermometers-thermodynamic and kinetic guidelines for optimized performance. Adv. Theor. Simul. 3, 2000176 (2020). https://doi.org/10.1002/adts.202000176 8. A.N.C. Neto, E. Mamontova, A.M.P. Botas, C.D.S. Brites, R.A.S. Ferreira, J. Rouquette et al., Rationalizing the thermal response of dual-center molecular thermometers: The example of an Eu/Tb coordination complex. Adv. Opt. Mater. 10, 2101870 (2022). https://doi.org/10. 1002/adom.202101870 9. V. Trannoy, A.N. Carneiro Neto, C.D.S. Brites, et al., Engineering of mixed Eu 3+ /Tb 3+ metalorganic frameworks luminescent thermometers with tunable sensitivity. Adv. Opt. Mater. 9, 2001938 (2021). https://doi.org/10.1002/adom.202001938 10. V. Trannoy, A.N.C. Neto, C.D.S. Brites, L.D. Carlos, H. Serier-Brault, Engineering of mixed Eu3+ /Tb3+ metal-organic frameworks luminescent thermometers with tunable sensitivity. Adv. Opt. Mater., 2001938 (2021). https://doi.org/10.1002/adom.202001938

Critical Analysis of the Recent Advances, Applications and Uses …

393

11. C.D.S. Brites, S.V. Kuznetsov, V.A. Konyushkin, A.N. Nakladov, P.P. Fedorov, L.D. Carlos, Simultaneous measurement of the emission quantum yield and local temperature: the illustrative example of SrF2 :Yb3+ /Er3+ single crystals. Eur. J. Inorg. Chem. 1555–1561 (2020). https://doi.org/10.1002/ejic.202000113 12. J.C. Martins, A.R.N. Bastos, R.A.S. Ferreira, X. Wang, G. Chen, L.D. Carlos, Primary luminescent nanothermometers for temperature measurements reliability assessment. Adv. Photon. Res. 2, 2000169 (2021). https://doi.org/10.1002/adpr.202000169 13. L.D. Carlos, R.A.S. Ferreira, J.P. Rainho, V. De Zea Bermudez, Fine-tuning of the chromaticity of the emission color of organic-inorganic hybrids co-doped with Eu(III), Tb(III), and Tm(III). Adv Funct. Mater. 12, 819–823 (2002). https://doi.org/10.1002/adfm.200290012 14. J.F.C.B. Ramalho, S.F.H. Correia, L.S. Fu, L.L.F. Antonio, C.D.S. Brites, P.S. Andre, Luminescence thermometry on the route of the mobile-based internet of things (IoT): how smart QR codes make it real. Adv. Sci. 6, 1900950 (2019). https://doi.org/10.1002/advs.201900950 15. R.A.S. Ferreira, E. Mamontova, A.M.P. Botas, M. Shestakov, J. Vanacken, V. Moshchalkov et al., Synchronous temperature and magnetic field dual-sensing by luminescence in a dysprosium single-molecule magnet. Adv. Opt. Mater. 9, 2101495 (2021). https://doi.org/10.1002/ adom.202101495 16. J. Drabik, R. Kowalski, L. Marciniak, Enhancement of the sensitivity of single band ratiometric luminescent nanothermometers based on Tb3+ ions through activation of the cross relaxation process. Sci. Rep. 10, 11190 (2020). https://doi.org/10.1038/s41598-020-68145-5 17. K. Trejgis, K. Maciejewsk, A. Bednarkiewicz, L. Marciniak, Near-infrared-to-near-infrared excited-state absorption in LaPO4 :Nd3+ nanoparticles for luminescent nanothermometry. ACS Appl. Nano. Mater. 3, 4818–4825 (2020). https://doi.org/10.1021/acsanm.0c00853 18. K. Trejgis, R. Lisiecki, A. Bednarkiewicz, L. Marciniak, Nd3+ doped TZPN glasses for NIR operating single band ratiometric approach of contactless temperature readout. J. Lumin. 224, 117295 (2020). https://doi.org/10.1016/j.jlumin.2020.117295 19. K. Trejgis, A. Bednarkiewicz, L. Marciniak, Engineering excited state absorption based nanothermometry for temperature sensing and imaging. Nanoscale 12, 4667–4675 (2020). https://doi.org/10.1039/C9NR09740F 20. K.M.N. Souza, R.N. Silva, J.A.B. Silva, C.D.S. Brites, B. Francis, R.A.S. Ferreira et al., Novel and high-sensitive primary and self-referencing thermometers based on the excitation spectra of lanthanide ions. Adv. Opt. Mater. 10, 2200770 (2022). https://doi.org/10.1002/adom.202 200770 21. Y.C. Jiang, Y. Tong, S.Y.Z. Chen, W.N. Zhang, F.F. Hu, R. Wei, H. Guo, A three-mode selfreferenced optical thermometry based on up-conversion luminescence of Ca2 MgWO6 :Er3+ , Yb3 + phosphors. Chem. Eng. J. 413, 127470 (2020). https://doi.org/10.1016/j.cej.2020. 127470 22. I.E. Kolesnikov, D.V. Mamonova, M.A. Kurochkin, E.Y. Kolesnikov, E. Lähderanta, J. Lumin. 231, 117828 (2021) 23. Z. Cao, X. Wei, L. Zhao, Y. Chen, M. Yin, A.C.S. Appl, Mater. Interfaces 8, 34546 (2016) 24. X. Zhou et al., Opt. Mater. 66, 12–16 (2017) 25. L. Marciniak, K. Elzbieciak-Piecka, K. Kniec, A. Bednarkiewicz, Chem. Eng. J. 388, 124347 (2020) 26. K. Kniec, A. Kochanowska, L. Li, M. Suta, L. Marciniak, J. Mater. Chem. C 10, 9275–9286 (2022). https://doi.org/10.1039/D2TC00952H 27. P.Z. Ristíc, V. Ðordevíc, M. Sekulíc, T. Dramícanin, Ž. Antíc, M.D. Dramicanin, J. Lumin. 238, 118306 (2021) 28. K. Trejgis, D. Drami´canin, L. Marciniak, J. Alloys Comp. 875, 159973 (2021) 29. F.E. Maturi, C.D.S. Brites, E.C. Ximendes, C. Mills, B. Olsen, D. Jaque, S.J.L. Ribeiro, L.D. Carlos, Going above and beyond: a tenfold gain in the performance of luminescence thermometers joining multiparametric sensing and multiple regression. Laser Photon. Rev. 15, 2100301 (2021). https://doi.org/10.1002/lpor.202100301 30. J. Liu, X. Yue, Z. Wang, X. Zhang, Y. Xu, J. Mater. Chem. C 8, 13328 (2020)

394

M. C. Pujol Baiges and J. J. Carvajal Martí

31. L. Marciniak, K. Prorok, L. Francés-Soriano, J. Pérez-Prieto, A. Bednarkiewicz, Nanoscale 8, 5037 (2016) 32. S. Thiem, J. Spelthann, A. Neumann, A. Ruehl, D. Ristau, Three-dimensional nanothermometry below the diffraction limit. Opt. Lett. 46, 3352–3355 (2021) 33. J. Thiem, A. Ruehl, D. Ristau, Influence of pumping regime on temperature resolution in nanothermometry. Nanomaterials 11, 1782 (2021) 34. V. Westphal, S.O. Rizzoli, M.A. Lauterbach, D. Kamin, R. Jahn, S.W. Hell, Video-rate far-field optical nanoscopy dissects synaptic vesicle movement. Science 320, 246–249 (2008) 35. X.M. Liu, X.G. Kong, Y.L. Zhang et al., Breakthrough in concentration quenching threshold of upconversion luminescence via spatial separation of the emitter doping area for bioapplications. Chem. Commun. 47, 11957–11959 (2011) 36. F. Vetrone, R. Naccache, V. Mahalingam, et al., The active-core/active-shell approach: a strategy to enhance the upconversion luminescence in lanthanide-doped nanoparticles. Adv. Funct. Mater. 19, 2924–2929 (2009) 37. D.M. Yang, C.X. Li, G.G. Li et al., Colloidal synthesis and remarkable enhancement of the upconversion luminescence of BaGdF5 :Yb3+ /Er3+ nanoparticles by active-shell modification. J. Mater. Chem. 21, 5923–5927 (2011) 38. J. Wang, R.R. Deng, M.A. MacDonald, et al., Enhancing multiphoton upconversion through energy clustering at sublattice level. Nat. Mater. 13, 157–162 (2014) 39. H. Wang et al., Brain temperature and its fundamental properties: a review for clinical neuroscientists. Front. Neurosci. 8, 307 (2014) 40. L. Liu, K. Lu, D. Yan, et al., Concentration dependent optical transition probabilities in ultra-small upconversion nanocrystals. Opt. Express 26, 23471–23479 (2018) 41. L. Liu, Z.S. Sun, C. Ma, et al., Highly sensitive and accurate optical thermometer through Er doped tellurite glasses. Mater. Res. Bull. 105, 306–311 (2018) 42. Y.F. Shang, Q. Han, S.W. Hao, et al., Dual-mode upconversion nanoprobe enables broad-range thermometry from cryogenic to room temperature. ACS Appl. Mater. Inter. 11, 42455–42461 (2019) 43. Z. Chouahda, J.P. Jouart, T. Duvaut et al., the use of the green emission in Er3+ -doped CaF2 crystals for thermometry application. J. Phys. Condens. Matter 21, 245504 (2009) 44. W.L. Lu, L.H. Cheng, H.Y. Zhong et al., Dependence of upconversion emission intensity on Yb3+ concentration in Er3+ /Yb3+ co-doped flake shaped Y2 (MoO4 )3 phosphors. J. Phys. D Appl. Phys. 43, 085404 (2010) 45. V.K. Tikhomirov, K. Driesen, V. D. Rodriguez, et al., Optical nanoheater based on the Yb3+ Er3+ co-doped nanoparticles. Opt. Express 17, 11794–11798 (2009) 46. D. Errulat, R. Marin, D.A. Gálico, K.L.M. Harriman, A. Pialat, B. Gabidullin et al., A luminescent thermometer exhibiting slow relaxation of the magnetization: toward self monitored building blocks for next-generation optomagnetic devices. ACS Cent. Sci. 5, 1187–1198 (2019). https://doi.org/10.1021/acscentsci.9b00288 47. C. Viravaux, O. Oms, A. Dolbecq, et al., Temperature sensors based on europium polyoxometalate and mesoporous terbium metal–organic framework. J. Mater. Chem. C 9, 8323–8328 (2021). https://doi.org/10.1039/d1tc01532j 48. L.X. Peng, Q.Y. Meng, W.J. Sun, et al., Synthesis and optical temperature sensing performance of NaLa(MoO4 )2 :Tb3+ ,Eu3+ phosphors. Ceram. Int. 45, 20656–20663 (2019) 49. X. Lian, D. Zhao, Y. Cui, et al., A near infrared luminescent metal-organic framework for temperature sensing in the physiological range. Chem. Commun. 51, 17676–17679 (2015). https://doi.org/10.1039/C5CC07532G 50. J. Rocha, C.D.S. Brites, L.D. Carlos, Lanthanide organic framework luminescent thermometers. Chem. Eur. J. 22, 14782–14795 (2016). https://doi.org/10.1002/chem.201600860 51. Y. Cui, H. Xu, Y. Yue, et al A Luminescent Mixed-Lanthanide Metal–Organic Framework Thermometer. J. Amer. Chem. Soc. 134, 3979–3982 (2012). https://doi.org/10.1021/ ja2108036

Critical Analysis of the Recent Advances, Applications and Uses …

395

52. A. Kourtellaris, W. Lafargue-Dit-Hauret, F. Massuyeau, C. Latouche, A.J. Tasiopoulos, H. Serier-Brault, Tuning of thermometric performances of mixed Eu-Tb metal-organic frameworks through single-crystal coordinating solvent exchange reactions. Adv. Opt. Mater. 10, 2200484 (2022) 53. T. Xia, Y. Cui, Y. Yu, et al., A luminescent ratiometric thermometer based on thermally coupled levels of Dy-MOF. J. Mater. Chem. C 22, 14782 (2017). https://doi.org/10.1039/C7T C00921F 54. D. Yue, J. Zhang, D. Zhao et al., Ratiometric near infrared luminescent thermometer based on lanthanide metal-organic frameworks. J. Solid State Chem. 241, 99–104 (2016). https:// doi.org/10.1016/j.jssc.2016.06.005 55. D. Zhao, J. Zhang, D. Yue et al., A highly sensitive near-infrared luminescent metal-organic framework thermometer in the physiological range. Chem. Commun. 52, 8259–8262 (2016). https://doi.org/10.1039/c6cc02471h 56. D. Zhao, X. Han, S. Wang et al., 808 nm-light-excited near-infrared luminescent lanthanide metal-organic frameworks for highly sensitive physiological temperature sensing. Chem. A Eur. J. 26, 3145–3151 (2020). https://doi.org/10.1002/chem.201905216 57. L. Armelao, S. Quici, F. Barigelletti et al., Design of luminescent lanthanide complexes: from molecules to highly efficient photo-emitting materials. Coord. Chem. Rev. 254, 487–505 (2010) 58. R.F. D’Vries, S. Álvarez-García, N. Snejko, et al Multimetal rare earth MOFs for lighting and thermometry: tailoring color and optimal temperature range through enhanced disulfobenzoic triplet phosphorescence. J. Mater. Chem. C 1, 6316 (2013). https://doi.org/10.1039/c3tc30 858h 59. L. Li, Y. Zhu, X. Zhou, et al., Visible-light excited luminescent thermometer based on singlelanthanide organic frameworks. Adv. Funct. Mater. 26, 8677–8684 (2016). https://doi.org/10. 1002/adfm.201603179 60. J.F. Feng, T.F. Liu, J. Shi, et al., Dual-emitting UiO-66(Zr;Eu) metal−organic framework films for ratiometric temperature sensing. ACS Appl. Mater. Interfaces 10, 48 (2018). https:// doi.org/10.1021/acsami.8b04889 61. J. Liu, X. Han, Y. Lu et al., Isostructural single- and dual-lanthanide metal-organic frameworks based on substituent-group-modifying tetracarboxylate ligands for ratiometric temperature sensing. Inorg. Chem. 60, 4133–4143 (2021). https://doi.org/10.1021/acs.inorgchem.1c00310 62. S. Wang, J. Jiang, Y. Lu, et al., Ratiometric fluorescence temperature sensing based on singleand dual-lanthanide metal-organic frameworks. J. Lumin. 226, 117418 (2020). https://doi. org/10.1016/j.jlumin.2020.117418 63. Y. Cui, R. Song, J. Yu et al., Dual-emitting MOF-dye composite for ratiometric temperature sensing. Adv. Mater. 27, 1420–1425 (2015). https://doi.org/10.1002/adma.201404700 64. Y. Wan, T. Xia, Y. Cui et al., A two-photon luminescent dye-loaded metal-organic framework for physiological temperature sensing within biological windows. Chempluschem 82:1320– 1325 (2017). https://doi.org/10.1002/cplu.201700438 65. C. Liao, F. Chen, H. Wu, H. Wu, L. Zhang, G. Pan, F. Liu, X. Wang, J. Zhang, Afterglowintensity-ratio-based temperature sensing using a persistent phosphor. J. Mater. Chem. C 10, 11884 (2022). https://doi.org/10.1039/d2tc02417a 66. E. Martín Rodríguez, G. López-Peña, E. Montes, G. Lifante, J. García Solé, D. Jaque, L.A. Díaz-Torres, P. Salas, Appl. Phys. Lett. 111, 081901 (2017) 67. L. Zhao, J. Mao, B. Jiang, X. Wei, Y. Chen, M. Yin, Opt. Lett. 43, 3882–3884 (2018) 68. L. Li, Z. Wu, C. Wang, X. Han, L. Marciniak, Y. Yang, Opt. Lett. 47, 1701–1704 (2022) 69. Z. Wu, L. Li, X. Lv, H. Suo, C. Cai, P. Lv, M. Ma, X. Shi, Y. Yang, L. Marciniak, J. Qiu, Persistent luminescence ratiometric thermometry. Chem. Eng. J. 438, 135573 (2022) 70. M. Wei, S. Feng, X. Tian, C. Ji, Z. Huang, J. Wen, X. Liu, F. Luo, C. Li, J. Li, Y. Peng, Albumin assisted sol-gel synthesized SrSnO3 : Pr3+ red persistent phosphors for temperature sensing. J. Lumin. 239, 118328 (2021). https://doi.org/10.1016/j.jlumin.2021.118328 71. R. Lei, X. Luo, Z. Yuan, H. Wang, F. Huang, D. Deng, S. Xu, Ultrahigh-sensitive optical temperature sensing in Pr3+ : Y2 Ti2 O7 based on diverse thermal response from trap emission and Pr3+ red luminescence. J. Lumin. 205, 440–445 (2019)

396

M. C. Pujol Baiges and J. J. Carvajal Martí

72. S. Das, S. Som, C.Y. Yang, C.H. Lu, Optical temperature sensing properties of SnO2 : Eu3+ microspheres prepared via the microwave assisted solvothermal process. Mater. Res. Bull. 97, 101–108 (2018) 73. M.G. Nikolíc, Z. Antíc, S. Culubrk, J.M. Nedeljkovíc, M.D. Dramícanin, Temperature sensing with Eu3+ doped TiO2 nanoparticles. Sensor. Actuator. B Chem. 201, 46–50 (2014) 74. J. Xue, F. Li, F. Liu, H.M. Noh, B.R. Lee, B.C. Choi, S.H. Park, J.H. Jeong, P. Du, Designing ultra-highly efficient Mn2+ -activated Zn2 GeO4 green emitting persistent phosphors toward versatile applications. Mater. Today Chem. 23, 100693 (2022) 75. P. Haro-González, L. Martínez Maestro, M. Trevisani, S. Polizzi, D. Jaque, J. García Sole, et al., Evaluation of rare earth doped silica sub-micrometric spheres as optically controlled temperature sensors. J. Appl. Phys. 112, 054702 (2012) 76. P. Haro-González, W.T. Ramsay, L. Martínez-Maestro, B. del Rosal, K. Santacruz-Gómez, M. del Carmen Iglesias-de la Cruz, et al., Quantum dot-based thermal spectroscopy and imaging of optically trapped microspheres and single cells. Small 9, 2162–2170 (2013) 77. P.B. Roder, B.E. Smith, X. Zhou, M.J. Crane, P.J. Pauzauskie, Laser refrigeration of hydrothermal nanocrystals in physiological media. Proc. Nat. Acad. Sci. 112, 15024–15029 (2015) 78. A. Andres-Arroyo, F. Wang, W.J. Toe, P. Reece, Intrinsic heating in optically trapped Au nanoparticles measured by dark-field spectroscopy. Biomed. Opt. Express 6, 3646–3654 (2015) 79. H. Rodríguez-Rodríguez, P. Rodríguez Sevilla, E. Martin Rodríguez, D.H. Ortgies, M. Pedroni, A. Speghini, et al., Enhancing optical forces on fluorescent up-converting nanoparticles by surface charge tailoring. Small 11, 1555–1561 (2015) 80. P. Rodríguez-Sevilla, Y. Zhang, P. Haro-González, F. Sanz-Rodríguez, F. Jaque, J.G. Solé et al., Thermal scanning at the cellular level by an optically trapped upconverting fluorescent particle. Adv. Mater. 28, 2421–2426 (2016) 81. S. Baral, S.C. Johnson, A.A. Alaulamie, H.H. Richardson, Nanothermometry using optically trapped erbium oxide nanoparticle. Appl. Phys. A. 122, 340 (2016). https://doi.org/10.1007/ s00339-016-9886-0 82. S. Drobczy´nski, K. Prorok, K. Tamarov, K. Dus-Szachniewicz, V.P. Lehto, A. Bednarkiewicz, Toward controlled photothermal treatment of single cell: optically induced heating and remote temperature monitoring in vitro through double wavelength optical tweezers. ACS Photon. 4, 1993–2002 (2017) 83. A. Lehmuskero, P. Johansson, H. Rubinsztein-Dunlop, L. Tong, M. Kall, Laser trapping of colloidal metal nanoparticles. ACS Nano 9, 3453–3469 (2015) 84. E. Ortiz-Rivero, K. Prorok, M. Skowickł, D. Lu, A. Bednarkiewicz, D. Jaque et al., Single-cell biodetection by upconverting microspinners. Small 15, 1904154 (2019). https://doi.org/10. 1002/smll.201904154 85. E. Ortiz-Rivero, K. Prorok, I.R. Martín, R. Lisiecki, P. Haro-González, A. Bednarkiewicz et al., Laser refrigeration by an Ytterbium-Doped NaYF4 microspinner. Small 17, 2103122 (2021). https://doi.org/10.1002/smll.202103122 86. L. Shao, M. Käll, Light-driven rotation of plasmonic nanomotors. Adv. Funct. Mater. 28, 1706272 (2018) 87. P. Rodríguez-Sevilla, Y. Arita, X. Liu, D. Jaque, K. Dholakia, The temperature of an optically trapped rotating microparticle. ACS Photon. 5, 3772–3778 (2018). https://doi.org/10.1021/ acsphotonics.8b00822 88. K. Setoura, Y. Okada, D. Werner, S. Hashimoto, Observation of nanoscale cooling effects by substrates and the surrounding media for single gold nanoparticles under CW-laser illumination. ACS Nano 7, 7874–7885 (2013) 89. D. Andrén, L. Shao, N.O. Länk, S.S. A´cimovi´c, P. Johansson, M. Käll, Probing photothermal effects on optically trapped gold nanorods by simultaneous plasmon spectroscopy and brownian dynamics analysis. ACS Nano. 11, 10053–61 (2017) 90. S. Kalytchuk, K. Polakova, Y. Wang et al., Carbon dot nanothermometry: intracellular photoluminescence lifetime thermal sensing. ACS Nano 11, 1432–1442 (2017)

Critical Analysis of the Recent Advances, Applications and Uses …

397

91. G. Baffou, Anti-stokes thermometry in nanoplasmonics. ACS Nano. 15, 5785–5792 (2021) 92. A. Carattino, M. Caldarola, M. Orrit, Gold nanoparticles as absolute nanothermometers. Nano Lett. 18, 874–880 (2018) 93. Y.Y. Cai, L.J. Tauzin, B. Ostovar, S. Lee, S. Link, Light emission from plasmonic nanostructures. J. Chem. Phys. 55, 060901 (2021) 94. T. Cai, M. Khodsiani, B. Hallak, C. Abram, F. Beyrau, E. Specht, Phosphor thermometry at the surface of single reacting large-diameter spherical coke particles to characterise combustion for packed bed furnaces. Proc. Comb. Inst. 38, 4225–4232 (2021). https://doi.org/10.1016/j. proci.2020.06.174 95. T. Cai, S. Mohammadshahi, T. Lee, K.C. Kim, Simultaneous measurement of two-dimensional temperature and strain fields based on thermographic phosphor and digital image correlation. Meas. Sci. Techn. 32, 095204 (2021). https://doi.org/10.1088/1361-6501/ac0468 96. S. Jones, D. Andren, P. Karpinski, M. Kall, Photothermal heating of plasmonic nanoantennas: influence on trapped particle dynamics and colloid distribution. ACS Photon. 5, 2878–2887 (2018) 97. P. Karpinski, S. Jones, H. Sipova-Jungova, R. Verre, M. Käll, Optical rotation and thermometry of laser tweezed silicon nanorods. Nano Let. 20, 6494–6501 (2020) 98. T.P. van Swieten, A. Meijerink, F.T. Rabouw, Impact of noise and background on measurement uncertainties in luminescence thermometry. ACS Photon. 9, 1366–1374 (2022). https://doi. org/10.1007/978-3-030-68056-510 99. O.A. Savchuk, J.J. Carvajal, J. Massons, C. Cascales, M. Aguiló, F. Díaz, Novel low-costcompact and fast signal processing sensor for ratiometric luminescent nanothermometry. Sens. Actuat. A 250, 87–95 (2016) 100. O.A. Savchuk, J.J. Carvajal, C. Cascales, M. Aguiló, F. Díaz, Benefits of silica core-shell structures on the temperature sensing properties of Er, Yb:GdVO 4 up-conversion nanoparticles. ACS Appl. Mater. Interfaces 8, 7266–7273 (2016) 101. Y. Shen, H.D.A. Santos, E.C. Ximendes, J. Lifante, A. Sanz-Portilla, L. Monge, N. Fernández, I. Chave-Coira, C. Jacinto, C.D.S. Brites, L.D. Carlos, A. Benayas, M.C. Iglesias-de la Cruz, D. Jaque, Ag2 S nanoheaters with multiparameter sensing for reliable thermal feedback during in vivo tumor therapy. Adv. Funct. Mater. 30, 2002730 (2020) 102. J.S. Donner, S.A. Thompson, M.P. Kreuzer, G. Baffou, R. Quidant, Mapping intracellular temperature using green fluorescent protein. Nano Lett. 12, 2107–2111 (2012) 103. Ferdinandus, et al., Facilely fabricated luminescent nanoparticle thermosensor for real-time microthermography in living animals. ACS Sens. 1, 1222–1227 (2016) 104. T. Miyagawa et al., Glue-free stacked luminescent nanosheets enable high-resolution ratiometric temperature mapping in living small animals. ACS Appl. Mater. Interfaces 8, 33377–33385 (2016) 105. X. Zhu et al., Temperature-feedback upconversion nanocomposite for accurate photothermal therapy at facile temperature. Nat. Commun. 7, 10437 (2016). https://doi.org/10.1038/nco mms10437 106. B. del Rosal, B., et al., Infrared-emitting QDs for thermal therapy with real-time subcutaneous temperature feedback. Adv. Funct. Mater. 26, 6060–6068 (2016) 107. E. Carrasco et al., Intratumoral thermal reading during photo-thermal therapy by multifunctional fluorescent nanoparticles. Adv. Funct. Mater. 25, 615–626 (2015) 108. E.G. Graham, C.M. Macneill, N.H. Levi-Polyachenko, Review of metal, carbon and polymer nanoparticles for infrared photothermal therapy. Nano Life 3, 1330002 (2013) 109. U. Rocha et al., Nd 3+ doped LaF3 nanoparticles as self-monitored photo-thermal agents. Appl. Phys. Lett. 104, 053703 (2014) 110. E. Ximendes et al., Infrared-emitting multimodal nanostructures for controlled in vivo magnetic hyperthermia. Adv. Mater. 33, 2100077 (2021) 111. E.C. Ximendes et al., Unveiling in vivo subcutaneous thermal dynamics by infrared luminescent nanothermometers. Nano Lett. 16, 1695–1703 (2016) 112. Y. Shen, Reliable and remote monitoring of absolute temperature during liver inflammation via luminescence-lifetime-based nanothermometry. Adv. Mater. 34, 2107764 (2022)

398

M. C. Pujol Baiges and J. J. Carvajal Martí

113. D. Jaque et al., Fluorescent nanothermometers for intracellular thermal sensing. Nanomedicine 9, 1047–1062 (2014) 114. K.N. Fountas, et al., Intracranial temperature. Neurocrit. Care 1, 195–199 (2004) 115. J. Delgado, T. Hanai, Intracerebral temperatures in free-moving cats. Americ. J. Physiol. Legacy Cont. 211, 755–769 (1966) 116. S. Luan et al., Neuromodulation: present and emerging methods. Front. Neuroeng. 7 (2014) 117. E.A. Kiyatkin, Brain temperature homeostasis: physiological fluctuations and pathological shifts. Front. Biosci. 15, 73 (2010) 118. E. Colombo et al., Nanoparticles: a challenging vehicle for neural stimulation. Front. Neurosci. 10, 105 (2016) 119. S. Shimizu, Chapter 32—Routes of Administration, in The Laboratory Mouse. ed. by H.J. Hedrich, G. Bullock (Academic Press, London, 2004), pp.527–542 120. S. Diao et al., Biological imaging without autofluorescence in the second near-infrared region. Nano Res. 8, 3027–3034 (2015) 121. B. del Rosal, B., et al., In vivo contactless brain nanothermometry. Adv. Funct. Mater. 28, 1806088 (2018) 122. W.W. Hsiao, Y.Y. Hui, P.C. Tsai et al., Fluorescent nanodiamond: a versatile tool for long-term cell tracking, super-resolution imaging, and nanoscale temperature sensing. Acc. Chem. Res. 49, 400–407 (2016) 123. Z. Ji, Y. Cheng, X. Cui et al., Heating-induced abnormal increase in Yb3+ excited state lifetime and its potential application in lifetime luminescence nanothermometry. Inorg. Chem. Front. 6, 110–116 (2019) 124. X. Su, Y. Wen, W. Yuan et al., Lifetime-based nanothermometry in vivo with ultra-long-lived luminescence. Chem. Commun. 56, 10694–10697 (2020) 125. M.L. Tan, F. Li, N. Cao et al., Accurate in vivo nanothermometry through NIR-II lanthanide luminescence lifetime. Small 16, 2004118 (2020) 126. J.F.C.B. Ramalho, L.D. Carlos, P.S. André, R.A.S. Ferreira, Optical sensing for the internet of things: a smartphone-controlled platform for temperature monitoring. Adv. Photon. Res. 2, 2000211 (2021) 127. J. Othong, J. Boonmak, F. Kielar, S. Youngme, S.C.A. Appl, Mater. Interfaces 12, 41776 (2020) 128. P. Kumbhakar, A.R. Karmakar, G.P. Das, J. Chakraborty, C.S. Tiwary, P. Kumbhakar, Nanoscale 13, 2946 (2021) 129. J. Lee, W.K. Kim, J. Ind. Eng. Chem. 94, 457 (2021) 130. L. Shi, W. Song, C. Lian, W. Chen, J. Mei, J. Su, H. Liu, H. Tian, Adv. Opt. Mater. 8, 1800190 (2018) 131. Y. Ma, Y. Dong, S. Liu, P. She, J. Lu, S. Liu, W. Huang, Q. Zhao, Adv. Opt. Mater. 8, 1901687 (2020) 132. W. Piotrowski, K. Trejgis, K. Maciejewska, K. Ledwa, B. Fond, L. Marciniak, A.C.S. Appl, Mater. Interfaces 12, 44039 (2020) 133. J.F.C.B. Ramalho, S.F.H. Correia, L. Fu, L.M.S. Dias, P. Adao, P. Mateus, R.A.S. Ferreira, P.S. Andre, NPJ Flex. Electron. 4, 11 (2020) 134. N. Katumo, G. Gao, F. Laufer, B.S. Richards, I.A. Howard, Adv. Opt. Mater. 8, 2000507 (2020) 135. A. Nexha, J.J. Carvajal, M.C. Pujol, F. Díaz, M. Aguiló, Short-wavelength infrared selfassessed photothermal agents based on Ho,Tm:KLu(WO4 )2 nanocrystals operating in the third biological window (1.45–1.96 μm wavelength range). J. Mater. Chem. C 8, 180 (2020) 136. D. Ross, M. Gaitan, L.E. Locascio, Temperature measurement in microfluidic systems using a temperature-dependent fluorescent dye. Anal. Chem. 73, 4117–4123 (2001). https://doi.org/ 10.1021/ac010370l 137. F. Ayela, M. Medrano-Muñoz, D. Amans, C. Dujardin, T. Brichart, M. Martini, O. Tillement, G. Ledoux, Experimental evidence of temperature gradients in cavitating microflows seeded with thermosensitive nanoprobes. Phys. Rev. E 88, 043016 (2013). https://doi.org/10.1103/ PhysRevE.88.043016

Critical Analysis of the Recent Advances, Applications and Uses …

399

138. J. Massing, N. van der Schoot, C.J. Kähler, C. Cierpka, A fast start up system for microfluidic direct methanol fuel cells, Int. J. Hydrog. Energy 44, 26517–26529. https://doi.org/10.1016/ j.ijhydene.2019.08.107 139. R. Weser, Z. Deng, V.V. Kondalkar, A.N. Darinskii, C. Cierpka, H. Schmidt, J. König, Threedimensional heating and patterning dynamics of particles in microscale acoustic tweezers. LabChip 22, 2886–2901 (2022). https://doi.org/10.1039/D2LC00200K 140. R.G. Geitenbeek, J.C. Vollenbroek, H.M.H. Weijgertze, C.B.M. Tregouet, A.E. Nieuwelink, C.L. Kennedy, B.M. Weckhuysen, D. Lohse, A. van Blaaderen, A. van den Berg, M. Odijk, A. Meijerink, Luminescence thermometry for in situ temperature measurements in microfluidic devices. Lab Chip 19, 1236–1246 (2019). https://doi.org/10.1039/C8LC01292J 141. T. Seoudi, D. Philippon, N. Fillot, L. Lafarge, N. Devaux, A. Mondelin, P. Vergne, Cdse-based quantum dots as in situ pressure and temperature non-intrusive sensors in elasto hydrodynamic contacts. Tribol. Lett. 68, 73 (2020). https://doi.org/10.1007/s11249-020-01312-x 142. K.E. Mitchell, M. Aryal, S. Allison, F. Sabri, Remote optical detection of geometrical defects in aerogels and elastomers using phosphor thermometry. Opt. Mater. 119, 111378 (2021). https://doi.org/10.1016/j.optmat.2021.111378 143. E.M.C. Jones, A.R. Jones, C. Winters, Combined thermographic phosphor and digital image correlation (TP + DIC) for simultaneous temperature and strain measurements. Strain 58, e12415 (2022). https://doi.org/10.1111/str.12415 144. A. Sposito, E. Heaps, G. Sutton, G. Machin, R. Bernard, S. Clarke, Phosphor thermometry for nuclear decommissioning and waste storage. Nucl. Eng. Design 375, 111091 (2021). https:// doi.org/10.1016/j.nucengdes.2021.111091 145. A. Omrane, F. Ossler, M. Aldén, J. Svenson, J.B.C. Pettersson, Surface temperature of decomposing construction materials studied by laser-induced phosphorescence. Fire Mater. 29, 39–51 (2005). https://doi.org/10.1002/fam.876 146. A. Omrane, F. Ossler, M. Aldén, Two-dimensional surface temperature measurements of burning materials. Proc. Combust. Inst. 29, 2653–2659 (2002) 147. A.D. Casey, Z.A. Roberts, A. Satija, R.P. Lucht, T.R. Meyer, S.F. Son, Dynamic imaging of the temperature field within an energetic composite using phosphor thermography. Appl. Opt. 58, 164320–164325 (2019). https://doi.org/10.1364/AO.58.004320 148. L. Rosso, S. Tabandeh, G. Beltramino, V. Fernicola, Validation of phosphor thermometry forindustrial surface temperature measurements. Meas. Sci. Technol. 31, 034002 (2019). https://doi.org/10.1088/1361-6501/ab4b6b 149. G. Sutton, Imaging luminescence thermometry to 750 °C for the heat treatment of common engineering alloys and comparison with thermal imaging. Int. J. Thermophys. 43, 36 (2022) 150. R.G. Geitenbeek, A.E. Nieuwelink, T.S. Jacobs, B.B.V. Salzmann, J. Goetze, A. Meijerink, B.M. Weckhuysen, In situ luminescence thermometry to locally measure temperature gradients during catalytic reactions. ACS Catal. 8, 2397–2401 (2018). https://doi.org/10.1021/acs catal.7b04154 151. F. Rouzé l’Alzit, T. Bazin, A. Poulon-Quintin, E. Champion, C. Bertrand, T. Cardinal, A. Rougier, I. Mjejri, V. Jubera, M. Gaudon, Thermo chromo luminescent Mn2+ -doped ox-ides as thermal sensor for selective laser sintering. Opt. Mater. 110, 110542 (2020). https://doi. org/10.1016/j.optmat.2020.110542 152. Y. Cao, G. Koutsourakis, G.J.M. Sutton, J.W.E. Kneller, S. Wood, J.C. Blakesley, F.A. Castro, In situ contactless thermal characterisation and imaging of encapsulated photovoltaic devices using phosphor thermometry. Prog. Photovolt: Res. Appl. 27, 673–681 (2019) 153. P. Schreivogel, C. Abram, B. Fond, M. Straußwald, F. Beyrau, M. Pfitzner, Simultaneous khz-rate temperature and velocity field measurements in the flow emanating from angled and trenched film cooling holes. Int. J. Heat Mass Transf. 103, 390–400 (2016). https://doi.org/ 10.1016/j.ijheatmasstransfer.2016.06.092 154. M. Stephan, H. Lee, B. Albert, A. Dreizler, B. Böhm, in Simultaneous Planar Gas-Phase Temperature and Velocity Measurements Within a Film Cooling Configuration Using Thermographic Phosphors. 18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics (Lisbon Portugal, July 4–7 2016)

400

M. C. Pujol Baiges and J. J. Carvajal Martí

155. M. Strauswald, C. Abraham, T. Sander, Time-resolved temperature and velocity field measurements in gas turbine film cooling flows with mainstream turbulence. Exp. Fluids 62, 3 (2021). https://doi.org/10.1007/s00348-020-03087-2 156. M. Strauswald, T. Sander, C. Abram, M. Pfitzner, Investigating the influence of mainstream turbulence and large vortices on gas turbine trenched film cooling flows using planar temperature-velocity imaging. J. Turbomachin. 144, 101013 (2022) 157. M. Greifenstein, Flame–cooling air interaction in an effusion cooled model gas turbine combustor at elevated pressure. Exp. Fluids 60, 10 (2019). https://doi.org/10.1007/s00348018-2656-3 158. A. Kopf, V. Frattina, M. Bardi, T. Endres, G. Bruneaux, C. Schulz, In-cylinder thermographic PIV combined with phosphor thermometry using ZnO:Zn, Int. J. Engine Res. https://doi.org/ 10.1177/14680874211048559 159. D. Peral, A. Zaid, C. Benninghoven, S. Aragúas-Rodríguez, D. Klus, S. Karagiannopoulos, R. Krewinkel, J.P. Feist, High-resolution thermal profiling of a combustor in a non-dedicated test using thermal history coatings. J. Turbomach. 144, 111007 (2022) 160. C.C. Pilgrim, J. Ehrhard, M. Schinnerl, S. Aragúas-Rodríguez, D. Peral, M.S. Straka, M. Genschmar, S. Karagiannopoulos, S.P. Gutierrez, J.P. Feist, Thermal Profiling of Automotive Turbo chargers in Durability Tests, J. Eng. Gas Turbines Power, 144, 021009 (2022). https:// doi.org/10.1115/1.4052497 161. A. Nexha, J.J. Carvajal, M.C. Pujol, Luminescence Nanothermometry and Photothermal Conversion Efficiency for Particles Operating in the SWIR Region, Book: Short-Wavelength Infrared Windows for Biomedical Applications (2021). https://doi.org/10.1117/3.2604326.ch2 162. L. Labrador-Páez, M. Pedroni, A. Speghini, J. García-Solé, P. Haro-González, D. Jaque, Reliability of rare-earth-doped infrared luminescent nanothermometers. Nanoscale 10, 22319–22328 (2018) 163. D. Ning, L. Xu, Y. Zhu, D. Li, H. Jiang, J.J. Carvajal, H. Li, J. Ren, L. Liu, J. Zhang, Bessel beam induced deep-penetrating bioimaging and self-monitored heating using Nd/Yb heavily doped nanocrystals. Appl. Phys. Lett. 121, 04701 (2022) 164. E. Ximendes, R. Marin, L.D. Carlos, D. Jaque, Less is more: dimensionality reduction as a general strategy for more precise luminescence thermometry. Light Sci. Appl. 11, 237 (2022). https://doi.org/10.1038/s41377-022-00932-3 165. D. Lu, M. Pedroni, L. Labrador-Páez, M.I. Marqués, D. Jaque, P. Haro-González, Nanojet trapping of a single sub-10 nm upconverting nanoparticle in the full liquid water temperature range. Small 17, 2006764 (2021) 166. J.R. Casar, C.A. McLellan, C. Siefe, J.A. Dionne, Lanthanide-based nanosensors: refining nanoparticle responsiveness for single particle imaging of Stimuli. ACS Photon. 8, 3–17 (2021) 167. F. Hajizadeh, L. Shao, D. Andrén, P. Johansson, H. Rubinsztein-Dunlop, M. Käll, Brownian fluctuations of an optically rotated nanorod. Optica 4, 746–751 (2017) 168. P. Haro-Gonzalez, L. Martínez-Maestro, I.R. Martin, High-sensitivity fluorescence lifetime thermal sensing based on CdTe quantum dots. Small 8, 2652–2658 (2012) 169. T. Liu, J.P. Sullivan, K. Asai, C. Klein, Y. Egami, Applications of TSP (Springer International Publishing, Cham), pp. 345–398. https://doi.org/10.1007/978-3-030-68056-510 170. A.V. Orlova, V.Y. Kozhevnikova, L.S. Lepnev, A.S. Goloveshkin, I.M. Le-Deigen, V.V. Utochnikova, NIR emitting terephthalates (Smx DyGd1-xy )2 (tph)3 H2 O4 for luminescence thermometry in the physiological range. J. Rare Earth 38, 492–497. https://doi.org/10.1016/ j.jre.2020.01.010

Critical Analysis of the Recent Advances, Applications and Uses …

401

171. L.X. Peng, Y. Zhou, F. Qin et al., High performance FIR thermometry on the basis of the redshift of CTB by dual-wavelength alternative excitation in Eu3+ :YVO4 . Opt. Lett. 46, 5818– 5821 (2021) 172. Z. Wang, D. Ananias, A. Carné-sánchez et al., Lanthanide—organic framework nanothermometers prepared by spray-drying. Adv. Funct. Mater. 20, 2824–2830 (2015). https://doi. org/10.1002/adfm.201500518 173. Y. Zhou, L.X. Peng, P. Tan et al., Luminescence intensity ratio thermometry based on combined ground and excited states absorptions of Tb3+ doped CaWO4 . Opt. Express 29, 22805–22812 (2021)