Nanotechnology for Microfluidics 3527345337, 9783527345335

Citation preview

Nanotechnology and Microfluidics

Nanotechnology and Microfluidics Edited by Xingyu Jiang Series Editors Chunli Bai Minghua Liu

Editor Prof. Xingyu Jiang

National Center for Nanoscience and Technology Southern University of Science and Technology No.11 ZhongGuanCun BeiYiTiao Haidian District 100190 Beijing China

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate. Library of Congress Card No.:

Series Editors Prof. Chunli Bai

applied for

Chinese Academy of Sciences 52 Sanlihe Rd Xicheng District 100864 Beijing China

British Library Cataloguing-in-Publication Data

Prof. Minghua Liu

National Center for Nanoscience and Technology No.11 ZhongGuanCun BeiYiTiao Haidian District 100190 Beijing China

A catalogue record for this book is available from the British Library. Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at . © 2020 Wiley-VCH Verlag GmbH & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law. Print ISBN: 978-3-527-34533-5 ePDF ISBN: 978-3-527-81833-4 ePub ISBN: 978-3-527-81835-8 oBook ISBN: 978-3-527-81834-1

Adam-Design, Weinheim, Germany Typesetting SPi Global, Chennai, India Cover Design

Printing and Binding

Printed on acid-free paper 10 9 8 7 6 5 4 3 2 1

v

Contents Preface xiii

1

1

Micro/Nanostructured Materials from Droplet Microfluidics Xin Zhao, Jieshou Li, and Yuanjin Zhao

1.1 1.2 1.2.1 1.2.2 1.2.3 1.2.4 1.2.5 1.3 1.3.1 1.3.2 1.3.3 1.4 1.4.1 1.4.2 1.4.3 1.4.4 1.4.5 1.5

Introduction 1 MMs from Droplet Microfluidics 4 Simple Spherical Microparticles (MPs) 4 Janus MPs 7 Core–Shell MPs 7 Porous MPs 9 Other MMs 10 NMs from Droplet Microfluidics 13 Inorganic NMs 13 Organic NMs 16 Other NMs 16 Applications of the Droplet-Derived Materials 18 Drug Delivery 18 Cell Microencapsulation 23 Tissue Engineering 25 Biosensors 29 Barcodes 32 Conclusion and Perspectives 35 References 36

2

Digital Microfluidics for Bioanalysis 47 Qingyu Ruan, Jingjing Guo, Yang Wang, Fenxiang Zou, Xiaoye Lin, Wei Wang, and Chaoyong Yang

2.1 2.2 2.2.1 2.2.1.1 2.2.1.2 2.2.1.3 2.2.2 2.2.3 2.3

Introduction 47 Theoretical Background 48 Theoretical Background 48 Thermodynamic Approach 49 Energy Minimization Approach 50 Electromechanical Approach 52 Contact Angle Saturation 53 Basic Microfluidic Functions by EWOD Actuation 53 Device Fabrication 55

vi

Contents

2.4 2.4.1 2.4.1.1 2.4.1.2 2.4.1.3 2.4.2 2.4.2.1 2.4.2.2 2.4.2.3 2.5 2.5.1 2.5.2 2.5.3 2.5.4 2.6

Digital Microfluidics Integrated with Other Devices 56 Sample Processing Systems Integrated with Digital Microfluidics 56 World-to-chip Interface 56 Magnet Separation 58 Heater Module 59 Detection Systems Integrated with Digital Microfluidics 59 Optical Methods 59 Electrochemical Methods 61 Other Detection Methods 62 Biological Applications on DMF 63 Enzyme Assays 63 Immunoassay 63 DNA-Based Applications 66 Cell-Based Applications 68 Conclusions and Perspectives 72 References 73

3

Nanotechnology and Microfluidics for Biosensing and Biophysical Property Assessment: Implications for Next-Generation in Vitro Diagnostics 83 Zida Li and Ho Cheung Shum

3.1 3.1.1 3.2 3.2.1 3.2.2 3.3 3.3.1 3.3.1.1 3.3.1.2 3.3.1.3 3.3.1.4 3.3.2 3.3.2.1 3.3.2.2 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.5

Introduction 83 Nanotechnology and Microfluidics 84 Fundamentals of Nanotechnology and Microfluidics 86 Nanotechnology 86 Microfluidics 87 Biomolecule Sensing 88 Techniques Based on Optical Readout 89 Localized Surface Plasmon Resonance 89 Surface-Enhanced Raman Spectroscopy 90 Nanoengineered Fluorescence Probes 91 Nanotopography-Based Cell Capturing 93 Techniques Based on Electrical Readouts 93 Electrochemical Reactions 93 Nanotransistor-Based Assays 94 Biophysical Property Sensing 95 Cell Contractility Measurement 96 Cell Deformability 98 Fluid Rheology 99 Electrophysiology 99 Concluding Remarks 100 Acknowledgments 100 References 101

4

Microfluidic Tools for the Synthesis of Bespoke Quantum Dots 109 Shangkun Li, Jeff C. Hsiao, Philip D. Howes, and Andrew J. deMello

4.1

Introduction 109

Contents

4.1.1 4.1.2 4.1.3 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.1.1 4.3.1.2 4.3.2 4.3.2.1 4.3.2.2 4.3.2.3 4.3.2.4 4.3.3 4.3.3.1 4.3.4 4.4 4.4.1 4.4.1.1 4.4.1.2 4.4.1.3 4.4.1.4 4.4.2 4.4.2.1 4.4.3 4.4.4 4.5

Microfluidics in the Chemical and Biological Sciences 109 Compound Semiconductor Nanoparticles 109 Microfluidic Tools for Nanoparticle Synthesis 112 Design Considerations 114 Continuous-Flow Microfluidics 115 Segmented-Flow Microfluidics 115 Continuous-Flow Microfluidic Synthesis of Quantum Dots 118 Homogenous Core-Type Quantum Dots in Continuous Flow 118 Cadmium Sulfide (CdS) 118 Cadmium Selenide (CdSe) 119 Heterogenous Core/Shell Quantum Dots in Continuous Flow 121 Zinc Selenide/Zinc Sulfide (ZnSe/ZnS) 121 Cadmium Selenide/Zinc Sulfide (CdSe/ZnS) and Cadmium Telluride/Zinc Sulfide (CdTe/ZnS) 121 Copper Indium Sulfide/Zinc Sulfide (CuInS2 /ZnS) 123 Indium Phosphide/Zinc Sulfide (InP/ZnS) 125 Heterogenous Core/Multishell Quantum Dots in Continuous Flow 125 Cadmium Selenide/Cadmium Sulfide/Zinc Sulfide (CdSe/CdS/ZnS) 126 Summary of QD Classes 128 Segmented-Flow Microfluidic Synthesis of Quantum Dots 128 Homogenous Structure Quantum Dots in Segmented Flow 129 Cadmium Sulfide (CdS) 129 Cadmium Selenide (CdSe) 130 Lead Sulfide (PbS) and Lead Selenide (PbSe) 131 Perovskite QDs 132 Heterogenous Core/Shell Quantum Dots in Segmented Flow 134 Copper Indium Sulfide/Zinc Sulfide (CuInS2 /ZnS) 134 Multistep Synthesis of QDs in Segmented Flow 135 Nucleation and Growth Studies of Quantum Dots 138 Conclusions and Outlook 140 References 141

5

Microfluidics for Immuno-oncology 149 Chao Ma, Jacob Harris, Renee-Tyler T. Morales, and Weiqiang Chen

5.1 5.2 5.2.1 5.2.1.1 5.2.1.2 5.2.1.3 5.2.1.4 5.2.2 5.2.2.1 5.2.2.2 5.2.2.3 5.3

Introduction 149 Microfluidics for Single Immune Cell Analysis 153 Single Immune Cells 153 T Cells 153 MΦs 156 DCs 157 B Cells 158 Microfluidics for Immune and Tumor Cell Interaction Analysis 159 T-cell Priming and Activation by APCs 159 Killing of Cancer Cells by Immune Effector Cells 162 Interaction Between Cancer Cells and MΦs 163 Microfluidics for Tumor Immune Microenvironment Analysis 163

vii

viii

Contents

5.3.1 5.3.1.1 5.3.1.2 5.3.1.3 5.3.1.4 5.3.2 5.3.2.1 5.3.2.2 5.4

Modeling the Tumor Immune Microenvironment 163 T-cell Trafficking and Migration 164 T-cell Priming and Activation by APCs 165 APC Processing and Presentation of TAAs 165 Interaction Between Cancer Cells and MΦs 166 On-chip Testing of Tumor Immunotherapy 166 TCR T Cells 167 Immune Checkpoint Blockade 167 Concluding Remarks and Future Perspectives 170 Acknowledgments 171 References 172

6

Paper and Paper Hybrid Microfluidic Devices for Point-of-care Detection of Infectious Diseases 177 Hamed Tavakoli, Wan Zhou, Lei Ma, Qunqun Guo, and XiuJun Li

6.1 6.2 6.2.1 6.2.1.1 6.2.1.2 6.2.1.3 6.2.1.4 6.2.2 6.3

Introduction 177 Fabrication of Paper-Based Microfluidic Devices 179 Fabrication Techniques for Paper-Based Microfluidic Platforms 179 Physical Blocking of Pores in Paper 180 Physical Deposition of Reagents on Paper Surface 181 Chemical Modification 182 Other Techniques 183 Fabrication of Paper Hybrid Microfluidic Devices 183 Application of Paper and Paper Hybrid Microfluidic Devices for Infectious Disease Diagnosis 184 Colorimetric Detection 185 Fluorescence Detection 187 Electrochemical Detection 191 Integration of Nanosensors on Paper and Paper Hybrid Microfluidic Devices for Infectious Disease Diagnosis 193 Carbon-Based Nanosensors 195 Gold-Based Nanosensors 198 Other Nanosensors 200 Summary and Outlook 202 Acknowledgment 202 References 203

6.3.1 6.3.2 6.3.3 6.4 6.4.1 6.4.2 6.4.3 6.5

7

Biological Diagnosis Based on Microfluidics and Nanotechnology 211 Navid Kashaninejad, Mohammad Yaghoobi, Mohammad Pourhassan-Moghaddam, Sajad R. Bazaz, Dayong Jin, and Majid E. Warkiani

7.1 7.2

Introduction 211 Quantum Dot-Based Microfluidic Biosensor for Biological Diagnosis 212 Qdot-Based Disease Diagnosis Using Microfluidics 213 Upconversion Nanoparticles 219

7.2.1 7.3

Contents

7.4 7.5 7.6 7.6.1 7.6.2 7.6.2.1 7.6.2.2 7.6.3 7.6.3.1 7.6.3.2 7.7

Fluorescent Biodots 221 Digital Microfluidic Systems for Diagnosis Detection 223 Paper-Based Diagnostics 226 Structure and Chemistry of Paper 226 Applications of Paper-Based Devices in the Diagnostics 227 Labeled Biosensing 228 Label-Free Biosensing 228 Integration of Nanoparticles with Paper-Based Microfluidic Devices 228 Gold Nanomaterials 228 Fluorescent Nanomaterials 229 Conclusion and Future Perspective 231 Conflicts of Interest 231 Acknowledgment 231 References 232

8

Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses 239 Dong Wang, Ho N. Chan, Zeyu Liu, Sean Micheal, Lijun Li, Dorsa B. Baniani, Ming J. A. Tan, Lu Huang, Jiantao Wang, and Hongkai Wu

8.1 8.2 8.2.1 8.2.2 8.2.3 8.2.4 8.3 8.3.1 8.3.2 8.4 8.4.1 8.4.2 8.4.3 8.5 8.6

Introduction 239 Cell 240 Blood Cell Counting 240 Characterization of CD64 Expression 241 Enumeration of CD4+ T Lymphocytes for HIV Monitoring 242 Circulating Tumor Cell (CTC) Isolation and Analysis 243 Nucleic Acid 245 Nonisothermal Amplification 245 Isothermal Amplification 246 Protein 253 Novel Chemistry and Nanomaterials 253 3D-Printed Microfluidic Devices 256 Digital and Droplet Microfluidics 259 Metabolites and Small Molecules 262 Conclusion and Outlook 271 Acknowledgments 271 References 271

9

Microfluidics in Microbiome and Cancer Research 281 Barath Udayasuryan, Daniel J. Slade, and Scott S. Verbridge

9.1 9.2 9.2.1 9.2.2 9.2.3 9.2.4 9.2.5

Introduction 281 What is the Microbiome? 282 Composition and Biogeography 282 The Microbiome and Cancer 285 Helicobacter pylori and Gastric Cancer 286 Fusobacterium nucleatum and CRC 287 Bacterial Invasion 288

ix

x

Contents

9.3 9.3.1 9.3.2 9.3.3 9.4 9.4.1 9.4.2 9.4.3 9.4.4 9.4.5 9.4.6 9.5

Studying the Microbiome 289 2D Models 291 3D Models 291 Organ-on-a-Chip and the Application of Microfluidics 295 Microfluidic Intestine Chip Models 297 Gut-on-a-Chip Model 297 Co-culture of the Gut-on-a-Chip with Microbiota 298 The HuMiX Model 299 Anaerobic Human Intestine Chip 301 Anoxic-Oxic Interface (AOI)-on-a-Chip 303 Future Directions 304 Concluding Remarks and Future Perspectives 306 Acknowledgments 308 References 308

10

Microfluidic Synthesis of Functional Nanoparticles 319 Ziwei Han and Xingyu Jiang

10.1 10.2 10.2.1 10.2.2 10.2.3 10.3 10.3.1 10.3.1.1 10.3.1.2 10.3.2 10.3.2.1 10.3.2.2 10.3.2.3 10.3.2.4 10.4

Introduction 319 Fabrication of Microfluidic Chips 320 Fabrication of Microchannels: Photolithography 321 Fabrication of PDMS-Based Microfluidic Chips 321 Pressure Tolerance 321 Microfluidic Synthesis of Functional Nanoparticles 323 Mixing Strategy 323 Hydrodynamic Focusing 323 Microstructure to Enhance Mixing Efficiency 324 Bionanoparticle Interactions 325 Well-Controlled Size and Monodispersity 325 Surface Modification 326 Mechanical Properties 327 Controllable Multilayer Structure 328 Microfluidic Assembly of Nanoparticles for Biological and Medical Applications 329 Drug Delivery 330 pH-Sensitive Drug Release 330 Hydrophilic Drug Delivery 331 Photoresponsive Drug Release 332 Gene Delivery 332 Imaging 332 MRI 332 Fluorescence Imaging 333 Ultrasonic Imaging 334 Biosensing 334 Theranostics 336 Prospects of Microfluidic Synthesis 337 Acknowledgment 338 References 339

10.4.1 10.4.1.1 10.4.1.2 10.4.1.3 10.4.1.4 10.4.2 10.4.2.1 10.4.2.2 10.4.2.3 10.4.3 10.4.4 10.5

Contents

11

Design Considerations for Muscle-Actuated Biohybrid Devices 347 Yoshitake Akiyama, Sung-Jin Park, and Shuichi Takayama

11.1 11.2

Introduction 347 Characteristics and Applicability of Muscles for Biohybrid Devices 348 11.2.1 Heart Muscle (Cardiomyocytes) 348 11.2.2 Skeletal Muscle Cells 350 11.2.3 Smooth Muscle Cells 351 11.2.4 Nonmammalian Muscle Cells 352 11.3 Arrangement of Muscle Cells and Tissues on Biohybrid Devices 352 11.3.1 Interfaces Between Muscle Cells and Material 353 11.3.1.1 Interfaces in 2D Culture 353 11.3.1.2 Interfaces in 3D Culture 354 11.3.2 Mechanical Pairing of Muscles 355 11.3.3 Interface Between Medium and Air 356 11.4 Oxygen Supply in Muscle Tissue Engineering 356 11.4.1 Equation and Conditions for Numerical Simulations 357 11.4.2 Oxygen Distribution under Static Culture 357 11.4.3 Oxygen Distribution in Microfluidic Devices 359 11.4.4 Other Approaches to Improve Oxygen Supply 360 11.5 Contractile Force of Muscle Bundles and Stimulations 361 11.5.1 Tissue-Engineered Muscle Consisting of C2C12 Cells 361 11.5.2 Tissue-Engineered Muscle Consisting of Primary Myoblasts 364 11.6 Control of Muscle Contractions 366 11.6.1 Electrical Stimulation 366 11.6.2 Optical Stimulation 367 11.6.3 Others 368 11.7 Conclusions and Future Challenges 368 11.7.1 Completely 3D-Printed Biohybrid Devices 368 11.7.2 Integration with Other Tissues 369 11.7.3 Long-Term Maintenance and Self-healing 369 11.7.4 Exploring Applications 370 Acknowledgments 370 References 370 12

Micro- and Nanoscale Biointerrogation and Modulation of Neural Tissue – From Fundamental to Clinical and Military Applications 383 Jordan Moore, Diego Alzate-Correa, Devleena Dasgupta, William Lawrence, Daniel Dodd, Craig Mathews, Ian Valerio, Cameron Rink, Natalia Higuita-Castro, and Daniel Gallego-Perez

12.1 12.2 12.2.1 12.2.2 12.3

Introduction 383 General Principles 385 Physics of Miniaturized Systems Material Properties 385 Areas of Study 386

385

xi

xii

Contents

12.3.1 12.3.2 12.3.3 12.3.4 12.4 12.4.1 12.4.2 12.4.3 12.4.4 12.4.5 12.4.6 12.5 12.6

Neurodevelopment 386 Neuro-oncology 388 Neurodegenerative Disorders 389 Traumatic Brain Injury 392 Applications 394 Neuron-Directed Cellular Reprogramming 394 Tissue Nanotransfection 396 Cancer Interrogation 398 FISH On-Chip for Alzheimer’s Disease 401 On-chip Brain Injury 403 Military 405 Limitations and Future Outlook 406 Summary 407 References 408 Index 419

xiii

Preface The rapid development of microfluidics technology has promoted new innovations in nanoscience and nanotechnology based on precise manipulation of fluids within microscale confinements. On the other hand, nanotechnology also plays a key role in developing microfluidics. The far-reaching progress in microfluidics technology often follows from the development of new nanomaterials and nanotechnology that can provide robust and reliable performance and functionalities. The combination of nanotechnology on microfluidic devices offers great potential for various applications, such as biomedical diagnostics, biochemical analysis, drug delivery, and tissue engineering. In this book, we summarize the recent progress of microfluidics technology and examine the impact of microfluidics on nanotechnology over the past decade. Compared to materials assisted by conventional strategies, advanced materials synthesized by droplet microfluidics are with superior properties and performances. Chapter 1 by Professor Yuanjin Zhao focuses on the recent progress in the design principles of droplet microfluidics for synthesizing micro/nanostructured materials. The practical applications and remaining technical challenges in droplet microfluidics are also discussed. Digital microfluidics, a new concept in the field of microfluidics, is based on an array of electrodes with hydrophobic surface for precise manipulation of discrete droplets in picoliter to microliter sizes. Chapter 2 by Professor Chaoyong Yang provides a comprehensive review on the status quo of digital microfluidics from the aspects of theoretical background, chip fabrication, device integration, and biological applications. Nanotechnology enables biosensing techniques with enhanced performance, while microfluidics offers automated sample preparation and exquisite fluid handling. Chapter 3 by Professor Ho Cheung Shum provides an overview of the emerging techniques for in vivo diagnostics and in vitro diagnostics (IVD) based on nanotechnology and microfluidics, covering the underlying mechanisms and discussing representative works. Chapter 4 by Professor Andrew J. deMello focuses on the microfluidics-based synthesis of bespoke quantum dots (QDs), covering continuous- and segmented-flow platforms, strategies for homogenous and heterogenous QD synthesis and integration of in situ monitoring techniques. They provide basic knowledge on microfluidics and semiconductor nanoparticles at the very beginning of the chapter, which makes it easy for nonprofessionals to understand the following contents. Microfluidics has already been shown to be an invaluable and robust

xiv

Preface

tool for cancer immunotherapy, from single-cell analysis to tumor immunity modeling. Chapter 5 by Professor Weiqiang Chen summarizes the current state of the art of microfluidics technologies that can be used to analyze the key cell players during immunotherapy progression and resistance from the levels of isolated single cells and paired cell interactions. Chapter 6 by Professor XiuJun Li focuses on recent advances of point-of-care detection of infectious diseases using paper-based microfluidic platforms, in particular fully paper-based and paper hybrid microfluidic platforms, especially in resource-poor settings. Several examples that integrate nanosensors on paper and paper hybrid microfluidic devices have been presented. Chapter 7 by Professor Majid E. Warkiani describes the applications of nanoparticle-based microfluidics for disease diagnosis. Applications of digital and paper-based microfluidic platforms for biological diagnosis have been discussed. Chapter 8 by Professor Hongkai WU reviews the recent development and improvement of microfluidic platforms for point-of-care diagnosis tests based on the detection of different types of biomarkers. Representative microfluidic devices are demonstrated for the detection of cells, nucleic acids, proteins, metabolites, and small molecules for point-of-care disease diagnosis. Chapter 9 by Professor Scott S. Verbridge focuses on the topic of microfluidic in microbiome and cancer research. This chapter is divided into sections on concept of microbiome, the models of microbiome study, microfluidic intestine chip models, as well as future directions in this field. Chapter 10 by Professor Xingyu Jiang focuses on the microfluidic assembly of nanoparticles and their specific applications in medical and biological fields. A prospect is also provided to demonstrate other potential applications and remaining challenges of microfluidic assembly of nanoparticles. Chapter 11 by Professor Shuichi Takayama introduces the muscle-actuated biohybrid devices and summarizes the various factors to design a biohybrid device. Chapter 12 by Professor Daniel Gallego-Perez focuses on the application of micro- and nanotechnology on the biointerrogation and modulation of neural tissues. They summarize the use of micro- and nanotechnology on the context of neural development, neuro-oncology, neurodegenerative diseases, and traumatic brain injury. 28 May 2019 Beijing, China

Xingyu Jiang

1

1 Micro/Nanostructured Materials from Droplet Microfluidics Xin Zhao 1 , Jieshou Li 1 , and Yuanjin Zhao 1,2 1 Medical School of Nanjing University, Research Institute of General Surgery, Jinling Hospital, No. 305, East Zhongshan Road, Xuanwu District, Nanjing 210002, P. R. China 2 Southeast University, State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, No. 2, Sipailou, Xuanwu District, Nanjing 210096, P. R. China

1.1 Introduction Since the emergence of microfluidics at the beginning of 1980s, microfluidic technologies have been extensively applied in the fabrication of materials with specific physicochemical features and versatile applications [1–3]. This relatively new field is the synergy of science and technology of systems with integrated channels on the microscale dimensions, through which small quantities of fluids (usually 10−9 to 10−18 l) can flow in designed configurations and are precisely controlled and manipulated [4–6]. In the field of microfluidics, as fluid dimensions shrink to the microscale level, their specific surface area increases, thus showing behaviors divergent from those of macroscopic fluids, which can be characterized by three major phenomena: highly efficient mass–heat transfer, relative dominance of viscous force over inertial force, and significant surface effects [7, 8]. In addition, the high integration of microfluidic channels facilitates the coexistence and diverse interactions of multiple fluid phases and paves the way for miniaturized systematic control over individual fluids and fluid interfaces [9, 10]. These features offer obvious advantages over bulk synthesis, most notably in their ability to ensure monodispersity and control the structure of final products [11–13]. Therefore, microfluidics has promoted the development of multidisciplinary research in physical, chemical, biological, medical, and engineering fields. Droplet microfluidics is an important subcategory of the microfluidic technologies, which generates and manipulates discrete droplets through immiscible multiphase flows inside the microchannels [14–16]. In the past two decades, fostered by great progress in both theoretical and technical aspects, droplet microfluidics has fulfilled original expectations and become a significant approach to generate materials for a broad range of applications [17–19]. The basic principles and microfluidic devices for droplet generation are shown

Nanotechnology and Microfluidics, First Edition. Edited by Xingyu Jiang. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.

2

1 Micro/Nanostructured Materials from Droplet Microfluidics

in Figure 1.1, including a T-junction chip (Figure 1.1a), a flow-focusing chip (Figure 1.1b), and a coaxial structured chip (Figure 1.1c) [20]. In the T-junction chip, the dispersed phase flows from a vertical channel to a horizontal channel filled with the continuous phase. Under the combined action of both shear force and extrusion pressure from the continuous phase, monodispersed droplets are generated. In the flow-focusing chip, the dispersed phase flows from the middle channel and undergoes extrusion force of the continuous phase from all directions. The dispersed phase experiences stretching and breakage, leading to droplet formation. In the coaxial structured chip, the dispersed phase channel is embedded in the continuous phase channel, and the dispersed phase flows parallel to the continuous phase toward the same direction. Also, the dispersed phase is broken into droplets. In microfluidic systems, droplet generation is influenced by microchannel construction, viscosity and flow velocity of each phase, and interfacial tension between adjacent flows. Therefore, the dimensions and production rates of droplets can be regulated by adjusting the above parameters. In addition, through a flexible design of microchannels, double or even multiple emulsions could be generated in a controlled manner (Figure 1.1d,e) [21, 22]. These microfluidic droplets have diverse morphologies and components and can serve as excellent templates to synthesize materials with specific structures and functions. With the development of microfabrication technology, considerable research has been made to synthesize microstructured materials (MMs)/nanostructured materials (NMs) because the microscopic architectures give additional properties to the materials [23–26]. Conventional bulk methods usually adopt a certain physical or chemical procedure (e.g. mechanical stirring) [27, 28]. These methods usually generate materials with a monotonous morphology, and the dispersity of products and synthetic processes are difficult to control [29, 30]. In particular, for fabrication of composite materials, such as “intelligent materials” or “core–shell materials,” the conventional approaches are insufficient to meet the requirements. MMs/NMs synthesized from droplet microfluidics possess narrow size distribution, flexible structures, and desired properties [31–33]. Compared to conventional methods, the advantages of microfluidic synthesis lie in the following aspects [20, 34–36]. The material size, structure, and composition are easily controlled, resulting in superior properties and functions. The addition of reagents is very simple, which is beneficial for the manipulation of multistep and multireagent synthesis. Through scale integration of microfluidic systems and equipment automation, the complex reaction process can be largely simplified. Because majority of materials used to make microfluidic chip are facilitated to be observed, real-time monitoring of the reaction process could be realized, which helps to clarify the synthesis mechanism. Therefore, the application of droplet microfluidics to design and prepare MMs/NMs has become a hot topic recently and will bring about infinite possibilities for the future development of materials science. In this chapter, we summarize the classical and recent achievements in the MMs/NMs engineered from droplet microfluidics and their various applications. We first provide an overview of MMs fabricated by droplet microfluidics, including the droplet formation mechanism and various microchips used to

1.1 Introduction

(a)

(b)

Continuous phase

Dispersed phase

Continuous phase

(c)

Dispersed phase

Continuous phase

Dispersed phase

(d) (i) (ii) (iii) (iv) (v)

(e)

D2 Inner fluid (Q1)

Middle fluid (l) (Q2)

Middle fluid (lI) (Q3) Outer fluid (Q4)

Transition tube (I)

D4

Injection tube

D3

Inner fluid (Q1)

Collection tube

D2

Transition tube

D1

Injection tube

d2

D3

D1

d1

Outer fluid (Q3)

Middle fluid (Q2)

Transition tube (II)

Collection tube

Figure 1.1 (a–c) The principles and chip designs with different flow regimes for droplet generation, including T-junction (a), flow-focusing (b), and coaxial (c) structured chip. Source: Ma et al. 2017 [20]. https://www.mdpi.com/2072-666X/8/8/255. Licensed under CCBY 4.0 (d) Generation of multiple emulsions in a stepwise flow-focusing device: (i–v) single-, double-, triple-, quadruple-, and quintuple-emulsion droplets, respectively. Source: Adapted with permission from Abate and Weitz [21]. Copyright 2009, John Wiley & Sons. (e) Generation of multiple emulsions in a stepwise coflow platform. Source: Adapted with permission from Chu et al. [22]. Copyright 2007, John Wiley & Sons.

3

4

1 Micro/Nanostructured Materials from Droplet Microfluidics

generate different droplets, the methods to prepare MMs templated from these droplets, and the unique and complex structures enabled by microfluidic techniques. We then present basic synthesis methods for inorganic and organic NMs through droplet microfluidics, and the heterogeneous and multifunctional nanostructures from microfluidic platforms are also introduced. Following these two sections, much emphasis will be laid on the applications of the generated MMs/NMs, including drug delivery, cell encapsulation, TE, and analytical applications. Finally, we will discuss the current status and existing challenges and provide opinions on the directions of future development of droplet microfluidics in the synthesis of advanced MMs/NMs.

1.2 MMs from Droplet Microfluidics Although the history of MMs with sizes ranging from 1 to 1000 μm has started in the 1960s, their application was only expanded recently after they were utilized as drug delivery agents by mimicking genetic materials carrying pollens [37– 39]. Thereafter, other studies have continuously investigated the functionalities of MMs and they are now being utilized in various fields including pharmaceuticals, food industry, cosmetics, photonics, coatings, and printing [40, 41]. These applications of MMs depend on their properties that correlate with their size, structure, composition, and configuration [42, 43]. Typically, MMs have been prepared through traditional methods including emulsion polymerization, dispersion polymerization, and spray drying [44]. These methods always result in MMs with large polydispersity, poor reproducibility, limited functionality, and less tunable morphology [44–46]. Therefore, it is becoming increasingly urgent to fabricate MMs with defined sizes, morphologies, and compartments in a controlled manner to improve their capability. Droplet microfluidics can generate emulsion droplets with a precisely controlled size, shape, and composition, which provide excellent templates for the synthesis of functional MMs with uniform size, controllable shape, and versatile compositions [47–49]. Moreover, precise control over single emulsion droplets by microfluidics allows further creation of multiple emulsions with highly controllable, nested, and droplet-in-droplet structures [50, 51]. Thus, using such multiple emulsions as templates, MMs with well-tailored internal compartments and specific functions can be successfully fabricated for many applications. 1.2.1

Simple Spherical Microparticles (MPs)

Simple spherical microparticles (MPs) are synthesized straightforwardly by solidification of the droplet templates, which involves a chemical or physical reaction process [52–54]. Photopolymerization is one of the most prevalent chemical processes because it enables in situ solidification and continuous fabrication in a fast response time, which helps to determine the particle location and better control the size distribution [55, 56]. For example, Jeong et al. developed a simple and cost-effective method for the fabrication of polymeric

1.2 MMs from Droplet Microfluidics

MPs in droplet microfluidics [57]. The polymerizable sample fluid and the immiscible nonpolymerizable sheath fluid (mineral oil) were introduced into the inlet channels of the sample and sheath flow, respectively. Both fluids were combined at the tip of the pulled micropipette (dotted area in Figure 1.2a), producing hydrogel droplets floating among the sheath stream. The separated droplets traveled through the main channel without touching the inner wall of the channel and then the hydrogel droplets were polymerized by continuous ultraviolet (UV) exposure at the unshielded area. The size of the MPs could be adjusted within the range of tens to hundreds of micrometers by changing the flow rates of the dispersed and continuous phases. This method has been extensively applied for the synthesis of a large variety of MPs by using precursors with unsaturated hydrocarbon chains. MPs synthesized via UV irradiation are inappropriate when considering their biotoxicity and biocompatibility. In those cases, physical gelation or ionic reactions are more applicable and has been employed to synthesize many kinds of MPs [62, 63]. For example, Tan and Takeuchi described the production of monodisperse alginate hydrogel MPs using a method that combined the internal gelation method with T-junction droplet formation in microfluidic devices (Figure 1.2b) [58]. They dispersed droplets of Na-alginate solution containing CaCO3 nanoparticles (NPs) in a continuous phase of corn oil at room temperature. Syringe pumps were used to infuse fluids into the microfluidic device fabricated from polydimethylsiloxane (PDMS) using soft lithography techniques. Corn oil with lecithin sheared off droplets of Na-alginate solution containing insoluble CaCO3 NPs one at a time to generate an inverse emulsion with a narrow size distribution. Lecithin was added to the corn oil to stabilize the droplets against coalescence, thereby preserving the monodispersity of the droplets. In downstream, acetic acid dissolved in corn oil was introduced and mixed with oil flowing in the main stream. The acetic acid then diffused into the aqueous Na-alginate droplets, reduced pH, and released Ca2+ ions from the insoluble calcium complex, causing gelation. Also, they demonstrated that the gelation conditions in this approach were mild enough to encapsulate cells without loss of their viability. The low production rate of microfluidic devices for generation of MPs has remained a key challenge to successfully translate many promising laboratory-scale results to commercial-scale production of microfluidicsgenerated materials [64, 65]. To address these challenges, Yadavali et al. incorporated an array of 10 260 (285 × 36) microfluidic droplet generators onto a 3D-etched single silicon wafer that was operated using only a single set of inlets and outlets (Figure 1.2c) [59]. The monolithic construction from a single silicon wafer obviated the alignment and bonding challenges of prior multilayer approaches and allowed high-pressure use. To demonstrate the power of this approach, they generated polycaprolactone solid MPs, with a coefficient of variation (CV) ds and h < ds, that is, the droplet flowed into a wide channel, it changed to a disk shape, and when both of the geometries were smaller than ds, the droplet was confined into a rod-like shape. By illumination with UV light, these droplets were polymerized into solid particles and their shapes were retained. Another facile approach was introduced by Wang et al.; they encapsulated monodisperse droplets generated from microfluidics into cross-linked polymeric networks via interfacial cross-linking reaction in microchannel to produce droplet-containing fiber-like matrices [93]. By stretching and twining the dried fiber-like matrices, the encapsulated droplets could be engineered into versatile shapes from tablet to helix. Based on these deformed droplet templates, versatile MMs were synthesized, such as tablet-like, rod-like, needle-like MMs and complex 3D helices for magnetic-driven rotational and translational motion (Figure 1.4B). The material shape can also be tailored by controlling the reaction parameters during the solidification process. For example, Lin et al. synthesized tail-shaped alginate MPs in a slow cross-linking process. As shown in Figure 1.4C, sodium alginate droplets were first generated in a microfluidic device and then fell from the oil carrier phase into a CaCl2 solution under gravity [94]. During the slow sedimentation process, the droplets were deformed under the competitive effects of viscous deformation and the interfacial restoring forces, which resulted in the generation of alginate MPs with teardrop or tail shapes. By changing the alginate viscosity and the Ca2+ concentration, the size and the morphology of the particles could be tuned. In another interesting study, acorn-like or sharp-edged MMs were fabricated by tuning the wettability properties between two immiscible drops. When the drop pairs were emulsified together in a third carrier fluid phase and came into contact, an equilibrium structure formed in accordance with the spreading coefficient values and droplet volumes. By carefully controlling the interfacial tensions between the two fluid phases using surfactants, the spreading coefficients could be adjusted so that partial engulfment between the two dispersed phases occurred in compliance with the minimum total interfacial energy. Therefore, the drop pairs could finally form a dumbbell- or acorn-shaped configuration. Additionally, Nisisako and Torii synthesized particles with sharp edges by selective polymerization of one of the droplet pairs. They constructed a triphase microfluidic system composed of photocurable monomer 1,6-hexanediol diacrylate (HDDA), a silicone oil phase, and an aqueous phase [95]. Biphasic droplets were generated downstream

11

Microstructured materials

(B) Droplet-containing fiber-like matrix

(A)

Drying

Template synthesis

D L

Stretching Stretching

Twining

(C)

Continuous phase: oil

Dispersed phase: 1% Na-alginate in water

Flow direction

(D)

(a)

(b)

(c)

(e)

(f)

Na-alginate emulsion Continuous phase: oil

(d)

Oil

Ca-alginate particle CaCl2

Figure 1.4 (A) Representations of the shapes of droplets in the microfluidic channel, and optical microscopy images of microfluidic MMs based on these deformed droplets. Source: Adapted with permission from Xu et al. [92]. Copyright 2005, John Wiley & Sons. (B) Manipulation of the encapsulated droplets by drying, stretching, stretching, and twining for flexible deformation and synthesis of nonspherical MMs and helices from the deformed droplets. Source: Adapted with permission from Wang et al. [93]. Copyright 2017, John Wiley & Sons. (C) Formation of a tail-shaped Ca-alginate particle from a spherical Na-alginate droplet. Source: Adapted with permission from Lin et al. [94]. Copyright 2013, John Wiley & Sons. (D) MMs of various shapes engineered from Janus droplets containing different volume ratios of silicone oil and the monomer. The scale bars represent 100 μm. Source: Adapted with permission from Nisisako and Torii [95]. Copyright 2007, John Wiley & Sons.

1.3 NMs from Droplet Microfluidics

at equilibrium, and MPs were synthesized by photopolymerization of the monomer. It was demonstrated that, with increasing fraction of the HDDA monomer, the particle shape varied from convex to planar and further became concave (Figure 1.4D).

1.3 NMs from Droplet Microfluidics In recent decades, NMs have drawn significant attention in various applications [96, 97]. Because of electric confinement and surface asymmetry effects, NMs show distinct properties such as optical emission in semiconductor NMs and surface plasmon resonance in noble metal NMs. Additionally, as their dimension is similar to that of biomolecules, they can be tailored to coordinate with biological systems and showed unique properties in imaging, optoelectronics, catalysis, sensing, and drug delivery [98, 99]. The synthesis of NMs goes through four steps: supersaturation, nucleation, growth, and aggregation, and there have emerged many techniques for NM synthesis, which can be classified into two main categories: “top-down” and “bottom-up.” [100] Although the properties of NMs are highly related to their size and morphology, it is technically challenging in conventional batch processes to reproducibly fabricate NMs having a desired morphology with a small standard deviation [101]. Since 1990s, much research effort has been devoted to the synthesis of NMs by droplet microfluidics. In contrast to conventional batch systems, the microfluidic method stands out for its intrinsic advantages, including miniaturization, enhanced mass and heat transfer, and reduced time and reagent consumption. Especially, the droplet reactor offers additional fascinating strengths. For example, as the reaction is confined in the microscale droplets, toxic or volatile chemicals can be utilized, and the resultant NMs would not contact the channel walls, thus avoiding possible contamination and blocking. In addition, the advection flow field within the droplets further improves the mixing efficiency, thus offering a well-defined starting point and a consistent residence time, which contributes to a narrower size distribution of the final NMs. Moreover, local control over the synthetic environments could be exerted on separate droplet reactors. Therefore, the reaction parameters scale up linearly, enabling homogeneous synthesis and quantity production [102–105]. Here, we reviewed recent developments in synthesizing inorganic, organic, and other composite NMs through droplet microfluidics. 1.3.1

Inorganic NMs

Inorganic NMs can be classified as amorphous and crystalline ones. Their chemical and physical properties are related to the size, shape, and structure of the particles. Thus, the monodispersity of inorganic NMs should be considered during the formation process, which is largely dependent on the specially controlled reaction kinetics, rapid mixing of the injected precursors, and well-defined time–temperature profiles [106, 107]. All of these parameters

13

14

1 Micro/Nanostructured Materials from Droplet Microfluidics

can be precisely regulated in microfluidics based on microscale characteristics [108, 109]. Controllable synthesis of colloidal semiconductor, metal, metal oxide, and hybrid NMs has been demonstrated. For example, Lazarus et al. used a two-phase microfluidic droplet device to synthesize AuNPs and silver nanoparticles (AgNPs) [110]. As shown in Figure 1.5a, a carrier oil phase was injected through inlet 1. The two reaction reagents of a metal salt precursor and the reductant flowed via inlets 2 and 4, respectively. Also, an ionic liquid stream flowed through inlet 3 and served as a stabilizer and also prevented contact of the reaction reagents before droplet formation. When droplets were generated in the T-junction and kept in the dripping regime, the recirculating streamline induced a convective flow, which largely accelerated mixing of the two reaction reagents. This promoted a rapid nucleation burst and thus ensured a homogeneous synthesis environment. As demonstrated by the authors in Figure 1.5b,c, well-dispersed spherical NPs were obtained that were smaller and more monodisperse than those produced in analogous batch reactions as a result of the rapid mixing and the homogeneous reaction environment afforded by the discrete droplets within an immiscible carrier phase. Besides the demands for homogeneity in NMs, controllable nanostructures are another significant consideration. Based on the flexible design of microfluidic chips and flows, various complex nanostructures have been achieved, such as core–shell NPs, nanogels with controlled pore size, Janus NPs, and other complex materials [100–102]. Shestopalov et al. reported a plug-based, microfluidic method for performing multistep chemical reactions to synthesize CdS/CdSe core–shell NPs with millisecond time control [111]. In this microfluidic method, the investigators first generated droplets from the initial reaction mixture. Two aqueous reagent streams were brought together into the channel where they were allowed to flow laminar alongside each other (labeled R1 and R2 in Figure 1.5d). These reagent streams were then sheared into droplets by an inert stream (labeled S in Figure 1.5d). The winding channels induced mixing by chaotic advection, and the droplets were allowed to react rapidly. To initiate the second stage of the multistep reaction, an aqueous stream of an additional reagent (labeled R3 in Figure 1.5d) was directly injected into the droplets at a junction. The second reaction proceeded as droplets flowed through another length of serpentine channel (Figure 1.5e). Therefore, by conducting multistep reactions in droplets, the CdS/CdSe core–shell NPs were produced and prevented from aggregating on the walls of the microchannels. Because of the small dimensions of droplets and the closed environment of a pressure-driven flow system, it is typically difficult to control the delivery of chemicals across the continuous phase into the microdroplets. As a result, most applications involve only an initial mixing of all the chemicals prior to the generation of microdroplets. However, many chemical reaction systems require addition of chemicals into generated microdroplets at precise time intervals [113, 114]. An effective method is to directly inject chemicals into the droplets through a side channel, where the chemical solution forms a pendant drop at the junction and merges with the microdroplets passing by. Based on previous studies, Gu et al. presented a novel method that could control material transport into the microfluidic droplets [112]. Instead of using an inert oil as the carrier fluid,

1.3 NMs from Droplet Microfluidics

(a) 2

S

(d)

3

R1

R2

4 1000 μm 200 μm

Oil Out

1

R3

(b)

(e)

100 μm

wf = 0.2, U = 67 mm s–1

(f)

Nanodroplet

Aqueous DP

Convection Electrocoalescence

(c) W/O miniemulsion CP

HAuCI4/PVP solution Ascorbic acid emulsion Na2PdCI4 emulsion Ascorbic acid emulsion

Au-Pd core–shell nanoparticles

Figure 1.5 (a–c) Schematic representation of the multiple inlet T-junction microfluidic device used to synthesize AuNPs and AgNPs (a); transmission electron microscopy micrographs of AuNPs (b) and AgNPs (c) produced in a droplet-based microfluidic device. Scale bars were 50 nm. Source: Adapted with permission from Lazarus et al. [110]. Copyright 2012, American Chemical Society. (d, e) A schematic diagram of the microfluidic network (d) and a micrograph showing droplets merging with the aqueous stream (e). Source: Adapted with permission from Shestopalov et al. [111]. Copyright 2004, Royal Society of Chemistry. (f ) Experimental setup for generating aqueous microdroplets and Au-Pd core–shell NPs with a W/O miniemulsion and electrocoalescence. Source: Gu et al. 2018 [112]. https://pubs.rsc.org/en/ content/articlelanding/2018/lc/c8lc00114f#!divAbstract. Licensed under CCBY 3.0.

a W/O miniemulsion was used by the investigators as the continuous phase to generate aqueous monodisperse microdroplets as dispersed phase (Figure 1.5f ). The W/O miniemulsion was a thermodynamically metastable system, composed of 50–500 nm aqueous nanodroplets dispersed in an immiscible organic solvent with a stabilizing surfactant. Via electrocoalescence, these nanodroplets served as carriers for chemicals and were transported into the microdroplets. As the nanodroplets were 3 orders of magnitude smaller than the microdroplets, the investigators could easily achieve a nanodroplet-to-microdroplet population ratio of greater than 1 million. Such a large population ratio made it possible to control the chemical addition rate over a wide range, and the addition was in fact “quasi-continuous.” Finally, this method was successfully applied to a single-step

15

16

1 Micro/Nanostructured Materials from Droplet Microfluidics

synthesis of AuNPs and a multistep flow synthesis of Au-Pd core–shell NPs with a narrow size distribution. 1.3.2

Organic NMs

Amphiphilic molecules such as block copolymers and lipids can self-assemble into NPs when they experience a change in solvent quality. A common and flexible way to accomplish such a change is by mixing the solvent with the antisolvent, where the mixing time directly influences the final size and size distribution of the resultant NPs [115, 116]. However, the heterogeneous environment prevents stabilization of the nascent NPs, facilitates their aggregation, and leads to formation of larger and polydisperse products. In microfluidic systems, the highly efficient mixing of the flow has resulted in polymeric and lipid NPs with tunable size, narrow distribution, and batch-to-batch reproducibility [117, 118]. For example, Hung et al. presented droplet microfluidics-based solvent evaporation and extraction process to enable the controlled generation of monodisperse PLGA particles (Figure 1.6a,b) [119]. A mixture of PLGA-DMSO and water droplets was formed in a carrier phase of silicon oil. After droplet coalescence, DMSO was extracted out into water, and the PLGA nanospheres precipitated as a result of supersaturation. By tuning the PLGA concentration in solvent and the relative flow rates of oil and aqueous phases in the system, they were able to synthesize particles ranging from 70 nm to 30 μm in diameter. Although the size-controlled synthesis of polymeric NPs was achieved by the aforementioned one-stage microfluidic chips, it is still difficult to synthesize the hybrid core–shell NPs with tunable sizes because of their complex structures. Recently, via a specifically designed two-stage microfluidic chip, Zhang et al. realized the synthesis of controllable core–shell NPs with polymer cores and lipid shells. The first stage of the chip consisted of three inlets and one straight synthesis channel, whereas the second stage had one middle inlet and a spiral synthesis channel (Figure 1.6c) [104]. In mode A, they introduced PLGA solution into the first stage of chip to precipitate intermediate PLGA NPs and injected lipid solution into the second stage to assemble lipid monolayer shell onto the surface of PLGA NPs by hydrophobic attraction between lipid tail and PLGA. In mode B, they generated an intermediate liposome by injecting lipid solution into the first stage, which could reassemble onto the PLGA NPs when PLGA solution was injected into the second stage. In mode A, the NPs were covered by lipid–monolayer–shell, whereas in mode B, NPs were coated by lipid–bilayer–shell. The results indicated an enhanced mixing effect at the high flow rate in microfluidic chips, thus resulting in the assembly of small and monodisperse hybrid NPs. 1.3.3

Other NMs

Metal–organic frameworks (MOFs) are porous crystalline materials consisting of metal clusters or ions that act as connecting nodes and rigid organic bridging ligands. They have attracted immense attention because of their potential for extremely diverse structural topologies and tunable chemical

1.3 NMs from Droplet Microfluidics

(a)

1% PVA Continuous phase

(e)

g

rgin

me

Ni3BTC2 (shell)

Co3BTC2 crystals in droplets

PLGA/DMC Discrete phase

Co3BTC2 (core)

PLGA particles

So eva lvent por atio n

Mineral oil

BPR

140 °C

Core MOF precursor

140 °C, 4.5 MPa

Co3BTC2@Ni3BTC2

Shell MOF precursor

(f)

(b)

PS-CdS quantum dot building block

PS-b-PAA block copolymer stabilizing chains

PACd layer

PS block

CdS quantum dot

PAA block PS corona

Water addition Quantum dot compound micelle

1 μm

10 nm

(c)

PAA

30 μm Second stage

Stage Inlet 1

Outlet

Inlet 2

5 mm MP Mode A: PLGA

Lipid

DMF/water inlet Argon inlet

PS-CdS

Mixing channel Flow direction

DMF separator inlet

BP Mode B: Lipid

PLGA

Injector DMF/solids inlet

(d)

Outlet

Solvents Organic Metallic ligand salt Reaction time

1 cm

Post formation channel

Oil

Droplet generation

Solvothermal synthesis

Figure 1.6 (a, b) Schematic of device design for the solvent evaporation method (a) and size graph showing that a wide range of particle sizes were achieved with the extraction and evaporation methods (b). Source: Adapted with permission from Hung et al. [119]. Copyright 2010, Royal Society of Chemistry. (c) Schematic of the two-stage microfluidic chip and generated monolayer-covered or bilayer-covered PLGA NPs. Source: Adapted with permission from Zhang et al. [104]. Copyright 2015, American Chemical Society. (d, e) Schematic representation of the general microchemical process (d) and the integrated hydrothermal microchemical process for synthesis of core–shell MOFs (e). Source: Adapted with permission from Faustini et al. [120]. Copyright 2013, American Chemical Society. (f ) Schematic of the QDCM assembly process and the microfluidic reactor. Source: Adapted with permission from Wang et al. [121]. Copyright 2010, American Chemical Society.

functionalities [122, 123]. Microfluidics has recently been employed for the synthesis of MOFs. For example, Faustini et al. reported an ultrafast and continuous synthesis of versatile MOFs with unique morphologies based on microfluidic strategy. Compared to conventional batch processes, the reaction kinetics of MOFs preparation were tremendously improved in confined droplets

17

18

1 Micro/Nanostructured Materials from Droplet Microfluidics

(Figure 1.6d) [120]. Representative MOF structures, such as HKUST-1, MOF-5, IRMOF-3, and UiO-66, were synthesized within a few minutes. In addition, three different types of core–shell MOFs composites, i.e. Co3 BTC2 @Ni3 BTC2 , MOF-5@diCH3 -MOF-5, and Fe3 O4 @ZIF-8, were synthesized by exploiting a unique two-step integrated microfluidic system (Figure 1.6e). Unique features such as anisotropic crystal growth or enhanced stability against moisture were observed in these MOFs. Therefore, the microfluidic strategy allowed continuous fabrication of high-quality MOFs crystals and composites exhibiting distinct morphological characteristics in a time-efficient manner. The droplet platform also provides a microchamber for NM assembly into high-order structures, which exhibit distinct features and potential application values. Wang et al. used a two-phase gas–liquid-segmented microfluidic reactor to control the self-assembly of polystyrene-coated quantum dots (PS-CdS) and stabilizing PS-b-PAA block copolymer into quantum dot compound micelles (QDCMs) (Figure 1.6f ) [121]. Mixing of water with polymeric constituents was greatly enhanced because of chaotic advection in the liquid plugs as they traveled through the sinusoidal channel. In addition, circulating flow within the liquid plugs provided two convenient handles for size control following initial self-assembly: shear-induced particle breakup and collision-induced particle coalescence. The resulting particle size was a balance of the relative rates of these fluids under a specific set of experimental parameters. By changing the self-assembly conditions, the mean particle size was tuned in the range of 40–137 nm. These results demonstrate that microfluidic strategies have obvious advantages in controlling self-assembly of block copolymers and other colloids.

1.4 Applications of the Droplet-Derived Materials Droplet microfluidics enables synthesis of materials with uniform and highly controlled sizes and structures. By incorporating specific ingredients, such materials could be endowed with distinct physical and chemical properties such as optical features, mechanical strength, selective permeability, and stimulus-responsive capacity. Therefore, they are highly applied in various fields. Herein, we categorize their applications into several aspects. 1.4.1

Drug Delivery

Encapsulation and controlled release of active agents are of significant interest in developing advanced delivery system for drugs, nutrients, fragrances, and cosmetics. Especially, for efficient drug delivery, the agents should be encapsulated within carriers with desired doses and then released in a specific target location [124, 125]. However, the clinical translation of drug delivery systems is relatively slow, which can be partially ascribed to the poor control of the preparation processes in the conventional batch method [126]. For example, the polymeric particles prepared by the conventional batch method usually show high batch-to-batch variations in physicochemical properties, such as the average

1.4 Applications of the Droplet-Derived Materials

particle size, size distribution, surface charge, and drug release profiles. Materials synthesized by droplet microfluidics possess several advantages to overcome the dilemma, which lie in the following aspects. First, droplet-synthesized materials are highly tunable and uniform in size, structure, and encapsulation efficiency. This provides a guarantee for maintaining a consistent release response and thus for regulating the release rate. Second, the droplet-based drug delivery system allows for a wide range of material choices, and multiple drugs could be loaded simultaneously for investigating their synergetic effects. Third, by using different matrix materials, various release profiles can be achieved under an external stimulus, which are important for specific usages [127, 128]. Many MP-based drug delivery platforms are developed through O/W single emulsion, where the drugs are encapsulated into a biocompatible polymeric particle matrix, including PLGA, poly-𝜀-caprolactone, and hydroxypropyl methylcellulose acetate succinate (HPMCAS). The polymers used in O/W emulsions are typically dissolved in a volatile solvent that can evaporate or diffuse out from the droplets. The size of MPs is primarily controlled by the droplet size and how much the final droplet shrinks during solvent removal, an adjustable variable by employing different polymer concentrations [124]. For example, Xu et al. fabricated monodisperse biodegradable drug-loaded PLGA MPs by combining the formation of droplets in a microfluidic flow-focusing generator with rapid evaporation of solvent from the droplets (Figure 1.7a,b) [129]. By comparison, they demonstrated that microfluidics-based, monodisperse MPs exhibited significantly reduced burst release and slower release rates than conventional, polydisperse MPs of similar average sizes and overall loading of drug. Therefore, the ability of droplet microfluidics to produce monodisperse particles for drug delivery has several practical advantages. In addition, the reduced shear stresses used to prepare particles in a microfluidic device may assist in maintaining the bioactivity of shear-sensitive biomolecular drugs. Given that the conventional emulsion approach typically produces aggregates that must be removed by filtration, particles prepared using the microfluidic method can be produced with higher yields, which is also a significant advantage particularly for expensive drugs. Single emulsions do not ensure the simultaneous loading of multiple therapeutics, especially when the payloads present different solubility. Therefore, double emulsions are also widely used in drug delivery applications. Li et al. presented a new type of MPs with gelatin methacrylate (GelMa) cores and PLGA shells for synergistic and sustained drug delivery applications (Figure 1.7c) [130]. The MPs were fabricated by using GelMa aqueous solution and PLGA oil solution as the raw materials of the microfluidic double-emulsion templates, in which hydrophilic and hydrophobic actives, such as doxorubicin hydrochloride (DOX, hydrophilic) and camptothecin (CPT, hydrophobic), could be loaded, respectively. As the inner cores were polymerized in the microfluidics during the formation of the double emulsion, the solute actives could be trapped in the cores with high efficiency, and the rupture or fusion of the cores could be avoided during the solidification of the MP shells with other actives. It was also demonstrated that the core–shell MPs with DOX and CPT codelivery could significantly reduce the viability of liver cancer cells. These features made the solid core–shell MPs ideal for synergistic and sustained drug delivery applications.

19

Outer fluid

(d) Middle fluid

(a) (b)

Ethanol solution

Inner fluid

Injection tube

Collection tube

Transition tube

(D)

(C) Washing

(c)

(e) Aqueous solution

(A)

Microcapsules in water

(B) Crosslinking

Microfluidic chip

O/W/O emulsions

Microcapsules in oil

Drug loading

(E)

pHpKa

T

(G)

CHOL DOPC

T

pH-responsive chitosan microcapsule membrane

Magnetic nanoparticle

Temperature-responsive sub-microsphere

Encapsulated drug

DPPC Membrane Proteins

Figure 1.7 (a, b) Schematic illustration of the procedure to fabricate monodisperse polymer MPs (a), and scanning electron microscopy images of monodisperse PLGA MPs with a narrow size distribution (b). Source: Adapted with permission from Xu et al. [129]. Copyright 2009, John Wiley & Sons. (c) Schematic diagram of a capillary microfluidic system for generating the W/O/W double-emulsion templates with polymerized cores and the fabrication process of the drug-loaded GelMa–PLGA core–shell MPs. Source: Adapted with permission from Li et al. [130]. Copyright 2017, Springer Nature. (d) Schematic illustration of fabrication process and controlled release mechanism of the proposed multi-stimuli-responsive microcapsules. Source: Adapted with permission from Wei et al. [131]. Copyright 2014, John Wiley & Sons. (e) Schematic process for the assembly of biomimetic nanovesicles using NanoAssemblr platform. Source: Adapted with permission from Molinaro et al. [132]. Copyright 2018, John Wiley & Sons. (f ) Schematic of the two-stage microfluidic chip for synthesizing the lipid-PLGA hybrid NPs. Source: Adapted with permission from Feng et al. [133]. Copyright 2015, American Institute of Physics. (g) High-resolution scanning electron microscopy images of TOPSi NPs, TOPSi@AcDEX nanosystems, and TOPSi@AcDEX@CCM nanovaccines. Source: Adapted with permission from Fontana et al. [134]. Copyright 2017, John Wiley & Sons.

(f) PLGA

Water

Lipid-PEG PLGA-NPs

Lipid-PEG Lipid

Figure 1.7 (Continued)

(g)

22

1 Micro/Nanostructured Materials from Droplet Microfluidics

In passive modes, drug release depends on molecular diffusion and matrix degradation, and the release profile generally shows an initial burst step and a following sustained pattern. MPs with multi-stimuli-responsive properties have been prepared by droplet microfluidics to achieve enhanced control over drug release. For instance, based on W/O/W emulsions, Wei et al. fabricated microcapsules composed of cross-linked chitosan acting as a pH-responsive capsule membrane (Figure 1.7d) [131]. When the local pH was lower than the pK a of chitosan, the membrane swelled, resulting in a high drug release rate. The release rate could be further tuned by varying the interspace distance between the nanosphere in the capsule membrane, which was achieved by the temperature-regulated volume change of the nanospheres. In addition, the magnetic NPs were embedded to realize “site-specific targeting,” and the temperature-responsive submicrospheres were embedded to serve as “microvalves.” Therefore, this kind of multi-stimuli-responsive MPs provided a new direction for designing “intelligent” controlled release systems and expected to realize more rational drug administration. NMs with unique physical and chemical properties can also serve as drug carriers. In addition, more recently, advances in biomimicry, i.e. the biologically inspired design of materials, has spurred the development of novel strategies to bestow NMs with multiple functionalities necessary to negotiate biological barriers. Current approaches for drug delivery carriers include mimicking of leukocytes, red blood cell platelets, and cancer cells to achieve superior delivery of therapeutics compared to conventional NMs. These biomimetic strategies demonstrated innate biological features and intrinsic functionalities typical of the donor cell source. For example, leukocyte-like nanovesicles showed prolonged circulation and preferential targeting of inflamed vasculature, while platelet-like NPs displayed platelet-mimicking properties such as adhesion to damaged vasculature and binding to platelet-adhering pathogens. Molinaro et al. successfully applied the microfluidics-based NanoAssemblr platform for the incorporation of membrane proteins within the bilayer of biomimetic nanovesicles (leukosomes) (Figure 1.7e) [132]. The physical, pharmaceutical, and biological properties of microfluidics-formulated leukosomes (called NA-Leuko) were characterized, which showed extended shelf life and retention of the biological functions of donor cells (i.e. macrophage avoidance and targeting of inflamed vasculature). Thus, the microfluidic approach represents as a universal, versatile, robust, and scalable tool, which is extensively used for the manufacturing of biomimetic nanovesicles. Core–shell NPs have gained increasing interest for drug delivery because of their high flexibility and biocompatibility. However, it is too complex and laborious to synthesize controlled core–shell NPs in bulk approaches. Microfluidic systems integrated with the precise flow control and hydrodynamic flow focusing have been applied to fabricate size-tunable lipid–polymer NPs. Jiang’s group designed a two-stage, high-throughput microfluidic chip to fabricate monodisperse lipid–polymer NPs with a controlled size (Figure 1.7f ) [133]. The core of the NPs was a PLGA polymer and the shell consisted of dipalmitoyl phosphatidylcholine (DPPC), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-Npoly(ethylene glycol) (DSPE-PEG), and cholesterol. It showed that the higher the flow rate, the better the mixing performance was, which led to smaller sized

1.4 Applications of the Droplet-Derived Materials

NPs in the microfluidic platform. In vitro experiments showed that the large hybrid NPs more likely to be aggregated in serum and exhibit a lower cellular uptake efficacy than the smaller ones. In another example, Fontana et al. adopted the glass capillary microfluidic nanoprecipitation technique to produce the core–shell NPs. Thermally oxidized porous silicon (TOPSi) NPs were encapsulated into acetalated dextran (AcDEX) or spermine-modified acetalated dextran (SpAcDEX) polymeric particles. The particles coated with AcDEX were then co-extruded together with vesicles derived from cancer cell membrane (CCM) to obtain the final core–shell system (TOPSi@AcDEX@CCM) (Figure 1.7g) [134]. TOPSi@SpAcDEX particles were functionalized with a model antigen, Trp2, to provide the second system (TOPSi@SpAcDEX-Trp2). They employed this model antigen to evaluate the ability of the system to work as an adjuvant. The adjuvant NPs presented high monodispersity because of the efficient mixing produced in the microfluidic device and were shown to be highly cytocompatible. In addition, the NPs induced the expression of costimulatory signals both in the immortal cell lines and in peripheral blood monocytes. TOPSi@AcDEX@CCM greatly enhanced the secretion of IFN-γ in peripheral blood monocytes and did not induce the secretion of IL-4, thereby orienting the polarization of the newly primed T-cells toward a Th1 cell-mediated response. Therefore, the developed nanovaccines showed promising adjuvant properties and the possibility of encapsulating the nanosystems with materials derived from the patient’s tumor opened new prospects in the field of personalized cancer medicine. 1.4.2

Cell Microencapsulation

Cell microencapsulation is a decade-old concept with many applications, such as cell therapy, cell biosensors, cell immobilization for antibody production, and probiotic encapsulation by the food industry [135, 136]. In vitro recapitulation of the 3D microenvironment that cells face in the human body has emerged as a powerful approach for addressing biomedical challenges such as control of stem cell fate or assessment of drug efficacy. In order to achieve this goal, biochemical and biophysical aspects of the 3D cellular microenvironment are typically imitated. Recently, microfluidics-enabled cell encapsulation has emerged as an interesting strategy to construct hydrogels and establish customized cellular microenvironments. Based on the microfluidic emulsification in which highly monodisperse surfactant-stabilized aqueous microdroplets are formed at the junction of microfluidic channels or in coaxial capillaries, the high-throughput handling of cells could be achieved. In the simplest case, cell-laden microgels can be prepared by encapsulating cells in W/O emulsions, followed by gelation. For example, Kapourani et al. employed biorthogonal, microfluidic templating to produce three different cell-laden polyglycerol-based matrices (Figure 1.8a–c) [137]. Star-shaped polyglycerol hexaazide, α,𝜔-bis azido-linear polyglycerol or polyethylene glycol, as well as dendritic polyglycerol-(polycyclooctyne) served as bioinert hydrogel precursors. Because of the multifunctional nature of the cytocompatible polymeric building blocks and the microfluidic patterning, the authors demonstrated the generation of entirely polyglycerol-based microcapsules with excellent stability and full retention of viability of the packed cells for longer than three weeks.

23

(a)

Microdroplet fabrication

(b)

Microgel formation

(c)

Microgel formation

(d)

Core

Solid shell

50 μm

Cells Thiol-one (OSTE) monomer PEGDA monomer Dispersed phase Continuous phase Crosslinked PEGDA

Cell Initiator Radical

UV (365 nm) exposure area

Shell Flow direction

shell

cells

1 Droplets containing cells,

2 UV exposure creates PEGDA monomers and radicals from initiators and initiators are formed via flow PEGDA monomers in the focusing in a continuous phase droplet start to polymerize. of OSTE monomers. Meanwhile, the radicals are diffusing in the network.

Aqueous shell flow

(e) W1

Aqueous core flow

(f)

Oil flow

(g) Shell (thin)

Oil

H1 H3

W2

core

The PEGDA inside the 3 As the droplet droplet and the OSTE polymerizes, radicals start to migrate outside around the droplet boundary the droplet boundary, are polymerized simultanecrosslinking a thin shell ously, creating an evenly distributed core-shell of OSTE at the droplet structure. boundary.

H2 H1

Shell (thick) Core

H2

Core W3 H3 100 μm

100 μm

Figure 1.8 (a–c) Droplet-templated cell encapsulation: a set of azide-functionalized precursors and dPG-polyC was prepared (a), cell-laden microgel particles deriving from S-PG-hexaazide (b), and from α,𝜔-bis-azido-LPG and α,𝜔-bis-azido-PEG (c), were formed, respectively. Source: Adapted with permission from Kapourani et al. [137]. Copyright 2018, John Wiley & Sons. (d) Illustration of the one-step curing mechanism for core–shell particle synthesis in a droplet microfluidic setup, and microscopy image of the generated gel core–solid shell particle that contains cells. Source: Adapted with permission from Zhou et al. [138]. Copyright 2018, John Wiley & Sons. (e–g) Nonplanar microfluidic flow-focusing device for one-step generation of core–shell microcapsules from two aqueous fluids (e), typical phase contrast images of microcapsules of ∼240 μm with a thin (∼10 μm) shell (f ), and ∼260 μm with a thicker (∼50 μm) shell (g). Source: Adapted with permission from Agarwal et al. [139]. Copyright 2013, Royal Society of Chemistry

1.4 Applications of the Droplet-Derived Materials

Other than simple spherical particles, microcarriers with complex shapes and microstructures are also designed. Microcapsules templated from W/O/W double emulsions have been generated for cell encapsulation and can be categorized into two types: matrix-core/shell and liquid-core/shell microcapsules. Incorporating cells in the core of the microcapsules enhances their resistance against external effects, such as enzymatic attack and UV irradiation, and prevents them from regressing. For example, Zhou et al. introduced cell encapsulation in core–shell particles with a synthetic polyethylene glycol diacrylate (PEGDA) hydrogel core and a solid off-stoichiometry thiol-ene polymer (OSTE) shell [138]. The MPs were synthesized by using a droplet microfluidic chip with a flow focusing junction to generate droplets that consisted of a dispersed phase of PEGDA monomers, cells, and UV initiator and a continuous phase of OSTE monomers (Figure 1.8d). Core–shell cell encapsulation was achieved after a single UV exposure. The encapsulated human cells in 100 μm diameter particles had >90% viability. The average shell thickness was controlled between 7 and 13 μm by varying the UV exposure, and the shell was measured to be permeable to low-molecular-weight species (480 Da). The unique material properties and the orthogonal control of the MP core size, shell thickness, shell permeability, and shell surface properties addressed the key unresolved challenges in the field and were expected to promote translation of novel cell therapy concepts from research to clinical practice. Microcapsules with liquid–core/shell structure allow encapsulated cells to form cell aggregates in the liquid core because of enhanced cell–cell interactions. This strategy is of special value for stem cell studies. For example, the comparison between cells cultured in solid microbeads and liquid–core/shell microcapsules showed that the latter allowed formation of single spherical embryoid body cells within two days. In contrast, cells in the solid microbeads only formed bumpy shapes from several clusters of cells. Agarwal et al. microfabricated a 3D microfluidic flow-focusing device to achieve one-step generation of core–shell microcapsules with an alginate hydrogel shell of controllable thickness and an aqueous liquid core of embryonic stem (ES) cells without using any cytotoxic chemicals or organic solvents (Figure 1.8e) [139]. The core–shell architecture of the microcapsules resembled that of a prehatching embryo where ES cells resided naturally (Figure 1.8f,g). ES cells encapsulated in the liquid core of the microcapsules were found to survive well (>92%) and proliferate to form a single aggregate in each microcapsule within seven days. Furthermore, the aggregated cells were found to have significantly higher expression of pluripotency marker genes compared to the ES cells cultured on 2D substrates and they could be efficiently differentiated into beating cardiomyocytes under the induction of a single small molecule without complex combination of multiple growth factors. 1.4.3

Tissue Engineering

In the past decade, TE has been demonstrated to be increasingly potential for creating true biological alternatives for artificial implants and prostheses for becoming an on-demand regenerative tool alternative to harvested tissue/organ

25

26

1 Micro/Nanostructured Materials from Droplet Microfluidics

transplantation. The underlying concept of TE consists of few steps in which healthy cells are first isolated from a patient’s biopsy, expanded in vitro, and then seeded/encapsulated into a carrier. The resulting engineered construct is often precultured in vitro and then grafted back into the patient to regenerate and/or replace the damaged tissue [140, 141]. Tissues are integrated 3D structures of multiple types of cells and ECMs. The function of a tissue is typically governed by multiple cues, such as intercellular signaling and cell interactions with the surrounding ECMs. Cell-laden microgel “modules” carrying different types of cells can be combined or reconfigured to mimic various types of tissues. These microgels generated from microfluidics serve as building blocks for construction of TE scaffolds. Their applications in TE, including organ-on-a-chip, bone/cartilage regeneration, stem cell culture, and therapy, have been extensively studied [142, 143]. The small size of the hydrogel MPs is particularly attractive for injectable cell delivery systems in regenerative medicine, as it allows direct delivery of cells through needles to the damaged tissue area. The direct injection minimizes surgical invasiveness and thus is beneficial in practical clinical applications. Hou et al. developed injectable degradable PVA microgels using a microfluidic approach for encapsulation of mesenchymal stem cells (MSCs) and bone morphogenetic protein-2 (BMP-2), which provided favorable microenvironments for cell proliferation and controlled osteogenic differentiation in vitro (Figure 1.9a–c) [144]. PVA microgels formed with various polymer concentrations exhibited different degradation and mechanical properties as well as the release profiles of growth factors. The mild cross-linking conditions and cell-compatible materials facilitated the encapsulation of MSCs with high bioactivity, which ensured prolonged cell survival, proliferation, and migration. Additionally, BMP-2 coencapsulated into the microgel environments enhanced osteogenic differentiation of MSCs. In another separate work, bioactive fibrin microbeads with bioactive molecules were produced using a droplet microfluidic platform. α2 PI1-8 -MMP-IGF-1-conjugated microbeads were mixed with genipin-cross-linked collagen, resulting in a novel injectable bulking agent, which promoted human urinary tract smooth muscle cell (hSMC) migration in vitro (Figure 1.9d) [145]. This injectable material showed similar rheological characteristics as the reference sample. The regenerative potential of this bioactive bulking agent was evaluated in a rabbit model and might promote functional regeneration of human urinary tract smooth muscle tissue for long-term treatment of stress urinary incontinence. With the advantages of using microfluidics to fabricate MPs, the dimensions of the microspheres could be precisely controlled to form desirable scaffold structures. However, it was not easy to fabricate ultra-high-porosity scaffolds because the porosity of the scaffolds is determined by the gaps between microspheres after packing. For microspheres with diameters of ≈100 μm, the gap was only several micrometers. Hollow microspheres were investigated because they have more adjustable parameters to provide greater flexibility as well as a larger contact surface area to interact with cells. Yu et al. used a droplet-based microfluidic process to fabricate hollow bacterial cellulose (BC) MPs by culturing Gram-negative bacterium Gluconacetobacter xylinus (G. xylinus) inside

1.4 Applications of the Droplet-Derived Materials

(a) O

x

yn OH

PVA-VEA-SH

Oil

O

O

O

(b)

y n OH

x O

O

O

Cells and growth factor

O S

PVA-VEA

O 2

SH

PVA-VEA-SH

PVA-VEA

Michael-type addition

Cells

Growth factors

Oil

(c)

Cells encapsulated In PVA hydrogel microcapsules

Oil

Fibrinogen

(d)

T junction

8 °C

Oil

Fibrinogen

FXllla + Thrombin +α2Pl1-8-MMP-IGF-1

RT

Fibrin microbeads in collection chamber

Figure 1.9 (a–c) Schematic illustration of PVA microgels formed by PVA–VEA and PVA–VEA–SH through a Michael-type addition reaction for encapsulation of stem cells and growth factors (a); image of the microfluidic device used for PVA microdroplet formation, the arrows indicated the cells suspending in the culture medium (b); confocal microscopy images of MSCs cultured in PVA microgels, live cells were labeled with calcein AM (green) (c). Source: Adapted with permission from Hou et al. [144]. Copyright 2018, Elsevier. (d) Schematic of microfluidic channels of the microchip used for fibrin microbead fabrication and bright-field microscopy image of the microbeads. Source: Adapted with permission from Vardar et al. [145]. Copyright 2018, Elsevier. (e) The microfluidic process for producing the double-layer alginate core agarose shell microdroplet as cellulose secretion template and the following steps to produce hollow BC microsphere including gelling, cellulose secretion, purification, and the application of the microsphere as a cell culture scaffold in vitro and an injectable scaffold for wound healing in vivo. Source: Adapted with permission from Yu et al. [146]. Copyright 2016, John Wiley & Sons.

a double-layer hydrogel template [146]. The generated microgels were washed and stored under static culture conditions for the production of cellulose from G. xylinus. The cellulose fibers became gradually entangled and confined in the shell part of the particles, which formed the desirable hollow morphology. The template gel was then removed using thermal and chemical treatments

27

28

1 Micro/Nanostructured Materials from Droplet Microfluidics

(e)

Culture medium

Oil

Agarose monomer

Alginate particle Bacteria

Agarose gelling

Removing agarose template

Cellulose secretion

Assembly of microspheres as a 3-D scaffold for cell culture

Cellulose microsphere formation

Removing alginate core and bacteria

Application as an Injectable porous scaffold for wound healing

Figure 1.9 (Continued)

without affecting the cellulose, thus generating hollow BC microspheres. With the advantage of hollow BC microspheres, scaffolds with high porosity were developed for 3D cell culture and could be used as an injectable scaffold in vivo for wound healing (Figure 1.9e). Currently, droplet microfluidics is mainly based on the W/O system, in which the organic reagents are used as continuous phase. However, the system is limited by the addition of the oil phase and/or surfactants, thus limiting their applications in TE. Liu et al. described a new microfluidic strategy for controllable and high-throughput generation of monodispersed W/W droplets

1.4 Applications of the Droplet-Derived Materials

[147]. Solutions of polyethylene glycol and dextran were used as continuous and dispersed phases, respectively, without any organic reagents or surfactants. The size of W/W droplets could be precisely adjusted by changing the flow rates of dispersed and continuous phases and the valve switch cycle. In addition, uniform cell-laden microgels were fabricated by introducing the alginate component and rat pancreatic islet cell suspension to the dispersed phase. The encapsulated islet cells retained high viability and the function of insulin secretion after cultivation for seven days. The high-throughput droplet microfluidic system with high biocompatibility was stable, controllable, and flexible, which could boost various chemical and biological applications, such as bio-oriented MPs synthesizing, microcarriers fabricating, TE, etc. 1.4.4

Biosensors

Biosensors are devices composed of biological recognition elements and signal transduction elements for quantitative and semiquantitative analysis [148, 149]. An ideal biosensor can detect analytes, such as glucose, enzymes, DNA, and antibodies in a rapid, efficient, and convenient manner. They are increasingly in demand for many biomedical applications from fundamental biological studies to clinical diagnostics. Metal NP-based sensors have been extensively studied. These biosensors are based on the localized surface plasmon resonance (LSPR) phenomenon of metal NPs and is highly associated with the size, shape, dielectric properties, aggregate morphology, surface modification, and refractive index of the surrounding medium [150]. AuNPs and AgNPs are the most common NPs fabricated for LSPR, and the synthesis of other metal NPs such as Pt and copper has also been attempted. Microfluidic synthesis offers fine-tuning and convenience in adding new agents for multistep reactions. For instance, Lohse et al. demonstrated a simple microfluidic reactor constructed from commercial components for generate hydrophilic functionalized AuNPs (Figure 1.10A) [151]. The authors showed that the synthesis of AuNPs in the reactor could be fine-tuned to control the aspect ratios and absolute dimensions. The reactor could also be easily integrated with UV–vis absorbance spectroscopy analysis to monitor in real time and to analyze the purity of AuNPs as quality control. Binary noble metal NPs have been synthesized for LSPR based on microfluidic system. Knauer et al. generated noble metal NPs with multishell structures with a two-step microreactor (Figure 1.10B) [152]. It was shown that a more uniform particle growth was reached by the use of microfluidic techniques for the synthesis of multishell NPs. The improvement of the particle size distribution corresponded to an enhanced quality of the optical spectra. The absorption spectra of the particle dispersions synthesized in the microflow-based system showed a blue shift combined with a narrower bandwidth, which clearly emerged in the case of Au/Ag/Au core/double-shell particle. As expected, the spectral position of the plasmonic resonance shifted dramatically by covering a gold seed particle with an alternating shell structure of gold and silver. Thus, the position of the resonance peak could be adjusted within a wide range of optical spectra

29

1 Micro/Nanostructured Materials from Droplet Microfluidics

(A)

Flow reactor set-up for AuNP synthesis

(B)

Carrier medium

A 10 cm

B

c Flow reactor set-up for AuNP functionalization

b Standard Mixer Geometry

Reaction initiation by thermal activation

150 cm

Y-Mixer

(C) H2C H 2C

CH3 Cl– N+ CH3

Microreactor

Aqueous phase

radicals

O

H2C

OCH3 CH3

Emulsion polymerization

(a)

Organic phase During ongoing polymerization

particles

Single-step process

SDS domain

(Mixture of monomers, cross-linker, initiator)

In water

Water evaporated

Diffusion during dynamic desorption kinetics

Through nanopores

In water Shrinking

SDS DADMAC

Swelling

(b)

MMA domain

30

(c)

All is polymer

Hydrophobic–hydrophilic–hydrophobic assembly

Water evaporated

Figure 1.10 (A) The integrated millifluidic reactor used for AuNPs synthesis and functionalization. Source: Adapted with permission from Lohse et al. [151]. Copyright 2013, American Chemical Society. (B) General setup for microsegmented flow synthesis of bimetallic multishell NPs. Source: Adapted with permission from Knauer et al. [152]. Copyright 2011, Elsevier. (C) Schematic overview of in situ nanoarchitecture of the copolymer NPs by emulsion polymerization via micro-flowthrough technique, interparticles nanoassembling of polymer particles during ongoing polymerization, and controlled hydrophobic/hydrophilic/ hydrophobic electrostatic nanoassembly particles. Source: Adapted with permission from Visaveliya and Köhler [150]. Copyright 2015, John Wiley & Sons. (D) Schematic illustration of the buckling of semipermeable capsules subjected to positive osmotic pressure, and confocal microscopy images of a mixture of four distinct microcapsules dispersed in an aqueous solution of 475 mOsm/l, whereby the red, orange, yellow, and green capsules contain aqueous solutions of 33, 342, 649, and 950 mOsm/l, respectively. Source: Adapted with permission from Kim et al. [153]. Copyright 2014, John Wiley & Sons.

1.4 Applications of the Droplet-Derived Materials

(D)

Buckling

V0–ΔV *

V0

Isotropic shrinkage

Outward flux of water

200 μm

V0–ΔV, Cin~Cout

Cout

100 μm

Figure 1.10 (Continued)

only by the deposition of a next metal shell. The investigations showed that the micro-continuous-flow synthesis was well suited for the preparation of composite plasmonic NPs. The polymer/metal NPs could also be applied as biosensors. For example, Visaveliya and Köhler presented a semimicrofluidic approach for a single-step in situ copolymer interaction to generate core–shell-type PMMApolyDADMAC particles and their assemblies with polymer and metal NPs (Figure 1.10C) [150]. This study revealed that the mean polymer particle size could be tuned between 200 nm and 1 μm at different process conditions regulated by microfluidic platform. Interfacial copolymerization induced the controlled compartmentalization where a hydrophobic core adopted spherical shape in order to minimize the surface energy and simultaneously sheltered in the hydrophilic shell-like surface layer. Surface layer could swell in the aqueous medium and allow controlled entrapping of functional hydrophobic NPs in the hydrophilic interior via electrostatic interaction, which could be particularly interesting for combined fluorescence activity. Furthermore, the polymer–metal nanoassembly particles could be implemented as an ideal surface-enhanced Raman scattering substrate for detection of the trace amounts of various analytes. Apart from NPs, MPs could also serve as important sensors in response to environmental parameter variations. For example, Kim et al. reported a microfluidic approach to produce microcapsules with ultrathin and semipermeable membranes, providing a facile and direct measurement of the osmotic strength [153]. Using capillary microfluidic devices, W/O/W double-emulsion drops with an ultrathin middle phase were prepared, which transformed into polymeric microcapsules that contain an aqueous solution with standard osmotic strength

31

32

1 Micro/Nanostructured Materials from Droplet Microfluidics

upon solidification of the middle phase. The resultant microcapsules are highly sensitive to osmotic pressure differences because of the semipermeability and very small thickness of the membrane. Therefore, a small positive pressure could lead to buckling of the capsules by an outward flux of water through the membrane. Therefore, when a mixture of distinctively labeled microcapsules containing different standard osmotic solutions were dispersed in an unknown solution, the capsules were selectively buckled according to the positive pressure, thereby enabling an estimation of the osmotic strength (Figure 1.10D). 1.4.5

Barcodes

The increasing application of high-throughput assays in biomedical areas, including drug discovery and clinical diagnostics, demands effective strategies for multiplexing [154]. One promising strategy is the use of barcode particles that encode information about their specific compositions and enable simple identification. Microfluidics is an effective approach that has created exciting avenues of scientific research in barcode particle synthesis. The resultant particles have found important applications in the detection of multiple biological species as they have properties of high flexibility, fast reaction times, less reagent consumption, and good repeatability. For example, Gerver et al. used a customized and fully automated microfluidic device to mix different predetermined ratios of nanophosphors suspended in monomer, followed by photopolymerization, resulting in spectrally encoded polymer (Figure 1.11a,b) [155]. Using this method, the authors produced, imaged, and characterized thousands of encoded beads with 24 distinct spectral codes. The beads were highly uniform in size and had a very tight distribution of lanthanide ratios, so that investigators could distinguish between the codes with 0.1% error. These results established the practical feasibility of using lanthanide nanophosphors for spectral encoding and lay the foundation for future high-throughput multiplexing of biological assays. Quantum dots (QDs) have narrow Gaussian emission line shapes, resistance to photobleaching, high quantum yields, single-wavelength excitation, and a large number of codes. Thus, they are an ideal additive to emulsions for generating barcode particles [158]. Zhao et al. developed a new approach to prepare barcode particles by encapsulating QDs in double-emulsion templates generated in capillary microfluidic devices (Figure 1.11c) [156]. The resultant particles exhibited uniform spectral characteristics and allowed substantial numbers of coding levels for multiplexing (Figure 1.11d). In addition, the approach also enabled fabrication of anisotropic magnetic barcode particles, which could rotate under a rotating magnetic field and aggregate under a stationary magnetic field. This feature created new opportunities to perform magnetic separation of the barcode particles. Therefore, the QD-containing barcode particles were promising as microcarriers in biomedical applications, including high-throughput bioassays and cell culture research where multiplexing was needed. Photonic crystals (PhCs) are a kind of well-known photonic NMs with spatially ordered lattices that exhibit brilliant structural colors [159]. By constructing PhCs

1.4 Applications of the Droplet-Derived Materials

(a) T-junction

Herringbone mixer Water input

Oil input

(b)

Water input

Waste

T-junction

Herringbone mixer

Oil input

Waste UV

Bead output Eu Eu/Sm

Eu/Dy

Stage 1: Fill mixing channel (c)

Eu

Eu/Sm

Eu/Dy

Stage 2: Generate beads

Encapsulated material + ETPTA PEG-DA solution

O/W/O double emulsions

Hexadecane (d)

PEG hydrogel shell

QD-tagged ETPTA resin core

Figure 1.11 (a, b) Microfluidic bead synthesizer: (a) mixtures of lanthanide suspended in prepolymer bead mixture flowed into a microfluidic device at controlled ratios and were mixed on chip using a staggered herringbone mixer; (b) water pushed the lanthanide mixture toward a T-junction containing a continuously flowing oil stream, producing droplets, and were then polymerized into beads via illumination with UV light (b). Adapted with permission from Gerver et al. [155]. Copyright 2012, Royal Society of Chemistry. (c, d) Formation of O/W/O double emulsions in a glass microcapillary device (c); optical micrograph of the polymerized double emulsions with PEG hydrogel shells and QD-tagged ETPTA resin cores. Scale bar is 100 μm (d). Source: Adapted with permission from Zhao et al. [156]. Copyright 2011, American Chemical Society. (e) Schematic illustration of the microfluidic emulsification process and concentration and self-assembly of water-extraction-derived colloidal NPs into a hollow spherical PhC shell on the inner wall of the microcapsules. Adapted with permission from Shang et al. [157]. Copyright 2015, American Chemical Society.

33

34

1 Micro/Nanostructured Materials from Droplet Microfluidics

(e)

PS dispersed solution

Flow direction

Middel ETPTA

ctant

Water with surfa

Figure 1.11 (Continued)

with different structural periods or different refractive indices, a series of PhC particles with different diffraction peak positions can be obtained for encoding. Unlike the fluorescence, the spectra of the PhCs originate from the reflection of their physical structures. Thus, they are resistant to photobleaching or photoquenching, which makes them a kind of ideal color-encoding element. PhCs can also be incorporated into microfluidic droplets for generating barcode particles. Shang et al. presented a new type of barcode particle with a semipermeable membrane shell and a colloidal-NP-encapsulated liquid core, using droplet microfluidics [157]. By dispersing the microcapsules in an ethanol solution to extract water from the core, their encapsulated colloidal NPs could be gradually concentrated and self-assembled into the hollow spherical PhC shell on the inner wall of the microcapsules, as schematically shown in Figure 1.11e. The resultant PhC microbubbles were composed of an outer transparent polymeric shell, a middle PhC shell, and an inner bubble core. The encoded elements of the particles originated from their PhC structure with a coated shell, which not only improved the stability of the codes but also provided a flexible surface for bioassays. In addition, by using multicompartmental PhC microbubbles, the barcode particles allowed for a substantial number of coding levels and controllable movement in multiplexing applications. More importantly, as the size of the encapsulated core and the corresponding bubbles of the barcode particles could be tailored in their emulsification process, the overall density of the PhC microbubbles could be adjusted to match the density of a detection solution so that they remained in suspension. These features made the PhC microbubbles excellent functional barcode particles in biomedical applications.

1.5 Conclusion and Perspectives

1.5 Conclusion and Perspectives This part summarizes recent research progress on the synthesis of MMs/NMs based on droplet microfluidics and their applications in various fields. In the past decade, with the development of microchip preparation technology, materials engineered from microfluidic technologies have experienced transition from simplicity to complexity in material structure and simplification to diversity in material function, which largely compensates for the limitation of conventional synthesis methods. Based on the flexibility and integration of microfluidic manipulation, preparation of functional materials has extended from the original single emulsion droplet method to complex multiple emulsion droplet method. The resultant materials have developed from simple solid MPs to specially shaped MMs, Janus particles, porous particles, and core–shell-structured particles. A series of novel NMs has also been developed in the microfluidic system. These MMs/NMs have shown great potentials in a broad range of applications, such as drug carrier, cell encapsulation, TE, and analytical application. Because of the theoretical research and technological innovations, droplet microfluidics currently bears significant value in an extremely wide range of areas. Despite a lot of exciting and compelling developments, there remain challenges that pose a gap between academic proof-of-concept studies and practical techniques for addressing real-world problems. Therefore, several important issues need to be solved to achieve the wide applicability of droplet microfluidics. Firstly, since current works mostly adopt single channel as the reaction unit, the yield of microfluidic synthetic materials is low compared to conventional batch mode. However, mass production is an inevitable topic to bring droplet microfluidics out of the laboratory as an industrialized technology. Fortunately, there is already some research on the scale-up of droplet microfluidic platforms, based on coaxial annular interfaces, stacking multiple generator layers, and parallel multiple modular reactors [160, 161]. The second issue is about technology promotion. The current droplet systems are usually run in specialized academic laboratories, which restrict the access to microfluidics for nonexperts. Regarding this, much effort should be paid to the simplification and modularization of basic functional units, and automation control should be enhanced to reduce manual operation. It is also important for the technology holders to actively cooperate with the researchers of clinical, medical, food, and environment topics and get their authority recognized. In other words, the microfluidic materials have bright market prospects, but the realization of their commercial value still requires the efforts of researchers and entrepreneurs. Third, much emphasis should be put on designing novel materials with more applications. For analytical application, millions of barcodes have been designed by using droplet microfluidics. However, these numbers mean little in biological applications. The main reason is that the largest multiplicity in protein detection is in the hundreds and this will be further limited by nonspecific reactions because of the cross-reaction of antibodies. Meanwhile, for gene analysis, the production of so many barcode particles is time consuming, and the size of these microfluidic

35

36

1 Micro/Nanostructured Materials from Droplet Microfluidics

particles is relatively large, which might require the use of a large number of samples. Therefore, in the future, microfluidic material design and preparation should be application targeted and oriented. Finally, compared with the MMs, the diversity and functionalities of the droplet microfluidics-derived NMs are still lacking. For NM synthesis, purification and extraction processes, which are easily achievable in conventional approaches, should be improved in microfluidic platforms. In addition, the integration and development of online analytical methods with the microfluidic approach will significantly improve the synthesis performance by optimizing processes with an immediate feedback control. It is envisioned that, with further endeavors, droplet microfluidics is expected to be a promising platform to optimize and tailor NMs for different applications. In conclusion, droplet microfluidics has evolved into a powerful technique and possesses application values cutting across multiple fields and disciplines. However, a lot of effort is still needed to overcome existing challenges. In particular, in-depth collaborative efforts and communication from different areas should be aimed to bridge the gap between material synthesis and applications. After addressing the issues described above, we firmly believe that more exciting accomplishments will be achieved in droplet microfluidics.

References 1 Nge, P.N., Rogers, C.I., and Woolley, A.T. (2013). Advances in microflu-

2 3

4 5 6

7 8

9

idic materials, functions, integration, and applications. Chem. Rev. 113 (4): 2550–2583. Marre, S. and Jensen, K.F. (2010). Synthesis of micro and nanostructures in microfluidic systems. Chem. Soc. Rev. 39 (3): 1183–1202. Elvira, K.S., Casadevall i Solvas, X., Wootton, R.C., and deMello, A.J. (2013). The past, present and potential for microfluidic reactor technology in chemical synthesis. Nat. Chem. 5 (11): 905–915. Wang, X.H., Liu, J.F., Wang, P.Z. et al. (2018). Synthesis of biomaterials utilizing microfluidic technology. Genes 9 (6): 283. Hayat, Z. and El Abed, A.I. (2018). High-throughput optofluidic acquisition of microdroplets in microfluidic systems. Micromachines 9 (4): 183. Kaminski, T.S. and Garstecki, P. (2017). Controlled droplet microfluidic systems for multistep chemical and biological assays. Chem. Soc. Rev. 46 (20): 6210–6226. Gong, X., Wang, Y.W., Ihli, J. et al. (2015). The crystal hotel: a microfluidic approach to biomimetic crystallization. Adv. Mater. 27 (45): 7395–7400. Gijs, M.A., Lacharme, F., and Lehmann, U. (2010). Microfluidic applications of magnetic particles for biological analysis and catalysis. Chem. Rev. 110 (3): 1518–1563. Nguyen, N.T., Hejazian, M., Ooi, C.H., and Kashaninejad, N. (2017). Recent advances and future perspectives on microfluidic liquid handling. Micromachines 8 (6): 186.

References

10 Gholizadeh, S., Shehata Draz, M., Zarghooni, M. et al. (2017). Microfluidic

11

12 13

14

15

16

17 18

19

20

21 22 23 24

25

26

approaches for isolation, detection, and characterization of extracellular vesicles: current status and future directions. Biosens. Bioelectron. 91: 588–605. Wang, W., Zhang, M.J., and Chu, L.Y. (2014). Functional polymeric microparticles engineered from controllable microfluidic emulsions. Acc. Chem. Res. 47 (2): 373–384. Tumarkin, E. and Kumacheva, E. (2009). Microfluidic generation of microgels from synthetic and natural polymers. Chem. Soc. Rev. 38 (8): 2161–2168. Lee, S.S., Kim, B., Kim, S.K. et al. (2015). Robust microfluidic encapsulation of cholesteric liquid crystals toward photonic ink capsules. Adv. Mater. 27 (4): 627–633. Mei, L., Jin, M., Xie, S. et al. (2018). A simple capillary-based open microfluidic device for size on-demand high-throughput droplet/bubble/microcapsule generation. Lab Chip https://doi.org/10.1039/c8lc00479j. Ferreira, J., Castro, F., Rocha, F., and Kuhn, S. (2018). Protein crystallization in a droplet-based microfluidic device: hydrodynamic analysis and study of the phase behaviour. Chem. Eng. Sci. 191: 232–244. Caggioni, M., Traini, D., Young, P.M., and Spicer, P.T. (2018). Microfluidic production of endoskeleton droplets with controlled size and shape. Powder Technol. 329: 129–136. Shang, L., Cheng, Y., and Zhao, Y. (2017). Emerging droplet microfluidics. Chem. Rev. 117 (12): 7964–8040. Jiao, Y., Liu, Y., Luo, D. et al. (2018). Microfluidic-assisted fabrication of clay microgels for cell-free protein synthesis. ACS Appl. Mater. Interfaces https:// doi.org/10.1021/acsami.8b09324. Vladisavljevic, G.T., Kobayashi, I., and Nakajima, M. (2012). Production of uniform droplets using membrane, microchannel and microfluidic emulsification devices. Microfluid. Nanofluid. 13 (1): 151–178. Ma, J.Y., Wang, Y.C., and Liu, J. (2017). Biomaterials meet microfluidics: from synthesis technologies to biological applications. Micromachines 8 (8): 255. Abate, A.R. and Weitz, D.A. (2009). High-order multiple emulsions formed in poly(dimethylsiloxane) microfluidics. Small 5 (18): 2030–2032. Chu, L.Y., Utada, A.S., Shah, R.K. et al. (2007). Controllable monodisperse multiple emulsions. Angew. Chem. Int. Ed. 46 (47): 8970–8974. Ceylan, H., Yasa, I.C., and Sitti, M. (2017). 3D chemical patterning of micromaterials for encoded functionality. Adv. Mater. 29 (9): 1605072. Leijten, J., Rouwkema, J., Zhang, Y.S. et al. (2016). Advancing tissue engineering: a tale of nano-, micro-, and macroscale integration. Small 12 (16): 2130–2145. Lash, M.H., Jordan, J.C., Blevins, L.C. et al. (2015). Non-Brownian particle-based materials with microscale and nanoscale hierarchy. Angew. Chem. Int. Ed. 54 (20): 5854–5858. Ghanbarzadeh, B., Oleyaei, S.A., and Almasi, H. (2015). Nanostructured materials utilized in biopolymer-based plastics for food packaging applications. Crit. Rev. Food Sci. Nutr. 55 (12): 1699–1723.

37

38

1 Micro/Nanostructured Materials from Droplet Microfluidics

27 Laval, C., Bouchaudy, A., and Salmon, J.B. (2016). Fabrication of microscale

28

29 30

31 32

33

34

35 36

37

38

39

40

41 42

materials with programmable composition gradients. Lab Chip 16 (7): 1234–1242. Li, F., Du, M., and Zheng, Q. (2016). Dopamine/silica nanoparticle assembled, microscale porous structure for versatile superamphiphobic coating. ACS Nano 10 (2): 2910–2921. Peng, F., Su, Y., Zhong, Y. et al. (2014). Silicon nanomaterials platform for bioimaging, biosensing, and cancer therapy. Acc. Chem. Res. 47 (2): 612–623. Makgwane, P.R. and Ray, S.S. (2014). Synthesis of nanomaterials by continuous-flow microfluidics: a review. J. Nanosci. Nanotechnol. 14 (2): 1338–1363. Shiba, K. and Ogawa, M. (2018). Precise synthesis of well-defined inorganic–organic hybrid particles. Chem. Rec. 18: 950–968. Stolzenburg, P., Lorenz, T., Dietzel, A., and Garnweitner, G. (2018). Microfluidic synthesis of metal oxide nanoparticles via the nonaqueous method. Chem. Eng. Sci. 191: 500–510. Sun, L., Wang, J., Yu, Y. et al. (2018). Graphene oxide hydrogel particles from microfluidics for oil decontamination. J. Colloid Interface Sci. 528: 372–378. Connacher, W., Zhang, N., Huang, A. et al. (2018). Micro/nano acoustofluidics: materials, phenomena, design, devices, and applications. Lab Chip 18 (14): 1952–1996. Liu, Y. and Jiang, X. (2017). Why microfluidics? Merits and trends in chemical synthesis. Lab Chip 17 (23): 3960–3978. Liu, Z.M., Yang, Y., Du, Y., and Pang, Y. (2017). Advances in droplet-based microfluidic technology and its applications. Chin. J. Anal. Chem. 45 (2): 282–295. McClements, D.J. (2018). Encapsulation, protection, and delivery of bioactive proteins and peptides using nanoparticle and microparticle systems: a review. Adv. Colloid Interface Sci. 253: 1–22. Naghizadeh, Z., Karkhaneh, A., and Khojasteh, A. (2018). Simultaneous release of melatonin and methylprednisolone from an injectable in situ self-crosslinked hydrogel/microparticle system for cartilage tissue engineering. J. Biomed. Mater. Res. A 106 (7): 1932–1940. Chaurasiya, B., Huang, L., Du, Y. et al. (2018). Size-based anti-tumoral effect of paclitaxel loaded albumin microparticle dry powders for inhalation to treat metastatic lung cancer in a mouse model. Int. J. Pharm. 542: 90–99. Rodríguez Villanueva, J. and Rodríguez Villanueva, L. (2017). Turning the screw even further to increase microparticle retention and ocular bioavailability of associated drugs: the bioadhesion goal. Int. J. Pharm. 531 (1): 167–178. Mooberry, M.J. and Key, N.S. (2016). Microparticle analysis in disorders of hemostasis and thrombosis. Cytometry, Part A 89 (2): 111–122. Jy, W., Horstman, L.L., and Ahn, Y.S. (2010). Microparticle size and its relation to composition, functional activity, and clinical significance. Semin. Thromb. Hemost. 36 (8): 876–880.

References

43 Zhang, L., Yang, W., Hu, C. et al. (2018). Properties and applications of

44

45

46 47 48

49

50

51

52

53

54

55

56

57

58

nanoparticle/microparticle conveyors with adjuvant characteristics suitable for oral vaccination. Int. J. Nanomed. 13: 2973–2987. Ghosh Dastidar, D., Saha, S., and Chowdhury, M. (2018). Porous microspheres: synthesis, characterisation and applications in pharmaceutical & medical fields. Int. J. Pharm. 548 (1): 34–48. Wong, C.Y., Al-Salami, H., and Dass, C.R. (2018). Microparticles, microcapsules and microspheres: a review of recent developments and prospects for oral delivery of insulin. Int. J. Pharm. 537: 223–244. Lee, B.K., Yun, Y., and Park, K. (2016). PLA micro- and nano-particles. Adv. Drug Deliv. Rev. 107: 176–191. Li, W., Zhang, L., Ge, X. et al. (2018). Microfluidic fabrication of microparticles for biomedical applications. Chem. Soc. Rev. 47 (15): 5646–5683. Yeo, S.J., Park, K.J., Guo, K. et al. (2016). Microfluidic generation of monodisperse and photoreconfigurable microspheres for floral iridescence-inspired structural colorization. Adv. Mater. 28 (26): 5268–5275. Cui, J., Björnmalm, M., Liang, K. et al. (2014). Super-soft hydrogel particles with tunable elasticity in a microfluidic blood capillary model. Adv. Mater. 26 (43): 7295–7299. Seth, A., Béalle, G., Santanach-Carreras, E. et al. (2012). Design of vesicles using capillary microfluidic devices: from magnetic to multifunctional vesicles. Adv. Mater. 24 (26): 3544–3548. Deng, N.N., Yelleswarapu, M., Zheng, L., and Huck, W.T. (2017). Microfluidic assembly of monodisperse vesosomes as artificial cell models. J. Am. Chem. Soc. 139 (2): 587–590. Abate, A.R., Kutsovsky, M., Seiffert, S. et al. (2011). Synthesis of monodisperse microparticles from non-Newtonian polymer solutions with microfluidic devices. Adv. Mater. 23 (15): 1757–1760. Arcos-Turmo, I., Herrada, M.A., Lopez-Herrera, J.M. et al. (2018). Novel swirl flow-focusing microfluidic device for the production of monodisperse microbubbles. Microfluid. Nanofluid. 22 (8): 79. Liang, S.S., Li, J., Li, X.M. et al. (2018). Microfluidic fabrication of ceramic microspheres with controlled morphologies. J. Am. Ceram. Soc. 101 (9): 3787–3796. Visaveliya, N. and Kohler, J.M. (2015). Simultaneous size and color tuning of polymer microparticles in a single-step microfluidic synthesis: particles for fluorescence labeling. J. Mater. Chem. C 3 (4): 844–853. Liu, H., Qian, X., Wu, Z.J. et al. (2016). Microfluidic synthesis of QD-encoded PEGDA microspheres for suspension assay. J. Mater. Chem. B 4 (3): 482–488. Jeong, W.J., Kim, J.Y., Choo, J. et al. (2005). Continuous fabrication of biocatalyst immobilized microparticles using photopolymerization and immiscible liquids in microfluidic systems. Langmuir 21 (9): 3738–3741. Tan, W.H. and Takeuchi, S. (2007). Monodisperse alginate hydrogel microbeads for cell encapsulation. Adv. Mater. 19 (18): 2696–2701.

39

40

1 Micro/Nanostructured Materials from Droplet Microfluidics

59 Yadavali, S., Jeong, H.H., Lee, D., and Issadore, D. (2018). Silicon and glass

60

61

62 63

64

65

66 67

68

69 70 71 72

73

74

75

very large scale microfluidic droplet integration for terascale generation of polymer microparticles. Nat. Commun. 9 (1): 1222. Nisisako, T., Torii, T., Takahashi, T., and Takizawa, Y. (2006). Synthesis of monodisperse bicolored Janus particles with electrical anisotropy using a microfluidic co-flow system. Adv. Mater. 18 (9): 1152. Liu, M., Sun, X.T., Yang, C.G., and Xu, Z.R. (2016). On-chip preparation of calcium alginate particles based on droplet templates formed by using a centrifugal microfluidic technique. J. Colloid Interface Sci. 466: 20–27. Velasco, D., Tumarkin, E., and Kumacheva, E. (2012). Microfluidic encapsulation of cells in polymer microgels. Small 8 (11): 1633–1642. Utech, S., Prodanovic, R., Mao, A.S. et al. (2015). Microfluidic generation of monodisperse, structurally homogeneous alginate microgels for cell encapsulation and 3D cell culture. Adv. Healthc. Mater. 4 (11): 1628–1633. Jeong, H.H., Yadavali, S., Issadore, D., and Lee, D. (2017). Liter-scale production of uniform gas bubbles via parallelization of flow-focusing generators. Lab Chip 17 (15): 2667–2673. Ofner, A., Moore, D.G., Ruhs, P.A. et al. (2017). High-throughput step emulsification for the production of functional materials using a glass microfluidic device. Macromol. Chem. Phys. 218 (2): 1600472. Lee, K.J., Yoon, J., and Lahann, J. (2011). Recent advances with anisotropic particles. Curr. Opin. Colloid Interface Sci. 16 (3): 195–202. Sun, X.T., Yang, C.G., and Xu, Z.R. (2016). Controlled production of size-tunable Janus droplets for submicron particle synthesis using an electrospray microfluidic chip. RSC Adv. 6 (15): 12042–12047. Li, S.S., Yu, X.L., You, S.J. et al. (2014). Generation of BiFeO3 -Fe3 O4 Janus particles based on droplet microfluidic method. Appl. Phys. Lett. 105 (4): 042903. Nisisako, T. (2016). Recent advances in microfluidic production of Janus droplets and particles. Curr. Opin. Colloid Interface Sci. 25: 1–12. Teh, S.Y., Lin, R., Hung, L.H., and Lee, A.P. (2008). Droplet microfluidics. Lab Chip 8 (2): 198–220. Serra, C.A. and Chang, Z.Q. (2008). Microfluidic-assisted synthesis of polymer particles. Chem. Eng. Technol. 31 (8): 1099–1115. Haeberle, S., Naegele, L., Burger, R. et al. (2008). Alginate bead fabrication and encapsulation of living cells under centrifugally induced artificial gravity conditions. J. Microencapsulation 25 (4): 267–274. Choi, C.H., Weitz, D.A., and Lee, C.S. (2013). One step formation of controllable complex emulsions: from functional particles to simultaneous encapsulation of hydrophilic and hydrophobic agents into desired position. Adv. Mater. 25 (18): 2536–2541. Haase, M.F. and Brujic, J. (2014). Tailoring of high-order multiple emulsions by the liquid-liquid phase separation of ternary mixtures. Angew. Chem. Int. Ed. 53 (44): 11793–11797. Choi, C.H., Kim, J., Nam, J.O. et al. (2014). Microfluidic design of complex emulsions. ChemPhysChem 15 (1): 21–29.

References

76 Deng, N.N., Wang, W., Ju, X.J. et al. (2013). Wetting-induced formation of

77

78

79

80

81

82

83

84

85

86

87 88

89

90

91

controllable monodisperse multiple emulsions in microfluidics. Lab Chip 13 (20): 4047–4052. Jeong, G.Y., Ricco, R., Liang, K. et al. (2015). Bioactive MIL-88A framework hollow spheres via interfacial reaction in-droplet microfluidics for enzyme and nanoparticle encapsulation. Chem. Mater. 27 (23): 7903–7909. Hennequin, Y., Pannacci, N., de Torres, C.P. et al. (2009). Synthesizing microcapsules with controlled geometrical and mechanical properties with microfluidic double emulsion technology. Langmuir 25 (14): 7857–7861. Ren, P.W., Ju, X.J., Xie, R., and Chu, L.Y. (2010). Monodisperse alginate microcapsules with oil core generated from a microfluidic device. J. Colloid Interface Sci. 343 (1): 392–395. Montazeri, L., Bonakdar, S., Taghipour, M. et al. (2016). Modification of PDMS to fabricate PLGA microparticles by a double emulsion method in a single microfluidic device. Lab Chip 16 (14): 2596–2600. Amstad, E., Kim, S.H., and Weitz, D.A. (2012). Photo- and thermo-responsive polymersomes for triggered release. Angew. Chem. Int. Ed. 51 (50): 12499–12503. Mou, C.L., Ju, X.J., Zhang, L. et al. (2014). Monodisperse and fast-responsive poly(N-isopropylacrylamide) microgels with open-celled porous structure. Langmuir 30 (5): 1455–1464. Costantini, M., Guzowski, J., Zuk, P.J. et al. (2018). Electric field assisted microfluidic platform for generation of tailorable porous microbeads as cell carriers for tissue engineering. Adv. Funct. Mater. 28 (32): 1870223. Wang, J., Shang, L., Cheng, Y. et al. (2015). Microfluidic generation of porous particles encapsulating spongy graphene for oil absorption. Small 11 (32): 3890–3895. Maeda, K., Onoe, H., Takinoue, M., and Takeuchi, S. (2012). Controlled synthesis of 3D multicompartmental particles with centrifuge-based microdroplet formation from a multi-barreled capillary. Adv. Mater. 24 (10): 1340–1346. Kim, S.H., Shum, H.C., Kim, J.W. et al. (2011). Multiple polymersomes for programmed release of multiple components. J. Am. Chem. Soc. 133 (38): 15165–15171. Zhou, M.M., Shen, L., Lin, X. et al. (2017). Design and pharmaceutical applications of porous particles. RSC Adv. 7 (63): 39490–39501. Wan, J., Bick, A., Sullivan, M., and Stone, H.A. (2008). Controllable microfluidic production of microbubbles in water-in-oil emulsions and the formation of porous microparticles. Adv. Mater. 20 (17): 3314. Geng, Y., Dalhaimer, P., Cai, S. et al. (2007). Shape effects of filaments versus spherical particles in flow and drug delivery. Nat. Nanotechnol. 2 (4): 249–255. Fish, M.B., Thompson, A.J., Fromen, C.A., and Eniola-Adefeso, O. (2015). Emergence and utility of nonspherical particles in biomedicine. Ind. Eng. Chem. Res. 54 (16): 4043–4059. Peyer, K.E., Zhang, L., and Nelson, B.J. (2013). Bio-inspired magnetic swimming microrobots for biomedical applications. Nanoscale 5 (4): 1259–1272.

41

42

1 Micro/Nanostructured Materials from Droplet Microfluidics

92 Xu, S., Nie, Z., Seo, M. et al. (2005). Generation of monodisperse particles

93

94

95

96

97

98

99

100 101

102

103

104

105

106

107

by using microfluidics: control over size, shape, and composition. Angew. Chem. Int. Ed. 44 (5): 724–728. Wang, W., He, X.H., Zhang, M.J. et al. (2017). Controllable microfluidic fabrication of microstructured materials from nonspherical particles to helices. Macromol. Rapid Commun. 38 (23): 1700429. Lin, Y.S., Yang, C.H., Hsu, Y.Y., and Hsieh, C.L. (2013). Microfluidic synthesis of tail-shaped alginate microparticles using slow sedimentation. Electrophoresis 34 (3): 425–431. Nisisako, T. and Torii, T. (2007). Formation of biphasic Janus droplets in a microfabricated channel for the synthesis of shape-controlled polymer microparticles. Adv. Mater. 19 (11): 1489–1493. Luo, Z., Xu, Y., Ye, E. et al. (2018). Recent progress in macromolecule-anchored hybrid gold nanomaterials for biomedical applications. Macromol. Rapid Commun. https://doi.org/10.1002/marc.201800029. Lin, Y.J., Gao, Y., Fang, F., and Fan, Z.Y. (2018). Recent progress on printable power supply devices and systems with nanomaterials. Nano Res. 11 (6): 3065–3087. Cui, L., Li, C.C., Tang, B., and Zhang, C.Y. (2018). Advances in the integration of quantum dots with various nanomaterials for biomedical and environmental applications. Analyst 143 (11): 2469–2478. Qi, G.B., Gao, Y.J., Wang, L., and Wang, H. (2018). Self-assembled peptide-based nanomaterials for biomedical imaging and therapy. Adv. Mater. 30 (22): e1703444. Li, L.L., Li, X.D., and Wang, H. (2017). Microfluidic synthesis of nanomaterials for biomedical applications. Small Methods 1 (8): 1700140. Macchione, M.A., Biglione, C., and Strumia, M. (2018). Design, synthesis and architectures of hybrid nanomaterials for therapy and diagnosis applications. Polymers 10 (5): 527. Ma, J., Lee, S.M., Yi, C., and Li, C.W. (2017). Controllable synthesis of functional nanoparticles by microfluidic platforms for biomedical applications - a review. Lab Chip 17 (2): 209–226. Jung, J.H., Park, T.J., Lee, S.Y., and Seo, T.S. (2012). Homogeneous biogenic paramagnetic nanoparticle synthesis based on a microfluidic droplet generator. Angew. Chem. Int. Ed. 51 (23): 5634–5637. Zhang, L., Feng, Q., Wang, J. et al. (2015). Microfluidic synthesis of hybrid nanoparticles with controlled lipid layers: understanding flexibility-regulated cell-nanoparticle interaction. ACS Nano 9 (10): 9912–9921. Ran, R., Wang, H., Liu, Y. et al. (2018). Microfluidic self-assembly of a combinatorial library of single- and dual-ligand liposomes for in vitro and in vivo tumor targeting. Eur. J. Pharm. Biopharm. 130: 1–10. Ortiz de Solorzano, I., Prieto, M., Mendoza, G. et al. (2016). Microfluidic synthesis and biological evaluation of photothermal biodegradable copper sulfide nanoparticles. ACS Appl. Mater. Interfaces 8 (33): 21545–21554. Phillips, T.W., Lignos, I.G., Maceiczyk, R.M. et al. (2014). Nanocrystal synthesis in microfluidic reactors: where next? Lab Chip 14 (17): 3172–3180.

References

108 Nightingale, A.M. and Demello, J.C. (2013). Segmented flow reactors for

nanocrystal synthesis. Adv. Mater. 25 (13): 1813–1821. 109 McMullen, J.P. and Jensen, K.F. (2010). Integrated microreactors for reac-

110

111

112

113 114

115

116

117

118

119

120

121

122

123

tion automation: new approaches to reaction development. Annu. Rev. Anal. Chem. (Palo Alto Calif ) 3: 19–42. Lazarus, L.L., Riche, C.T., Marin, B.C. et al. (2012). Two-phase microfluidic droplet flows of ionic liquids for the synthesis of gold and silver nanoparticles. ACS Appl. Mater. Interfaces 4 (6): 3077–3083. Shestopalov, I., Tice, J.D., and Ismagilov, R.F. (2004). Multi-step synthesis of nanoparticles performed on millisecond time scale in a microfluidic droplet-based system. Lab Chip 4 (4): 316–321. Gu, T., Zheng, C., He, F. et al. (2018). Electrically controlled mass transport into microfluidic droplets from nanodroplet carriers with application in controlled nanoparticle flow synthesis. Lab Chip 18 (9): 1330–1340. Song, H., Chen, D.L., and Ismagilov, R.F. (2006). Reactions in droplets in microfluidic channels. Angew. Chem. Int. Ed. 45 (44): 7336–7356. Theberge, A.B., Courtois, F., Schaerli, Y. et al. (2010). Microdroplets in microfluidics: an evolving platform for discoveries in chemistry and biology. Angew. Chem. Int. Ed. 49 (34): 5846–5868. Rhee, M., Valencia, P.M., Rodriguez, M.I. et al. (2011). Synthesis of size-tunable polymeric nanoparticles enabled by 3D hydrodynamic flow focusing in single-layer microchannels. Adv. Mater. 23 (12): 79–83. Alam, M.P., Jagodzinska, B., Campagna, J. et al. (2016). C—O bond formation in a microfluidic reactor: high yield SN Ar substitution of heteroaryl chlorides. Tetrahedron Lett. 57 (19): 2059–2062. Capretto, L., Carugo, D., Mazzitelli, S. et al. (2013). Microfluidic and lab-on-a-chip preparation routes for organic nanoparticles and vesicular systems for nanomedicine applications. Adv. Drug Deliv. Rev. 65: 1496–1532. Su, Y.F., Kim, H., Kovenklioglu, S., and Lee, W.Y. (2007). Continuous nanoparticle production by microfluldic-based emulsion, mixing and crystallization. J. Solid State Chem. 180 (9): 2625–2629. Hung, L.H., Teh, S.Y., Jester, J., and Lee, A.P. (2010). PLGA micro/nanosphere synthesis by droplet microfluidic solvent evaporation and extraction approaches. Lab Chip 10 (14): 1820–1825. Faustini, M., Kim, J., Jeong, G.Y. et al. (2013). Microfluidic approach toward continuous and ultrafast synthesis of metal–organic framework crystals and hetero structures in confined microdroplets. J. Am. Chem. Soc. 135 (39): 14619–14626. Wang, C.W., Oskooei, A., Sinton, D., and Moffitt, M.G. (2010). Controlled self-assembly of quantum dot-block copolymer colloids in multiphase microfluidic reactors. Langmuir 26 (2): 716–723. Liu, B.T., Vellingiri, K., Jo, S.H. et al. (2018). Recent advances in controlled modification of the size and morphology of metal–organic frameworks. Nano Res. 11 (9): 4441–4467. Lu, K., Aung, T., Guo, N. et al. (2018). Nanoscale metal–organic frameworks for therapeutic, imaging, and sensing applications. Adv. Mater. https://doi .org/10.1002/adma.201707634.

43

44

1 Micro/Nanostructured Materials from Droplet Microfluidics

124 Liu, D., Zhang, H., Fontana, F. et al. (2017). Microfluidic-assisted fabrication

of carriers for controlled drug delivery. Lab Chip 17 (11): 1856–1883. 125 Hasani-Sadrabadi, M.M., Taranejoo, S., Dashtimoghadam, E. et al. (2016).

126

127 128

129

130

131

132

133

134

135

136 137

138

139

Microfluidic manipulation of core/shell nanoparticles for oral delivery of chemotherapeutics: a new treatment approach for colorectal cancer. Adv. Mater. 28 (21): 4134–4141. Zhang, H., Liu, D., Shahbazi, M.A. et al. (2014). Fabrication of a multifunctional nano-in-micro drug delivery platform by microfluidic templated encapsulation of porous silicon in polymer matrix. Adv. Mater. 26 (26): 4497–4503. Sanjay, S.T., Zhou, W., Dou, M. et al. (2018). Recent advances of controlled drug delivery using microfluidic platforms. Adv. Drug Deliv. Rev. 128: 3–28. Hasani-Sadrabadi, M.M., Karimkhani, V., Majedi, F.S. et al. (2014). Microfluidic-assisted self-assembly of complex dendritic polyethylene drug delivery nanocapsules. Adv. Mater. 26 (19): 3118–3123. Xu, Q., Hashimoto, M., Dang, T.T. et al. (2009). Preparation of monodisperse biodegradable polymer microparticles using a microfluidic flow-focusing device for controlled drug delivery. Small 5 (13): 1575–1581. Li, Y.N., Yan, D., Fu, F.F. et al. (2017). Composite core-shell microparticles from microfluidics for synergistic drug delivery. Sci. China Mater. 60 (6): 543–553. Wei, J., Ju, X.J., Zou, X.Y. et al. (2014). Multi-stimuli-responsive microcapsules for adjustable controlled-release. Adv. Funct. Mater. 24 (22): 3312–3323. Molinaro, R., Evangelopoulos, M., Hoffman, J.R. et al. (2018). Design and development of biomimetic nanovesicles using a microfluidic approach. Adv. Mater. 30 (15): e1702749. Feng, Q., Zhang, L., Liu, C. et al. (2015). Microfluidic based high throughput synthesis of lipid-polymer hybrid nanoparticles with tunable diameters. Biomicrofluidics 9 (5): 052604. Fontana, F., Shahbazi, M.A., Liu, D. et al. (2017). Multistaged nanovaccines based on porous silicon@acetalated dextran@cancer cell membrane for cancer immunotherapy. Adv. Mater. 29 (7): 1603239. Wieduwild, R., Krishnan, S., Chwalek, K. et al. (2015). Noncovalent hydrogel beads as microcarriers for cell culture. Angew. Chem. Int. Ed. 54 (13): 3962–3966. Allazetta, S., Kolb, L., Zerbib, S. et al. (2015). Cell-instructive microgels with tailor-made physicochemical properties. Small 11 (42): 5647–5656. Kapourani, E., Neumann, F., Achazi, K. et al. (2018). Droplet-based microfluidic templating of polyglycerol-based microgels for the encapsulation of cells: a comparative study. Macromol. Biosci. 18 (10): 1800116. Zhou, X.C., Haraldsson, T., Nania, S. et al. (2018). Human cell encapsulation in gel microbeads with cosynthesized concentric nanoporous solid shells. Adv. Funct. Mater. 28 (21): 1707129. Agarwal, P., Zhao, S., Bielecki, P. et al. (2013). One-step microfluidic generation of pre-hatching embryo-like core-shell microcapsules for miniaturized 3D culture of pluripotent stem cells. Lab Chip 13 (23): 4525–4533.

References

140 Lutolf, M.P. and Hubbell, J.A. (2005). Synthetic biomaterials as instructive

141 142 143 144

145

146

147

148 149

150

151

152

153 154 155

extracellular microenvironments for morphogenesis in tissue engineering. Nat. Biotechnol. 23 (1): 47–55. Place, E.S., Evans, N.D., and Stevens, M.M. (2009). Complexity in biomaterials for tissue engineering. Nat. Mater. 8 (6): 457–470. Salgado, A.J., Coutinho, O.P., and Reis, R.L. (2004). Bone tissue engineering: state of the art and future trends. Macromol. Biosci. 4 (8): 743–765. Ma, P.X. (2008). Biomimetic materials for tissue engineering. Adv. Drug Deliv. Rev. 60 (2): 184–198. Hou, Y., Xie, W., Achazi, K. et al. (2018). Injectable degradable PVA microgels prepared by microfluidic technology for controlled osteogenic differentiation of mesenchymal stem cells. Acta Biomater. https://doi.org/10 .1016/j.actbio.2018.07.003. Vardar, E., Larsson, H.M., Allazetta, S. et al. (2018). Microfluidic production of bioactive fibrin micro-beads embedded in crosslinked collagen used as an injectable bulking agent for urinary incontinence treatment. Acta Biomater. 67: 156–166. Yu, J., Huang, T.R., Lim, Z.H. et al. (2016). Production of hollow bacterial cellulose microspheres using microfluidics to form an injectable porous scaffold for wound healing. Adv. Healthc. Mater. 5 (23): 2983–2992. Liu, H.T., Wang, H., Wei, W.B. et al. (2018). A microfluidic strategy for controllable generation of water-in-water droplets as biocompatible microcarriers. Small https://doi.org/10.1002/smll.201801095. Liu, Q. and Boyd, B.J. (2013). Liposomes in biosensors. Analyst 138 (2): 391–409. Kirsch, J., Siltanen, C., Zhou, Q. et al. (2013). Biosensor technology: recent advances in threat agent detection and medicine. Chem. Soc. Rev. 42 (22): 8733–8768. Visaveliya, N. and Köhler, J.M. (2015). Microfluidic assisted synthesis of multipurpose polymer nanoassembly particles for fluorescence, LSPR, and SERS activities. Small 11 (48): 6435–6443. Lohse, S.E., Eller, J.R., Sivapalan, S.T. et al. (2013). A simple millifluidic benchtop reactor system for the high-throughput synthesis and functionalization of gold nanoparticles with different sizes and shapes. ACS Nano 7 (5): 4135–4150. Knauer, A., Thete, A., Li, S. et al. (2011). Au/Ag/Au double shell nanoparticles with narrow size distribution obtained by continuous micro segmented flow synthesis. Chem. Eng. J. 166 (3): 1164–1169. Kim, S.H., Lee, T.Y., and Lee, S.S. (2014). Osmocapsules for direct measurement of osmotic strength. Small 10 (6): 1155–1162. Zhao, Y., Cheng, Y., Shang, L. et al. (2015). Microfluidic synthesis of barcode particles for multiplex assays. Small 11 (2): 151–174. Gerver, R.E., Gómez-Sjöberg, R., Baxter, B.C. et al. (2012). Programmable microfluidic synthesis of spectrally encoded microspheres. Lab Chip 12 (22): 4716–4723.

45

46

1 Micro/Nanostructured Materials from Droplet Microfluidics

156 Zhao, Y., Shum, H.C., Chen, H. et al. (2011). Microfluidic generation of

157 158

159 160

161

multifunctional quantum dot barcode particles. J. Am. Chem. Soc. 133 (23): 8790–8793. Shang, L., Fu, F., Cheng, Y. et al. (2015). Photonic crystal microbubbles as suspension barcodes. J. Am. Chem. Soc. 137 (49): 15533–15539. Medintz, I.L., Uyeda, H.T., Goldman, E.R., and Mattoussi, H. (2005). Quantum dot bioconjugates for imaging, labelling and sensing. Nat. Mater. 4 (6): 435–446. Pitruzzello, G. and Krauss, T.F. (2018). Photonic crystal resonances for sensing and imaging. J. Opt. 20 (7): 073004. Liu, D., Cito, S., Zhang, Y. et al. (2015). A versatile and robust microfluidic platform toward high throughput synthesis of homogeneous nanoparticles with tunable properties. Adv. Mater. 27 (14): 2298–2304. Huang, Y.C., Han, T.T., Xuan, J. et al. (2018). Design criteria and applications of multi-channel parallel microfluidic module. J. Micromech. Microeng. 28 (10): 105021.

47

2 Digital Microfluidics for Bioanalysis Qingyu Ruan, Jingjing Guo, Yang Wang, Fenxiang Zou, Xiaoye Lin, Wei Wang, and Chaoyong Yang Xiamen University, College of Chemistry & Chemical Engineering, Department of Chemistry, Lujiaxi Bldg, Room 532, No. 422, Siming South Road, Xiamen 361005, China

2.1 Introduction In recent decades, the concepts of micro-total-analysis system (μTAS) [1] and lab-on-a-chip (LOC) have been proposed. Microfluidics [2–5], a synergy of science and engineering, has greatly promoted the development of analytical chemistry and beyond. As a fluid-handling technology based on microscale effect, microfluidics integrates complex laboratory functions in a miniaturized single micro- or nanoanalytical device, making its applications manifold [6, 7]. The main body of the conventional microfluidic chip is typically enclosed in three-dimensional microchannels, which are coupled with pressure-controlled valves and pumps to realize the manipulations of the fluid flow. In recent years, microfluidics has gained widespread applications in chemistry, biology, medicine, etc. Compared with conventional techniques, microfluidics offers great advantages including miniaturization, low sample volume, good efficiency, and high throughput. However, such a valve-based chip with complicated structures often leads to difficulties in not only chip fabrication but also manipulation by operators. The valve-based system also requests external components, which limit the high integration. Moreover, it is easy to generate “dead volume” of the microchannel in the process of injection as well as the adherence of biofouling, causing cross-contamination easily. Therefore, with the growing need for better performance of the microfluidic system, an improved method of droplet manipulation is highly desired. Digital microfluidics (DMF) [8–10] is a new concept in the field of microfluidics. Broadly speaking, it is a general term for a series of new droplet manipulation technology, including electrowetting-on-dielectric (EWOD) [11, 12], dielectrophoresis (DEP) [13, 14], surface acoustic wave [15, 16], magnetic force [17, 18], thermocapillary force [19, 20], and optoelectrowetting [21]. In the narrow sense, it refers to EWOD-based techniques, which will be elaborated in this chapter. Nanotechnology and Microfluidics, First Edition. Edited by Xingyu Jiang. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.

48

2 Digital Microfluidics for Bioanalysis

DMF is based on an array of electrodes with a hydrophobic surface for precise manipulation of discrete droplets in picoliter to microliter sizes [22]. On the electrodes to which the potentials are applied, discrete droplets are capable of moving, merging, splitting, and dispensing from reservoirs individually [23] along different directions on the chip surface. Compared to the conventional microfluidics based on microchannels, DMF eliminates the need for valves, pumps, or mechanical mixers, conferring the benefits of chip design, fabrication, manipulation, and device integration while maintaining low sample and reagents consumption, high reaction rates, high parallelization, and automation [24–28]. Furthermore, DMF has great advantages over microchannel-based microfluidics especially in universality, reconfigurability, and flexibility, and embodied droplet path in one electrode chip can be reconfigured for different applications depending on the actuated programming. As a versatile platform, DMF is easy to integrate with other devices, including but not limited to mass spectrometry (MS) [29], surface-enhanced Raman scattering (SERS) [30], nuclear magnetic resonance (NMR) [31, 32], optical techniques [33, 34], and electrical techniques [28]. With its unique and powerful features well-suited for chemical [32, 35, 36] and biological [37–40] microanalysis, DMF has a remarkably wide range of applications including chemical reactions [41], cell-based [42–45] and DNA-based applications [46–48], immunoassays [30, 49, 50], enzyme assays [51], proteomics [52, 53], and clinical diagnostics [54, 55]. In this chapter, we outline the status quo of DMF from the aspects of theoretical background, chip fabrication, device integration, and biological applications, followed by a discussion of its underlying opportunities as well as existing challenges in analytical chemistry and concluded with a look forward.

2.2 Theoretical Background The driving force in EWOD-based DMF is electrostatic forces. The change in contact angle results from the electrostatic forces exerted on the droplet, which can be used for droplet actuation. Microfluidic manipulation of liquids by electrowetting (or electrocapillarity) was accomplished first with mercury droplets in water by Lippmann in 1875 [56]. He found that the rise of mercury in the capillary tube could be varied under the condition of contact with electrolyte when electric charges exist. Since then, researchers focused on the research that aqueous electrolytes are directly contiguous with mercury surfaces. However, when applying voltages over several hundred millivolts, electrolysis of electrolytes became a key obstacle to broad applications. It was not until the 1990s that Berge used an insulating layer to isolate the conductive liquid from the electrode to prevent the electrolysis [57, 58] and thus broaden the window of operations. This concept, electrowetting on this dielectric-coated surface, has been regarded as EWOD. 2.2.1

Theoretical Background

Firstly, we define the contact angle and make the relation of contact angle and surface tension on the triple-phase line clear. When the liquid is partially wetting

2.2 Theoretical Background

Figure 2.1 Schematics of a liquid partially wetting on a solid surface.

γLG

Gas

Liquid γSL

θ

γSG

Solid

on the surface of the solid substrate, the contact angle in a liquid–vapor interface is called the contact angle, as shown in Figure 2.1. Thomas Young suggested that the contact angle is determined by the mechanical equilibrium of the three surface tensions of the droplet, which sat on a plane surface: γSG − γSL = γLG cos 𝜃

(2.1)

where 𝛾 SG , 𝛾 SL , and 𝛾 LG at the interface of the solid and gas, the solid and liquid, the liquid and gas, respectively. In electrowetting, the interfacial force is greatly strong nearby the liquid–gas contact line under the tip effect. Lippmann’s law, a formula, presented the theory of electrocapillarity (also electrowetting): 𝜀𝜀 𝛾SL = 𝛾SL0 − 0 d V 2 (2.2) 2d where V represents the electric potential externally, and 𝛾 SL0 is the 𝛾 SL when V = 0. Next, Berge showed the Lippmann–Young’s law by combining Lippmann’s law with Young’s law for explaining the EWOD: cos 𝜃 = cos 𝜃0 +

C V2 2𝛾LG

(2.3)

Here, c is the capacitance of the dielectric layer, which prevented electrodes to come in direct contact with the liquid. Although the EWOD seems easy to understand as depicted above, we have to realize that the detailed physics of it is unclear until now. The deduction of Eq. (2.3) is multipartite and the origins of some related phenomena are still controversial. Next, we showed the derivation of Lippmann–Young’s law, which could be explained by various approaches [59]: the thermodynamic approach, the energy minimization approach, and the electromechanical approach. 2.2.1.1

Thermodynamic Approach

Lippmann’s law based on classical thermodynamics [60] was developed from a conductive liquid that comes in contact with a perfectly smooth solid surface. When applying an elementary electric field at the solid–liquid interface, there is an electric double layer (EDL) formation in the side of liquid contacting the substrate. The relationship between the normal force density and a gradient of the interfacial tension based on Gibbs’ interfacial thermodynamics yields eff = −𝜌SL dV d𝛾SL

(2.4)

Here, 𝛾 eff is the effective surface tension at the interface of liquid and solid, 𝜌SL is the field-induced surface charge density of the counterions, and V is the electric potential. Assuming that all of counterions are placed at the fixed gap dH against

49

50

2 Digital Microfluidics for Bioanalysis

the surface (the Helmholtz simplifying assumption), the EDL should have a certain capacitance per unit area answered to the charging under the normal electric field: cH = 𝜀0 𝜀l /dH (𝜀1 is the dielectric constant of the liquid). We find: V eff 𝛾SL (V ) = 𝛾SL −

∫Vpzc

V

𝜌SL dV = 𝛾SL −

∫Vpzc

cH V dV = 𝛾SL −

CH (V − Vpzc )2 2 (2.5)

where V pzc is the potential at no charge (note that the voltage needed to make up for this idiopathic charging as mercury surfaces receive an idiopathic charge immersed into electrolytes at 0 V). Inserting Eq. (2.5) into Young’s equation (2.1), we obtain: 𝜀0 𝜀1 (U − Upzc )2 (2.6) cos 𝜃 = cos 𝜃Y + 2dH 𝜎1V From Eq. (2.6), we related the contact angle to the applied voltage regarding that an electrolyte droplet is placed directly at an electrode surface. However, the equation is only applicable in the range of a little voltage because of hydrolysis phenomena. Most applications of electrowetting overcome it by introducing a dielectric layer (the insulator thickness d is greater compared with dH ) as mentioned above. In this EWOD configuration, EDL forms at the interface of insulator and droplet [61, 62]. The system contains two serial capacitors, the one between the solid and insulator interface (capacitance cH ) and the other is the dielectric layer (cd = 𝜀0 𝜀d /d), where 𝜀d is the dielectric constant of the insulator. The total capacitance per unit area c ≈ cd by omitting cH (cd ≪ cH ). Based on this result, the electric field penetrating into the liquid is so small such that it is considered as an ideal conductor. Equation (2.5) is replaced by 𝜀𝜀 eff 𝛾SL (V ) = 𝛾SL − 0 d V 2 (2.7) 2d where V pzc has been overlooked, assuming that the insulating layer has not caused charge adsorption spontaneously. By introducing Young’s law, we get Lippmann–Young’s law for EWOD: cos 𝜃 = cos 𝜃0 +

C V2 2𝛾LG

(2.8)

Compared to Eq. (2.6), we find that the voltage needed to realize the same decrease of contact angle in EWOD is much higher from Eq. (2.8) because of the introduction of the insulating layer. In addition, in many other experiments, Eq. (2.8) is found to deviate the theoretical curve when the voltage is over a certain threshold. This phenomenon, alleged contact angle saturation, will be discussed individually in Section 2.2.2. 2.2.1.2

Energy Minimization Approach

For EWOD, Berge derived the Eq. (2.8) first grounded on energy minimization [57] instead of thermodynamics. The energy of the droplet can be expressed as E = E(R; 𝜃; p) mathematically. Here, R represents the droplet radius, 𝜃 is the contact angle, and p is a net of other parameters. Herein, there are two premises that

2.2 Theoretical Background

must be made clear: (i) the system achieves a balance only in the case of the energy E(R; 𝜃; p) at a minimum and (ii) R and 𝜃 are dependent on each other as shown below. 𝜋 (2.9) V (R, 𝜃) = R3 (2 − 3 cos 𝜃 + cos3 𝜃) 3 Thus, substituting dE = 0 and dV = 0, we find: ( ) ( ) ⎡ 𝜃 ⎤ 2 𝜃 cot 2cos ⎢ 2 2 ⎥ 𝜕E 𝜕E dE = −R ⎢ (2.10) ⎥ 𝜕R + 𝜕𝜃 = 0 d𝜃 2 + cos 𝜃 ⎢ ⎥ ⎣ ⎦ Equation (2.10) is the basis for the Lippmann–Young law. Next, considering that the liquid is regarded as a perfect conductor [63], the electric energy is provided in the electrostatic contribution by − → − 1→ dEdiel = ( D ⋅ E )d𝜈 (2.11) 2 → − where D is the electric displacement vector, which stands for the induced dipole moment in the dielectric per unit volume, and dv is an elementary volume of the → − → − dielectric. For a perfect dielectric, this moment D = 𝜀0 𝜀D E , we obtain: −→2 1 dEdiel = 𝜀0 𝜀D ∣ E ∣ dv (2.12) 2 → − When neglecting the edge effect, the electrical field should be uniform: E = {0, 0, −V ∕d}. We insert it into Eq. (2.12): ( )2 𝜀 𝜀 V2 1 V ASL d = 0 D 𝜋R2 sin2 𝜃 (2.13) Ediel (R, 𝜃) = 𝜀0 𝜀D 2 d 2d On the other hand, the total electric energy of the system is equal to the electric energy of the dielectric minus the energy stored in the external charging source. The energy of the external charging source is negative twice the energy stored in the dielectric, which can be shown by the law of conservation of energy. Hence, the total energy of the system, surface plus electric energy, is ] [( ) 𝜀 𝜀 V2 (2.14) 𝜋 sin2 𝜃 + 𝛾LG 2𝜋(1 − cos 𝜃) E = R2 𝛾SL − 𝛾SG − 0 D 2d As the interfacial energy can be expressed: Eint = 𝜋R2 [2𝛾 LG (1 − cos 𝜃) + (𝛾 SL − 𝛾 SG )sin2 𝜃]. Inserting Eq. (2.14) into Eq. (2.10) and substituting the term 𝜀 V2 (𝛾 SL − 𝛾 SG ) by (𝛾SL − 𝛾SG − D2d ), we get Lippmann–Young’s equation: cos 𝜃 = cos 𝜃0 +

𝜀0 𝜀D 2 V 2𝛾LG

(2.15)

However, this energy minimization approach also does not explain how the contact angle varies in terms of mechanics, which is paramount in the research of electrowetting dynamics. Besides, the approach becomes unuseful in the case of viscous liquids. Next, we present an analytical approach of the contact angle reduction in terms of mechanics because the actual force exerted on the liquid must be seized for the EWOD phenomenon fully understood.

51

52

2 Digital Microfluidics for Bioanalysis

F⃗

+ + +

+

Liquid (perfect conductor)

2.2.1.3

E⃗

n⃗

+

Figure 2.2 Electric force exerting at the interface of liquid as a perfect conductor.

+ +

Electromechanical Approach

This approach was first applied in electrowetting by Jones et al. [64, 65] These forces that affect the droplet movement by the electric field result from the polarization density and the free electric charge density (𝜌f ). For the model of simple liquids, the Korteweg–Helmholtz body force density [ ] 𝜕𝜀f 𝜀 → − → − 𝜀 fk = 𝜎f E − 0 E2 ∇𝜀f + ∇ 0 E2 𝜌 (2.16) 2 2 𝜕𝜌 Here, 𝜎 f is the electric charge, 𝜀f is the permittivity of the liquid, and 𝜌 stands for the mass density of the liquid. We omit the last term in Eq. (2.16) in the present context because it is corresponding to electrostriction. The Maxwell stress tensor is obtained by integrating Eq. (2.16): ) ( 1 (2.17) Tik = 𝜀0 𝜀 Ei Ek − 𝛿ik E2 2 → − Here, E2 is equivalent to ∣ E ∣2 and 𝛿 ik describes the Kronecker delta function: 𝛿 ik = 0 if i ≠ k and 𝛿 ii = 1; and i, k = x, y, z. The net force of the liquid volume element dA is Fi =

∮Ω

Tik nk dA

(2.18)

From Eq. (2.18), we observe that the force per unit surface area (dA) toward the − extrorse surface normal → n (Figure 2.2) is the only nonzero dedication: → − 𝜀 − 𝜌→ − F − n = 0 E2→ n = sE = Pe→ (2.19) 𝛿A 2 2 Here, Pe = 𝜀0 E2 /2 is the electrostatic pressure exerted on the droplet surface. Kang [66], Vallet et al. [67] used the Schwarz–Christoffel conformal mapping [68] to resolve the surface distortion. Obviously, the electric potential 𝜑 is a harmonic function because it satisfies Laplace’s equation ∇2 ∅ = 0. A conformal mapping that converts the functions E and 𝜑 to the moiety of the plane into the same functions for a wedge occurs there, which suggests that the field and charges gather at the tip of the wedge. In practice, the electric field is singular at a wedge under the condition that any harmonic function near a geometry is singular, but the Maxwell pressure is integrable and causes a finite Maxwell force at the contact line after Eq. (2.19) is integrated: Fhorizontal =

𝜀0 𝜀D V 2 2d

(2.20)

2.2 Theoretical Background

Figure 2.3 The net electric forces on the liquid–gas interface.

V

γLG Fvertical

Electrolyte(aq.) γ

EDL (1–10 nm)

θ

– – – – – – – – SL + + + ++++++ + + +

Electrode

Fhorizontal γSG

It indicated that the Maxwell stress can just result in the electric forces on the liquid–gas interface within an extremely small extent near the triple line, which is shown at the macroscale in Figure 2.3. Making the mechanical equilibrium in the horizontal direction at the triple line by Eq. (2.20), we obtain Yang–Lippmann’s equation again: 𝜀D (2.21) cos 𝜃 = cos 𝜃0 + V2 2𝛾LG d The electromechanical approach presents that the Lippmann term (electrowetting number 𝜂) is derived from the Maxwell stress at the liquid–gas interface. However, its origin should be gone back to the thin layer of the solid medium and the wedge geometry of the liquid. Absolutely, the wedge shape of the gas phase results in distortion of the contact angle because the permittivity (and conductivity) of the liquid outclasses the gas, which can be obviously shown by the Schwarz–Christoffel transformation. This doctrine actually forecasts that if the contact angle reaches zero, the force would reach infinity, which accounts for the contact angle saturation, accompanying by ejection of nanodrops usually [69]. 2.2.2

Contact Angle Saturation

Although the Lippmann–Young equation has shown that the contact angle could reach to 0∘ by applying a high-enough voltage, there is good experimental evidence that the contact angle saturates at high voltage eventually. The saturation effect has been a bottleneck for the microminiaturization of EWOD devices [70, 71], which limits the variable range of electrowetting force. In spite of this, a series of theories has been proposed to describe the saturation effect, including gas ionization at the contact line [67], charge trapping in the dielectric layer [72], ejection of satellite microdroplets [73], and a zero solid–liquid interface tension limit at the contact line [74]. The various explanations presented about contact angle saturation are debated so far. 2.2.3

Basic Microfluidic Functions by EWOD Actuation

To achieve the continuous droplet operation, an array of electrodes is necessary to control the surface wettability electrode-by-electrode programmably. The configuration of a typical EWOD device is the sandwich structure composed of a

53

54

2 Digital Microfluidics for Bioanalysis

(A) Moving

ON

OFF

OFF

(a)

Splitting

OFF

ON

ON

OFF

OFF

(b) ON

OFF

(a)

Merging

ON

OFF

OFF

ON

ON

ON

OFF

OFF

ON

OFF

ON

ON

(d) ON

ON

(c)

(b)

(a)

OFF

(c)

(b) ON

ON

OFF

ON

(d) OFF

OFF

(c)

ON

OFF

(d)

OFF OFF

OFF

OFF

ON

OFF

(a)

Dispensing

ON

ON

ON

OFF

ON

(b)

ON

ON OFF

OFF

OFF

ON

OFF

(c)

OFF

(d)

(B)

Split Mix

Merge

Dispense

Figure 2.4 (A) Four fundamental droplet manipulations by EWOD actuation. (B) Basic physical microfluidic functions can be performed on the DMF chip.

liquid droplet and two parallel hydrophobic plates. Using the two-plate EWOD device, we can perform four fundamental microfluidic manipulations by EWOD actuation [23], including moving, merging, splitting, and dispensing as shown in Figure 2.4. Merging two droplets can be achieved easily by driving two droplets toward each other. Splitting is a more demanding process and requires a series of optimization in parameters. In brief, the height of gap should be low enough so that the shear force can cut the elongated droplet when driving voltage and electrode (droplet) size in a certain value. Finally, dispensing, a medium between micro and macrofluids, is a key operation for DMF circuits. However, dispensing from a reservoir has extra challenges and requires various techniques to be effective.

2.3 Device Fabrication

2.3 Device Fabrication In DMF, it does not require discrete pumps, valves, or moving micromachined components, acquiring the system very portable, universal, and energy-saving (no continuous current). Moreover, in this array-based microfluidic device, the microchannel that usually calls for attention to fabricating, bonding, and sealing issues is not required. Some fluidic manipulation can be carried out at the same time. Generally, DMF devices are implemented in two different configurations: a single plane (open format) and a two plane (closed format) (Figure 2.5a,b). The difference between the open format and the closed format is only the ground electrode. In the open format [13, 75–77], both of the ground electrode and the actuation electrodes are on the bottom plate, so it does not need a cover plate. In addition, in the closed format [78–83], the droplet is sandwiched between the top and bottom plates, which are separated by a spacer. The bottom plate is patterned with an array of electrodes, and the top plate is used as the ground electrode (Figure 2.5). The materials of substrates for DMF plates have a variety of choices, including glass, silicon, printed circuit board (PCB), and other flexible substrates. Use of silicon and glass, which have been widely used as substrates because of their stable chemical properties [84, 85], however, is limited because of the low fabrication throughput and high cost. In recent years, there is a current toward the use of PCB substrates because of the low cost and batch fabrication [51, 75, 83, 86]. Specially, disposable DMF devices are highly desirable, which have become a powerful tool for immunoassay [87] and disease diagnosis [88]. Paper-based DMF is the most common disposable DMF device [87, 89–92] because of the low cost and batch fabrication for replacement of the chip, which will decrease the cross-contamination. The electrodes on the paper-based substrates can be formed by screen printing or inject printing [91]. Besides, disposable DMF can be fabricated from flexible substrates [51], which can be driven on nonplanar surfaces, allowing for the integration of multiple physical and chemical environments on the same device. An electrode layer is covered on the substrates. The array and pattern of electrodes in the electrode layer are designed according to the experimental requirements. Generally, DMF electrodes are formed from metals (A)

Open DMF device

(B)

Closed DMF device Top plate

Bottom plate Substrate

Electrode

Dielectric layer

Hydrophobic layer

Figure 2.5 EWOD actuation configurations: (A) single plane (open format) and (B) two plane (closed format).

55

56

2 Digital Microfluidics for Bioanalysis

(e.g. chromium [30], aluminum, gold [93], and copper [94]) or other conductive materials such as indium tin oxide (ITO) [95]. The properties of the dielectric layer and hydrophobic layer are much more crucial to lower the driving voltage for the droplet wetting. The insulating dielectric layer is used to accumulate the charge or electric field gradient for droplet actuation. The dielectric constant of the material has a great influence on the droplet operation. For instance, when the dielectric constant is higher, the voltage required will be lower. Further, to better drive the droplet and not breakdown the dielectric, the uniformity and fineness of the dielectric layer should be optimized. The stable dielectric layer can increase the use of chip. Materials such as polytetrafluoroethylene (PTFE) [67], SiO2 [96], parylene-C [62, 72, 74, 97–100], conventional Teflon films [67, 100, 101], polydimethylsiloxane (PDMS) [102, 103], and some other commercial polymer foils of variable surface qualities can be used [67, 102, 104, 105]. It can be formed by a variety of techniques, such as spin-coating, thermal growth, and vapor deposition. Moreover, the hydrophobic layer, which allows for droplet transport on the chip, is typically made of amorphous fluoropolymer materials such as Teflon [106] and CYTOP [107]. We can use spin-coating or dip-coating to form a layer of varying thickness according to the experimental requirements [108]. In brief, the insulating layers should fulfill the two main criteria, thinner dielectric layer and larger contact angle, so that the chip can achieve better performance.

2.4 Digital Microfluidics Integrated with Other Devices Biological samples often have complex contents and require complicated processing before analysis. Therefore, demand rises for automation and integration in the field of microfluidics. Because of their good compatibility, portability, and automated processing, DMF integrated with other methods has been recently developed for sample processing and signal readout rapidly. The combination of various conventional methods and DMF greatly reduces human labor and expands the scope for portable applications. 2.4.1 2.4.1.1

Sample Processing Systems Integrated with Digital Microfluidics World-to-chip Interface

World-to-chip interface is essential for DMF because the volumes of samples from the real world are typically larger than those that can be handled on a DMF device. Therefore, the concentration modules integrated with DMF have been developed in recent years. Here, we introduced five world-to-chip interfaces including dried blood spot (DBS), solid-phase microextraction (SPME), world-to-DMF interface, preconcentration by liquid intake by paper (P-CLIP), and tissue–liquid extraction. DBS samples are emerging rapidly as an advantageous sampling and storage carrier for broad clinical applications. The long and tedious preprocessing of DBS extraction has been automatized by DMF since 2011 [29], and it can be easily integrated with DMF devices by positioning on the chip directly without any

2.4 Digital Microfluidics Integrated with Other Devices

channel clogging problem. In addition, the filter paper punch of DBS remains stuck to the surface of the chip by friction and capillary forces, making various fluids manipulate on the filter paper possible. In a typical experiment for DBS analysis on DMF, a disk of 3.2 mm diameter is punched from a DBS sample on filter paper and then deposited on the device. The analytes of the DBS sample are then extracted, derivatized, and reconstituted by a series of reagent droplets, finally analyzed by MS [29, 38]. SPME uses a fiber that is coated with an extracting phase. The fiber enables extraction of different kinds of analytes for subsequent analyses but requires large elution volumes that add to the challenge of preconcentration. The integration of SPME into DMF is developed to achieve low elution volumes, allowing for preconcentration of analytes and compatibility with a wider range of analytical methods. For example, Wheeler’s group [25] demonstrated a quantification of pg/ml concentrations of free steroid hormones in urine by SPME–DMF–HPLC–MS method. In this method, the analytes in the sample solution are extracted into a SPME fiber first, then eluted into desorption solvent droplets on a DMF device, and finally analyzed by HPLC–MS. They also developed a custom DMF–HPLC–MS interface [109] for direct sample loading, which will be discussed in Section 2.4.2.3. Both the above methods are still based on initial off-line extraction, requiring more procedures out of the system. Recently, as shown in Figure 2.6, a novel world-to-DMF interface has been developed for online extraction of RNA from human blood [110], including two 500 μl microcentrifuge tubes, a transfer tube interfaced with DMF module, and a connector tube integrated with a peristaltic pump. This interface enables large-volume fluid preconcentration, enhances the extraction efficiency, and can be further integrated with more additional modules. Moreover, another world-to-DMF interface, P-CLIP, has been developed [54]. In P-CLIP, a virtual channel is created on the DMF device Extraction module Connector tube

Peristaltic pump

Connector tube Reaction chamber

Waste chamber

Transfer tube 1

Transfer tube 2

Bridging electrodes

Front view Extraction–purification modules interface Side view Transfer tubes

Top view

Actuation electrodes Top plate

Wash 2 solution

Top plate Air gap Bottom plate Magnet Glass Electrode Hydrophobic layer Dielectric layer

Transfer Tubes

1 mm

6 mm

Wash 1 solution Purification module

1 mm

5 mm

Exposed Unexposed Area of Bridging electrode

Figure 2.6 DMF devices integrated with a world-to-chip interface.

Elution buffer

57

58

2 Digital Microfluidics for Bioanalysis

by electrowetting forces and wicking forces generated from absorbent materials, which is useful for concentration of low-abundance analytes from large-volume samples. In addition to liquid extraction, tissue–liquid extraction integrated with DMF device has also been reported [111]. Driven by the need for personalized medicine, Wheeler and coworkers developed a technique integrating tissue–liquid extraction with DMF device for quantification of analytes from core needle biopsy samples. In this technique, a tissue homogenate was directly loaded on the device, then extracted and reconstituted by sequential droplets of reagents for immunoassay, and quantified by LC–MS/MS. 2.4.1.2

Magnet Separation

Magnet separation for magnetic particles is a common method for sample processing in bioassays. It is also essential for DMF devices to integrate magnet separation module for various applications. Pamula and coworkers [75] developed a DMF device with a permanent magnet embedded at appropriate locations underneath the chip. Considering the impact on the results by the immobilized magnet, they compared the chemiluminescence results of the incubation performed on the magnet or off the magnet. It turned out that the time to reach saturation of the former condition doubles the time of the latter condition. They hypothesized that this was because the recirculation patterns of the magnet affected the beads’ distribution. However, in most of the cases, the magnet is placed above the top plate [48, 112, 113] or underneath the bottom plate [114, 115] as per the requirement. For example, a magnet manifold was produced by deleting specific wells from a microplate and a magnet was integrated into the manifold [114]. The magnet position was selected manually by a metal clip. This is the first step for a DMF device to be integrated with a removable magnet manifold. Moreover, this team introduced a novel motor-driven magnetic lens composed of a magnet bar and two steels [116] (Figure 2.7), enabling automated and parallel assays with magnetic particle separation. To downsize the integrated instrument, they introduced a servo motor-controlled magnet lens [117], which could adjust (b)

(a)

Default: PMT disengaged

Measure: PMT engaged PMT

Pogo pin Interface

DMF Device

(c) Default: Magnet disengaged

Separation: Magnet engaged

Magnetic Lens

Figure 2.7 DMF devices integrated with magnetic lens.

2.4 Digital Microfluidics Integrated with Other Devices

Figure 2.8 DMF devices integrated with a heater design.

Electrowetting pixel

Droplet 4 μm

Sensor

10 μm

Heater Silicon dye 200 μm

the magnet condition by rotation rather than moving up and down, as described in previous works. In summary, the integration of magnet with DMF devices is being developed in the trend of simplicity, compactivity, automation, and parallel assays in recent years, increasing the potential for DMF devices to be utilized in more point-of-care testing (POCT) fields. 2.4.1.3

Heater Module

A heater module integrated with DMF devices enables various applications, including polymerase chain reaction (PCR), loop-mediated isothermal amplification (LAMP), recombinase polymerase amplification (RPA), and so on. The heater modules can be divided into two classes. The first class is typically composed of two or three different temperature-controlling parts on one chip each staying at constant temperature [75, 118–122]. The second class includes one temperature-controlling part with alternating temperatures when needed [45, 123, 124]. In the first class, for example, Pollack and coworkers [75] designed a heater in which thermocycling is performed by circularly actuating a droplet between two immobilized temperature zones. In this design, two aluminum bars settled underneath the chip were taken as thermal control. Each aluminum heater includes a silicone rubber heater and a thermistor sensor connected to a proportion integration differentiation (PID) module controller. The system is powered by a DC power supply. This instrument with the heater design was applied to a series of systems, such as pyrosequencing [121] and real-time polymerase chain reaction (RT-PCR) for detections of DNA levels in the clinical specimens [118–120]. As of the second class, Shepard and coworkers [124] reported a temperature control that was accomplished by an embedded resistive heater and temperature sensors regulating the chip surface temperature to an accuracy of 0.45 ∘ C (Figure 2.8). With a PID control loop, the surface temperature regulates in 13 seconds for each change in PCR. In addition to heater modules listed above for PCR requiring alternative changes among three temperatures, recently many designs have been developed for RPA [125, 126] and LAMP [127, 128], which only need fixed temperature. 2.4.2 2.4.2.1

Detection Systems Integrated with Digital Microfluidics Optical Methods

Fluorescence detectors assembled to DMF devices have been widely applied in immunoassay, quantitative detection of targets, and real-time detection for PCR

59

60

2 Digital Microfluidics for Bioanalysis

Light enclosure Camera

Macrolens

Light guide LED light source

Emission filter

Focusing lens Excitation filter EWOD device Aluminum holder

Heater

Figure 2.9 DMF devices integrated with fluorescence platform.

and RPA. A number of studies demonstrated fluorometric enzymatic assays performed on DMF platforms using miniature fluorometers or standard fluorescence microscopes [112, 129]. In addition, a fluorescence plate reader was used to evaluate extraction efficiency in solid-phase extraction [130]. Custom-built microfluorometers and fluorescence microscopes were also utilized for detection of various targets [40, 48, 88]. For example, a DMF device implemented with a fluorescence microscope and a CCD camera was developed for detection of bacterial RNA in Legionella [113]. In addition to quantitative detection, fluorescent monitoring of DNA amplification reaction on chips has also been reported. In a recent study, optical detection for real-time PCR on DMF devices was carried out using a home-made fluorimeter consisting of a photodiode and a light-emitting diode [75]. In order to improve the sensibility of the fluorescence detection, the fluorimeter was modified to four independent channels, each of which has a fluorescein isothiocyanate (FITC) filter set and a long-pass dichroic mirror [119]. In another study, a wide-field imaging fluorescence detection system integrated to DMF platforms was used for monitoring the RPA reaction process [126] (Figure 2.9). The system is composed of a LED light resource connected with a focusing lens as well as an excitation filter and a camera with a macrolens and an emission filter. In addition, Haig Norian et al. reported the use of complementary metal oxide semiconductor (CMOS) with single-photon avalanche diodes (SPADs) in it for real-time detection in qPCR [124]. Chemiluminescence is also a broadly used signal amplification method in immunoassay. Typically, horseradish peroxidase (HRP)-H2 O2 -chemiluminescence substrate system is utilized and a photomultiplier tube (PMT) is applied to detect chemiluminescence signal. In recent years, with DMF being applied for immunoassay, the chemiluminescence detection module is integrated into DMF device with the advantages of portability, reagent saving, and high sensitivity. One of the examples, by Zhou and coworkers [27], developed a single planar DMF chip with ITO electrode and directly stuck the chip onto a PMT in a lightproof box. They enhanced the chemiluminescence signal detected by the

2.4 Digital Microfluidics Integrated with Other Devices

PMT through changing the contact angle of the droplet, which changed the shape of the droplet and concentrated the chemiluminescence onto the active area of PMT consequently. They used luminol (3-aminophthalhydrazide) as the chemiluminescence substrate and added PIP (4-iodophenol) to improve chemiluminescence efficiency. Later [131], they replaced PMT with a smartphone above the chip and used the photo took by the smartphone to analyze the chemiluminescence intensity. Another example from Wheeler’s group [116] developed a shoebox-sized DMF device with a PMT and magnet inside. When PMT was needed, a motor inside the device would drive the PMT with a black box to less than 1 mm above the chip and the measurement of chemiluminescence intensity was conducted. After measurement, PMT would be moved to about 5 cm above the chip. With this highly integrated DMF device, they performed various immunoassays for different analytes such as thyroid-stimulating hormone (TSH) [116], rubella virus (RV) IgM and IgG [49], and estradiol [111]. Other optical systems, such as SERS and surface plasmon resonance (SPR), have been integrated to DMF devices as well. SERS has been an appealing method for multiplex and sensitive detection because of its advantages on signal amplification and spatial resolution. Our group [30] reported an integrated DMF–SERS platform for immunoassay of disease biomarkers using a core@shell nanostructure with 4-mercaptobenzoic acid (4-MBA) as tags. In this work, SERS spectra were achieved from a portable Raman instrument. Several applications using DMF–SPR have also been reported for the detection of DNA hybridization reactions [132, 133]. For a typical detection, the SPR detection spots were fabricated on the transparent top plate, and the SPR measurements were acquired by aligning the chip onto the stage and adjusting it to fit the SPRi prism. 2.4.2.2

Electrochemical Methods

In recent years, the tendency to integrate electrochemical detection systems into DMF devices has been obvious because of the simplicity, easy automation, and high compatibility of electrochemical sensor (ECS). Wheeler’s group introduced a DMF platform containing gold working electrodes and silver counter/reference electrodes on the top plate for amperometric detection [134]. Similar to the system aforementioned, an EWOD DMF device with parallel-plate structure was proposed where the three-microelectrode-system-integrated sensor, along with ground electrode, was patterned on the top plate [135]. To improve the sensitivity of the sensor, working electrodes were modified to Au-nanostructured microelectrodes (NMEs) with a three-dimensional structure [50]. In another study, DMF was integrated with cell culture and electroanalysis by five electrically isolated electrodes on the top plate. Four electrodes were composed of two working electrodes and two counter/reference electrodes, and a fifth electrode was designed for cell culture and functioned as a counter electrode [136]. Considering that conventional fabrication of electrochemical detection systems on DMF devices is labor- and time-consuming, several methods of ECS fabrication have been reported to simplify the procedure and expand its application. Recently, a paper-based microfluidic chip implemented with ECSs, which was modularly fabricated on chips by printing techniques, was developed for

61

62

2 Digital Microfluidics for Bioanalysis

detection of three biochemical markers. However, post-processing of the sensor to enhance sensitivity for electroanalysis still needs electrochemical procedures, such as reducing graphene and depositing gold nanoparticles [137]. More recently, a report has demonstrated the feasibility of on-chip fabrication of a potassium-selective sensor, showing improved lifetime and reconfigurability of the sensor integrated on DMF devices. The potassium-selective sensor array in this report comprised two Ag/AgCl redox electrodes on the bottom plate fabricated by electroplating Ag and oxidation of Ag layer, and the key to make the sensor work was covering a potassium ion-selective membrane on one of the redox electrodes. It is worth mentioning that all the fabrication steps were performed through droplet actuation on a DMF device [138]. 2.4.2.3

Other Detection Methods

MS is a popular technique for the identification and quantification of various analytes for bioanalysis. The interface between MS and DMF devices has been propelled rapidly by Wheeler’s group in recent years [38, 109, 139, 140]. In 2012 [38], they developed a method to integrate DMF with nanoelectrospray ionization mass spectrometry (nESI-MS), which was constructed by putting a glass capillary emitter between the top plate and bottom plate. For further integration, they reported an interface of HPLC–MS and DMF devices [109], allowing direct access to end users. This technique lowered the threshold of utilizing DMF–HPLC–MS device for users and provided great promise for a wide range of applications. In addition, field-effect transistor (FET) is utilized widely owing to its advantages of small size and high sensitivity for label-free detection. Choi’s group developed a FET-based sensor and integrated it into the DMF device [141]. FET consists of four parts: source, drain, gate, and body; source and drain are connected by the body, which is a semiconductor typically, and electrons can flow from source to the drain. The voltage between gate and body creates an electric field, which can influence the electron flow and finally change the drain current (I d ). In this work, they bound avian influenza virus antigen to the FET in a region between gate and drain to detect avian influenza antibody. When the charged antibody is bound to the antigen in FET, the electric field between the source and gate would be influenced and finally led to a drop of drain current. They integrated this FET between the DMF electrodes and created a hole above it to let the solution get into the FET to detect avian influenza antibody. As a rapidly developing field, DMF is expected to have a wide influence in bioanalysis. The integration of other devices facilitates their biological applications. This part has summarized different integrations with DMF devices for sample processing or detection methods. Integrated sample processing includes world-to-chip interfaces, the magnet separation module, and the heater module, enabling sample collection, conservation, preconcentration, separation, and heating. As of detection methods, optical detection, electrochemical detection, MS, and FET have been integrated with DMF devices, allowing diverse signal readout for further applications. As integrated DMF techniques mature rapidly in recent years, they show great promise for portable bioanalysis development for commercial applications beyond academic fields.

2.5 Biological Applications on DMF

2.5 Biological Applications on DMF DMF presents many dramatical advantages, which make it an attractive and promising candidate for various applications. A good deal of DMF studies has concentrated on biochemical and bioanalytical processes, such as enzyme assays, immunoassay, DNA-based applications, and cell-based applications. The automation of sample processing markedly simplifies the process and reduces the risk of exposure to infectious samples. Furthermore, the consumption of expensive reagents and precious samples is obviously reduced, making it cost-effective in medical and clinical applications. The biggest challenge of using DMF in biological systems is biofouling because nucleic acids and proteins tend to adsorb on the chip surface, which may cause sample loss or cross-contamination [142]. When a droplet contains a large amount of proteins, it becomes difficult to move because proteins accumulated on the surface of the chip, making the hydrophobic surface become hydrophilic, which shorten the chip lifetime [30]. Several strategies have been developed to prevent the “biofouling,” such as Pluronic additives [143], trying to make DMF compatible with more biological applications. 2.5.1

Enzyme Assays

Enzyme assays play a key role in most biochemical analysis, and a variety of enzyme assays have been implemented on DMF. Taniguchi et al. [41] demonstrated the feasibility of enzyme assay by moving and merging the luciferase and luciferin droplets on DMF surrounded with oil medium. Srinivasan et al. [95] developed the first quantitative enzyme assay for glucose detection in clinical diagnostic applications. Glucose was detected using an enzyme colorimetric approach, and a photodiode and a light-emitting diode were applied to sense color change. Compared with standard measurements by using a photometer, there were no apparent changes in enzymatic activity with electricity. The glucose detection exhibited a linear range with 2 orders of magnitude in physiological samples. However, no detailed analysis of enzyme activity were reported. Miller implemented homogenous enzyme assays to detect and quantify small molecules [144]. Alkaline phosphatase was chosen to catalyze fluorescein diphosphate into fluorescein. As shown in Figure 2.10a–d, two droplets containing alkaline phosphatase and fluorescein diphosphate were plugged from reservoirs and mixed to give a fluorescence signal. A lower detection limit with a decrease of 2 orders of magnitude was obtained in a small volume of 140 nl. In the meantime, enzyme kinetics experiments were carried out (Figure 2.10e) and the kinetic constants were consistent with those measured by the conventional well-plate method. 2.5.2

Immunoassay

Immunoassay is a classical and effective approach to detect relevant analytes based on specific antibody–antigen interactions, but it often requires complex, repetitive, and multistep protocols. DMF is a well-suited platform for applying

63

64

2 Digital Microfluidics for Bioanalysis

(a)

(b)

Reservoirs (c)

(d)

Figure 2.10 Fluorescence enzyme assay on DMF. Alkaline phosphatase droplet and fluorescein diphosphate droplet were dispensed from left and right reservoir, respectively (a, b), and merged (c). After active mixing, alkaline phosphatase catalyzed fluorescein diphosphate into fluorescein and gave a strong fluorescence signal, and the fluorescent signal was monitored at real time (e).

immunoassay, not just decreasing reagents and sample consumption but also the assay time is greatly reduced as there is no limitation of diffusion kinetics in droplets. So far, there are two immunoassay approaches on DMF. One is immobilizing antibody on the surface of DMF, and the other relies on antibody immobilized on another solid supports, such as magnetic beads. Miller et al. [145] implemented the first homogenous immunoassay on DMF for human IgG detection. As shown in Figure 2.11A, the anti-IgG antibody was physisorbed on the top plate of the DMF device to capture target protein from samples. Detection antibody was labeled with FITC to give a fluorescence signal. The fluorescence was acquired by the multiwell plate reader after the final wash. To eliminate nonspecific biofouling, pluronic F-127 was added in droplets. As a result, the total assay time was shortened to 10%). One approach to overcome such limitations is through the application of supercritical fluids [81, 82]. At high pressures, a solvent will eventually reach a supercritical regime, where the fluid solubilizes a wide range of compounds in a similar manner to a liquid, while displaying the high miscibility and fast diffusion rates typical of gases [83]. With these advantages, supercritical fluids offer new opportunities for the microfluidic synthesis of QDs. Marre et al.

119

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

used such solvents in the continuous-flow synthesis of CdSe QDs within a silicon/Pyrex microreactor that was able to withstand high pressures (Figure 4.2a) [58]. In this study, the entire setup was first pressurized using nitrogen gas. The precursor solutions were then delivered into the reactor and mixed at room temperature before entering a high-temperature reaction zone, where QD nucleation and growth was initiated. The reaction mixture ultimately exited the microreactor at a low temperature, quenching further QD growth, and was collected in a high-pressure reservoir. To provide a direct comparison for their studies, the authors synthesized CdSe QDs in both a conventional solvent (squalene) and a supercritical fluid (sc-hexane). They explored the influence of residence time, precursor concentration, and solvent phase on the QD size, nuclei concentration, and size distribution. Notably, and as shown in Figure 4.6b, results demonstrated that QD size varied as a function of residence time, leading to a red shift in the photoluminescence (PL) spectra. In comparison to squalene, the authors also observed that the use of sc-hexane led to a significant decrease in the QD size distribution, reported through a reduction in the full width at half-maximum (FWHM) of the emission peak (Figure 4.6b). The two main contributing factors to this decrease in size distribution are the RTD difference of the solvents and the kinetics of nanocrystal formation. Specifically, hexane has a 10–20-fold lower viscosity compared with squalene, leading to a much lower RTD. Accordingly, the reaction mixture in sc-hexane experiences more uniform reaction conditions. Furthermore, sc-hexane possesses a higher supersaturation point, which induces a higher initial concentration of nuclei for subsequent growth, fast depletion of precursors in solution, and a short nucleation phase.

FWHM = 41 nm

Squalane

Graphite layer Pressurized N2 gas cylinder

Halo etch

L1

QDs growth

Mixing area

N2

L1 L 2

L2

tR = 30 s tR = 60 s tR = 150 s

1 2

3

2

3

(a)

Pressurized reservoir (4 vials)

3

4

5

6

50

Outlet

Compression/ cooling parts

6

Sc-Hexane

Sample Valve

Water flow

2 3 4 5 30 σ/davg 25 = 10.3% 20 15 10 5 0 dcdse (nm)

FWHM = 26 nm Frequency (%)

High pressure syringe pump

N2

1 2 3

Frequency (%)

Heating block

Emission intensity (a.u.)

120

1 2

500

20 10 0

50 bars 400

σ/davg 40 = 4.6% 30

600

dcdse (nm)

700

Wavelength (nm)

(b)

Figure 4.6 (a) Schematic of experimental setup used for supercritical fluid-assisted synthesis of QDs. The high-pressure microreactor is integrated with compression-cooling parts, a high-pressure syringe pump, a high-pressure valve, and a high-pressure reservoir containing four vials; (b) PL spectra at varying residence times (tR ) and QD size distributions. Size distributions were determined for samples run at tR = 60 seconds. Source: Marre et al. 2008 [58]. Reproduced with permission of John Wiley & Sons.

4.3 Continuous-Flow Microfluidic Synthesis of Quantum Dots

4.3.2

Heterogenous Core/Shell Quantum Dots in Continuous Flow

Core-type homogenous structure QDs perform well in idealized conditions; however, in practice, they suffer from some innate problems. The size and shape of QDs make their surfaces highly energetic and reactive. Such a reactive state can be somewhat passivated by the presence of surface capping ligands. Nevertheless, in practice, it is extremely difficult to arrest surface reactions such as oxidation. When these reactions occur, energetic traps are created where electrons of the all-important exciton become ensnared, thus prohibiting PL. In the bulk synthesis of QDs, a range of strategies to cap QD cores through epitaxial growth of an inorganic shell have proved successful. Such a shell has a band gap offset versus the core, rendering it energetically unfavorable for the electrons or holes of the excitons to diffuse into the shell. This general approach has become a standard solution to the problem of fluorescence quenching of QD cores. Thus, it is imperative that microfluidic approaches are capable of multistep growth to create heterogenous, or core/shell, QDs. Fortunately, this challenge was met by a series of innovations, on a variety of QD material types, as described now. 4.3.2.1

Zinc Selenide/Zinc Sulfide (ZnSe/ZnS)

In 2012, Kwon et al. at the Korea Advanced Institute of Science and Technology [84] reported a novel approach for the microfluidic synthesis of ZnSe/ZnS QDs. This work is notable for its demonstration of the continuous synthesis of core/shell-type QDs via a single injection of precursors, thus avoiding the need for more complex multistep synthetic routes [85, 86]. Briefly, the authors fabricated a cyclic olefin copolymer (COC) microfluidic reactor via injection molding [87]. Importantly, COC is stable under acidic and basic conditions and resistant to many organic solvents, while also being transparent in the visible region of the electromagnetic spectrum [88]. The developed microreactor consisted of three sections, one for precursor mixing, one for ZnSe core formation, and one for ZnS shell coating (Figure 4.7a). The precursors (Zn, Se, and S) were introduced together, with Zn and S being consumed first to produce ZnSe cores, followed by a reaction between the Zn and S precursors to form the ZnS shells. The resulting UV/vis absorption and PL spectra of the ZnSe/ZnS QDs are shown in Figure 4.7b,c. Significantly, it was observed that ZnSe/ZnS QDs exhibited significantly stronger emission than the ZnSe QDs. After the optimization of flow rates, the emission intensities increased by 270% (compound 12 in Figure 4.7), with a narrower size distribution implied by a reduced FWHM of the band edge emission. Accordingly, through the use of thermoplastic microreactors, the authors opened a new and facile route for the microfluidic synthesis of core/shell QDs. 4.3.2.2 Cadmium Selenide/Zinc Sulfide (CdSe/ZnS) and Cadmium Telluride/Zinc Sulfide (CdTe/ZnS)

As noted, QDs have attracted much interest as tools in biological sensing [89]. For such applications, it is necessary to put the final QD constructs into aqueous media and ensure maintenance of their advantageous colloidal and optical properties. This is typically achieved through the addition of a layer of organic ligands bearing hydrophilic groups and QD-binding moieties. QD synthesis can

121

Injection

Mixing

ZnSe Core

ZnSe/ZnS Core/shell

Nomalized absorbance (a. u.)

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

(b)

300

FZn : FSe : FS (μL min–1) 15 : 30 : 30 (ODE No.9) 15 : 30 : 30 (No. 10) 10 : 20 : 20 (No. 11) 5 : 15 : 15 (No. 12) 5 : 10 : 10 (No. 13)

350 400 450 Wavelength (nm) FZn : FSe : FS (μL min–1)

Zn(OLA) TOPS TOPSe

(a)

Mixing zone

Heating zone Zn

Se

Intensity (a. u.)

122

ZnS

350

(c)

15 : 30 : 30 (ODE, No. 9) 15 : 30 : 30 (No. 10) 10 : 20 : 20 (No. 11) 5 : 15 : 15 (No. 12) 5 : 10 : 10 (No. 13)

400 450 500 Wavelength (nm)

Figure 4.7 (a) A schematic of ZnSe/ZnS QD synthesis in a thermoplastic microfluidic reactor; (b) UV/vis absorption spectra; and (c) PL spectra of synthesized QDs as a function of precursor flow rate. Sample No. 9 is composed of ZnSe QDs, while samples No. 10, 11, 12, and 13 are ZnSe/ZnS QDs. Source: Kwon et al. 2012 [84]. Reproduced with permission of John Wiley & Sons.

be performed in aqueous conditions; however, it is much more common to carry out the synthesis in organic solvents and then perform a post-processing step to transfer them into aqueous solution. Accordingly, such a process of QD core synthesis, followed by inorganic shell growth and hydrophilic ligand attachment, makes for relatively complex synthetic procedures. In 2010, Kikkeri et al. at the Max Planck Institute for Colloids and Interfaces [90] reported the synthesis of biofunctionalized core/shell QDs in microfluidic reactors. The researchers used a series of glass microfluidic reactors to synthesize and functionalize CdSe/ZnS and CdTe/ZnS core/shell QDs bearing carboxyl groups and carbohydrates. The synthesis of the QD cores, growth of the inorganic shells, and functionalization were performed in discrete stages using separate continuous-flow microfluidic chips (Figure 4.8a), with purification being performed off-chip after each step. Briefly, the Cd and Se or Te precursors were injected into a first device to form the QD cores. The QDs were then purified off-chip and injected into a second device with a ZnS precursor to grow the shell layer. The core/shell QDs were purified off-chip again and finally injected into a third device for ligand exchange and surface modification before being precipitated and taken into aqueous solution. Significantly, the authors were able to demonstrate the synthesis of CdSe and CdTe QDs of controllable size by adjusting the reaction time within the first microfluidic device (Figure 4.8b), noting that longer reaction times resulted in a larger PL FWHM and decreased PL quantum yield (QY). However, upon formation of the ZnS shell in the second device, the PL QY increased from 23% to 31% because of the stabilizing effect of the shell. Successful functionalization with carboxyl groups and carbohydrates in the third microfluidic device was confirmed by

4.3 Continuous-Flow Microfluidic Synthesis of Quantum Dots

(a)

(b)

Cd precursor

CdSe quantum dots

CdTe quantum dots

Core Se or Te precursor

OA/TOP 160˚C ZnS precursor

HO HO

90˚C Carbohydrate precursor

OH OH O O

O

Core ZnS OA/TOP

S

O

Residence time (min): 3

10

20

30

3

10

20

or

(c)

HO HO

O

O

O

O

S

OH

O

Core ZnS

or

HO S

S

1,6E+05 Fluorescence Intensity

OH

1,2E+05 8,0E+04 4,0E+04 0,0E+00 420

470

520

570

620

670

720

Wavelength (nm)

Figure 4.8 (a) Schematic of a microreactor setup for the synthesis of carboxylated or carbohydrate-coated CdTe/ZnS and CdSe/ZnS core/shell QDs. Three continuous-flow microfluidic chips were used for each step: (1) synthesis of the QD core, (2) formation of shell layer, and (3) functionalization with the biologically relevant molecule. (b) Images of CdSe and CdTe QDs under UV light and (c) normalized photoluminescence spectra of CdSe (---) and CdTe (—) QDs as a function of reaction residence time. Source: Kikkeri et al. 2010 [90]. Reproduced with permission of John Wiley & Sons.

studying the interactions between the sugar-coated QDs and lectin concanavalin A (ConA). Some QDs were functionalized with α-mannose, while others were functionalized with β-galactose. Because ConA binds with α-mannose and not β-galactose, addition of ConA to α-mannose-coated QDs caused the QDs to aggregate, whereas ConA had no effect on the β-galactose-coated QDs. This was verified via turbidity measurements, with the addition of ConA to α-mannose-coated QDs resulting in immediate turbidity, whereas little turbidity was observed with β-galactose-coated QDs. Overall, this work was notable for its synthesis of biologically applicable core/shell QDs, a feat that is relatively challenging even on a bulk scale. 4.3.2.3

Copper Indium Sulfide/Zinc Sulfide (CuInS2 /ZnS)

The presence of Cd in traditional QD formulations, and its associated toxicity, has led to efforts to find alternative formulations that use minimally toxic elements but still possess the advantageous properties of traditional QDs. Recently, copper indium sulfide (CuInS2 ) QDs have emerged as an interesting competitor in this field. They are heavy metal-free and can emit at wavelengths between the visible and NIR regions of the electromagnetic spectrum [91, 92]. The initial exploration of CuInS2 QD synthesis routes utilized bulk methods to good effect, but obvious inherent issues regarding varying local reaction conditions across bulk solutions remained. In 2016, Tian et al. from Bayer Technology Services [93] used a continuous-flow microfluidic approach for the facile and large-scale synthesis of CuInS2 and CuInS2 /ZnS QDs. This required reformulation of the bulk precursor solutions to ensure compatibility with microfluidic processes and involved replacing copper actinium (CuAc) with copper(I) iodide (CuI) because of its superior

123

124

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots Microheat exchanger N2 N2

Pump

CulnS2 core precursor

N2 CulnS2 core precursor

Microreactor

CulnS2 nanocrystals

Pump Microreactor Microheat exchanger N2 Microreactor

N2

Pump

ZnS shell precursor

CulnS2/ZnS nanocrystals

Figure 4.9 A simplified schematic of a CuInS2 core QD (top) and CuInS2 /ZnS core/shell QD (bottom) continuous-flow synthesis. Source: Tian et al. 2016 [93]. Reproduced with permission of Elsevier.

stability under ambient conditions. They synthesized core-type or core/shell QDs in microreactors that fed directly into a microheat exchanger to rapidly cool the reactions and halt particle growth, yielding populations of stable QDs (Figure 4.9). For the synthesis of core-type CuInS2 QDs, the authors used a single microreactor and discovered that lower flow rates led to higher PL QYs. Indeed, they reported an optimized value of 28%, which was the highest ever value obtained for CuInS2 QDs. The products exhibited a narrow-size distribution, and the workflow achieved a synthesis rate more than twice that of previous reports. For ZnS shell formation, the authors incorporated a second microreactor, with core and shell growth occurring serially before the reaction solution was fed into the heat exchanger. Interestingly, the authors implemented staged temperature variations and a distributed feed to ensure slow and homogenous shell formation. Motivated by multistep bulk synthesis procedures, the staged temperature approach incorporated three microreactors at different temperatures, with core growth at 250 ∘ C and shell growth at 180 and 230 ∘ C. This resulted in a reduction in crystal defects in the ZnS shell and led to CuInS2 /ZnS QDs with a 49% PL QY. In the distributed feed approach, the shell precursor was injected four times, instead of just once, to maintain a high concentration of precursor during shell formation.

4.3 Continuous-Flow Microfluidic Synthesis of Quantum Dots

The authors found that each additional injection of shell precursor led to an increase in PL QY. With the first demonstration of a fully continuous system for CuInS2 and CuInS2 /ZnS QD synthesis, this general approach is a promising step toward large-scale high-quality production of a promising class of Cd-free QDs. 4.3.2.4

Indium Phosphide/Zinc Sulfide (InP/ZnS)

As previously described, the production of microreactor systems that incorporate consecutive reaction steps is important in facilitating the transfer of bulk approaches to the microscale. In this vein, Baek et al. at the Massachusetts Institute of Technology [94] developed a multistage synthesis platform incorporating up to six high-temperature and high-pressure microchip reactors (Figure 4.10a). The system was used to synthesize InP/ZnS core/shell QDs with narrow-size distributions, tunable emission peaks, and high PL QYs. To enhance the versatility of the platform, the authors used silicon-Pyrex chips to enable high-temperature and high-pressure reactions, with octane being used as a solvent, because of its low dispersion for both core and core/shell growth at or near supercritical conditions. InP cores were synthesized within the first two microreactors, while nanocrystal growth was initiated in the third. The ZnS shell was formed in the following three microreactors, where 10 subchannels maintained a low concentration of the shell materials and prevented undesired ZnS nucleation (Figure 4.10b). A customized in-line optical device was used to characterize the absorption and emission of the synthesized QDs. By adjusting the number of precursor injections within the first two microreactors, InP QD size could be precisely controlled. Furthermore, the addition of myristic acid allowed the platform to overcome traditional limitations in the growth of large InP QDs (Figure 4.10e). The PL FWHMs were as low as 42 nm, while the PL QYs were as high as 40%. Importantly, the authors confirmed that their InP/ZnS QDs remained luminescent after phase transfer into aqueous solution. To demonstrate the versatility of the platform, InP/ZnS QDs were also synthesized using low boiling point short-chain alkyl thiols as solvents. Interestingly, the authors were able to adjust the temperature and injection sequence of their modular platform to synthesize InP/ZnSe, InP/CdS, and InAs/InP QDs. 4.3.3 Flow

Heterogenous Core/Multishell Quantum Dots in Continuous

When adding shell layers to QD cores, there are two major factors that must be considered. First, the lattice mismatch between the core and the shell should be minimized to allow the synthesis of QDs with minimal lattice defects. Second, the QD shell should have a band gap significantly higher than the core to prevent electrons tunneling from core to shell. Although heterogenous core/shell QDs demonstrate superior PL efficiencies compared to homogenous core-type QDs, it is uncommon for a single shell material to fulfill both criteria. To tackle this, researchers have produced core/multishell QDs, where epitaxial growth of multiple shells are sequentially performed, which provides an intermediate layer to transition between the lattice mismatches of the core and outermost shell

125

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

Shell formation

Shell Sequential format ion growth

Annealing

Aging Mixing (TMS) S 2 (TMS) S 2 (TMS) P 3

Zn(OA)

2

Zn(OA)

2

In(MA)

2

Detector

In(MA)

2

UV-VIS

MA

(TMS) P 3

OUT

(b)

(a)

(c)

Emission (a.u.)

Absorbance (a.u.)

126

400 500 600 700 800 900

400

Wavelength (nm)

(d)

500

600

700

800

H2O

H2 O

CHCl3

CHCl3

900

Wavelength (nm)

(e)

(f)

Figure 4.10 (a) A multistage microfluidic platform for the synthesis of InP/ZnS core/shell QDs. The first three stages are used to synthesize InP cores, while the last three are used to synthesize the ZnS shell layer; (b) a shell formation microreactor consisting of 10 subchannels; (c) InP/ZnS QDs of various sizes under 365 nm illumination; (d) absorption and (e) emission spectra of the InP/ZnS QDs; and (f ) phase transfer of the InP/ZnS QDs from chloroform (left) to water (right). Source: Baek et al. 2018 [94]. Reproduced with permission of John Wiley & Sons.

while maintaining their large band offset. These heterogenous core/multishell structures have become established as an important class of QD; therefore, it is critical that microfluidic reactors are able to produce QDs with multiple shells engineered in a precisely defined manner. 4.3.3.1

Cadmium Selenide/Cadmium Sulfide/Zinc Sulfide (CdSe/CdS/ZnS)

In contrast to systems that incorporate increasing numbers of processing units to mimic the complexity associated with bulk synthetic methods, other

4.3 Continuous-Flow Microfluidic Synthesis of Quantum Dots

studies have focused on reformulating reaction precursors to naturally favor the formation of specific QD structures. This approach is particularly attractive when the reaction products are complex, for example, when core/multishell structures are desired. A prominent example in this respect is CdSe/CdS/ZnS. Although ZnS shell formation is known to significantly stabilize CdSe QD cores, the extent of the improvement is inherently limited by the relatively large lattice mismatch between ZnS and CdSe [95], which creates lattice strains at the interface between the core and shell [96]. Through the introduction of a CdS intermediate layer, studies have demonstrated a reduction in lattice strain and a corresponding increase in PL QY using bulk reactors [96]. Because of the inherent complexity of such a structure, however, some innovation is required to transfer the synthesis to the microscale. In 2015, Naughton et al. from the University of Illinois at Urbana-Champaign developed a continuous-flow platform for the synthesis of CdSe/CdS/ZnS QDs [97]. Interestingly, the focus of this work was on the formulation of the Cd, Se, S, and Zn precursors, with a relatively simple continuous-flow setup being used (Figure 4.11). Precursor mixing could be done before injection, obviating the need for online mixers and simplifying microreactor design. The authors explored various core/shell compositions and compound alloys. For the synthesis of CdSe/CdS/ZnS QDs, the Cd precursor was cadmium oxide (CdO) in oleic acid, the Se precursor was trioctylphosphine selenide (TOP-Se) in octadecene (ODE), and the S precursor was trioctylphosphine sulfide (TOP-S) in ODE. Zn for the shell synthesis was supplied from zinc diethyldithiocarbamate (Zn DDTC2 ) in TOP, which thermally decomposes into ZnS. After mixing the precursors and injecting them into the flow reactor, the CdSe cores formed first, followed by the CdS intermediate layers, and finally the ZnS shells. This occurred since Se has a higher reactivity than S. ZnS was then formed over the CdSe/CdS NPs through Zn DDTC2 decomposition. The authors elegantly showed that they could tune the emission wavelength of the CdSe/CdS/ZnS QDs by adjusting the average residence time and the number of ZnS layers. By doing this, they were able to obtain emission wavelengths

Oil bath CdSe/CdS/ZnS

CdSeS/ZnS Syringe pump

Heater

CdS/ZnS

Figure 4.11 A schematic illustrating the continuous-flow setup used to synthesize heterogenous core–multishell QDs. The syringe pump is used to motivate reagents, while an oil bath maintains the temperature of the reactor. The synthesis of CdSe/CdS/ZnS, CdSeS/ZnS, and CdS/ZnS QDs was demonstrated. Source: Naughton et al. 2015 [97]. Reproduced with permission of Royal Society of Chemistry.

127

128

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

covering the entire visible region of the electromagnetic spectrum. Additionally, screening studies indicated that two ZnS layers led to the highest quality QDs, agreeing with literature reports [96]. Optimized CdSe/CdS/ZnS QDs had a 60% PL QY, which is significantly above the 12% for CdSe/ZnS core/shell QDs synthesized using the same platform. 4.3.4

Summary of QD Classes

In this section, we have introduced continuous-flow microfluidic synthesis of a wide range of QDs, including homogenous core-type, heterogenous core–shell, and heterogenous core/multishell QDs. Homogenous core-type QDs are the most classic structures and contain only a single compound. However, they are vulnerable to surface degradation and oxidation because of insufficient protection from their organic capping ligands. Therefore, they are commonly capped with a second semiconductor compound – epitaxially grown upon the core – to completely passivate the surface and provide a band gap offset to restrict electrons from tunneling into the shell. However, it is often advantageous to grow multiple shells of different compounds, which allows better tuning of the lattice mismatch and band gap offset. These are termed core/multishell QDs.

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots As discussed, continuous-flow reactors provide significant advantages over bulk synthesis methods. However, there are two common problems that ultimately limit their efficiency and application. First, in all continuous-flow microfluidic devices, viscous drag at the channel walls causes nonuniform fluid velocities across the width of the channel, leading to finite and appreciable RTDs and ultimately a broadening of achievable particle size distributions [51, 57]. This problem is especially acute in high-viscosity solvents, which are commonly employed in QD synthesis. Second, deposition of solid material on channel walls over time is almost unavoidable. This severe issue constrains reactor lifetimes, which in turn limits possibilities for reaction scale-out. An effective method to overcome both the above problems is to employ a segmented flow rather than a continuous flow. Droplet microfluidic systems not only minimize dispersive flow and solid deposition but also confer rapid mass and heat transfer, high-throughput operation, precise reagent control, and reliable automation. We now discuss some key droplet-based microfluidic strategies for the synthesis of QDs, with a focus on the QD materials and structures that have been explored. Additionally, examples of kinetic studies of QD nucleation and growth, multistep synthesis techniques for multicomponent QD synthesis, and in situ spectroscopic modules for QD characterization during (online) and after (in-line) particle formation are described.

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots

4.4.1

Homogenous Structure Quantum Dots in Segmented Flow

4.4.1.1

Cadmium Sulfide (CdS)

Reports on the synthesis of QDs in segmented flows started appearing in the literature shortly after the initial continuous-flow studies, with Shestopalov et al. at the University of Chicago initially reporting the synthesis of QDs in droplet-based microfluidic reactors in 2004 [98]. In this work, the authors described an aqueous synthesis of CdS QDs at room temperature within a PDMS reactor. They performed elegant control experiments to demonstrate the efficacy of their system, including a direct comparison between continuous flow and segmented flow in the same device (Figure 4.12). For the continuous-flow experiments, streams of the Cd (CdCl2 , cadmium chloride) and S (Na2 S, sodium sulfide) precursors were brought together, but separated by a stream of buffer, and injected into a water stream at a T-junction (Figure 4.12A). Using such an approach, solid deposition onto channel surfaces began almost immediately, rendering the device ineffective after a few tens of minutes. In contrast, when the water stream was replaced with an oil stream, the aqueous reaction solution spontaneously segmented into droplets because of the shear force from the fluorinated carrier oil (Figure 4.12B), with subsequent synthesis of CdS QDs occurring only inside the droplets. In (A) 6 minutes

Buffer CdCl2 Na2S

30 minutes

Water

Out

200 μm

(a) (B)

(b) 6 minutes

Buffer CdCl2

(c) 30 minutes

Na2S

Oil

Out (a)

(b)

(c)

Figure 4.12 A droplet-based microfluidic reactor for CdS QD synthesis developed by Shestopalov et al. prevents fouling of microchannels during QD synthesis. (A) (a) A schematic showing the continuous-flow microreactor. (b, c) Images demonstrating the accumulation of solid CdS on the channel walls after 6 and 30 minutes of continuous flow operation; (B) (a) A schematic of the segmented-flow microreactor. (b, c) Images demonstrating the lack of solid deposition after 6 and 30 minutes of droplet flow. Source: Shestopalov et al. 2004 [98]. Reproduced with permission of Royal Society of Chemistry.

129

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

comparison, after 30 minutes of operation, there was no observable deposition of solid on the channel walls. That said, the scope of aqueous syntheses of QDs is somewhat limited, and thus, it is important for segmented-flow microfluidic systems to adapt to the more common organic solvents. 4.4.1.2

Cadmium Selenide (CdSe)

Reactions in long-chain, high-viscosity solvents typically require high temperatures to trigger the nucleation and growth of QDs. This provides significant motivation to produce segmented-flow microfluidic reactor systems that can process organic solvents in a robust manner and operate at high temperatures. One of the first droplet-based microfluidic reactors for high-temperature QD synthesis was reported by Yen et al. from the Massachusetts Institute of Technology [99], who used a gas–liquid segmented-flow microreactor to synthesize CdSe QDs at 260 ∘ C. The reactor was fabricated in silicon and glass and incorporated multiple temperature-controlled zones (Figure 4.13). A heated aluminum block was used to define the high-temperature reaction zone, while water-chilled aluminum blocks were used for the inlet and quenching zones. The Cd and Se precursors were prepared by solubilizing cadmium 2,4-pentanedionate and selenium in squalene with long-chain ligands. By tuning the injection rates of the Cd and Se precursors, the authors attained good control of CdSe QD size, with significantly narrower size distributions and shorter reaction times Gas

ar

Precursor solutions

Precursors

Deep trench

Hea

ati

b

ting

zon e

a

ng

zon

e

(a)

He

130

Quench zone Outlet

Thermal insulation

1 mm

(b)

Thermal insulation

(c)

Figure 4.13 A gas–liquid slug microreactor with thermal control for CdSe QD synthesis. (a) Inlets for the precursors and gas before slug formation. (b) Reaction heating zone. (c) Time exposure image of the quenching region and outlet under UV radiation. Source: Yen et al. 2005 [99]. Reproduced with permission of John Wiley & Sons.

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots

200 μm

10 nm 300 °C

Carrier fluid (a)

Precursor solution (b)

Figure 4.14 CdSe QD synthesis at temperatures between 270 and 300 ∘ C in a liquid–liquid segmented microreactor with thermal control. (a) Droplets are generated by a cross-shaped nanojet injector. (b) Transmission electron micrographs of CdSe QDs synthesized in droplets of ODE within a PFPE carrier fluid at 300 ∘ C. Source: Chan et al. 2005 [100]. Reproduced with permission of American Chemical Society.

compared to continuous-flow microreactors. Moreover, the PL FWHMs of CdSe QDs from their two-phase segmented-flow reactor (Δ𝜆 = 30 ± 2 nm) were significantly narrower than those obtained in continuous-flow synthesis schemes (Δ𝜆 = 34 ± 5 nm), which was attributed to the nondispersive motion of the reagent droplets [51]. Despite the promise shown by gas-liquid segmented-flow microfluidic platforms, there are two important limitations. First, the carrier gas severely limits the operating stability because its volume can be heavily affected by small changes in temperature or pressure. Second, solid products may still deposit onto the channel walls as the walls come into direct contact with the reaction solution. To address these issues, Chan et al. at the University of California, Berkeley [100], developed a high-temperature liquid–liquid segmented-flow microreactor for CdSe QD synthesis (Figure 4.14). Here, ODE was used as the discrete phase and a high-boiling point perfluorinated polyether (PFPE) oil as the continuous phase. To make droplets of ODE in PFPE over a wide range of flow rates, the authors fabricated glass microreactors with a cross-shaped nanojet injector (Figure 4.14a). PFPE was injected through the side arms of the cross while the ODE solution was injected from the top. The generated droplets then traveled through the 2.4 mm-long nozzle into a heating region where the reaction was initiated. This method allowed their microfluidic devices to maintain reproducible droplet flow. The QD synthesis proceeded by reacting cadmium oleate with a solution of TOP-Se/TOP at temperatures up to 300 ∘ C, with successful product formation confirmed by TEM (Figure 4.14b). The liquid–liquid segmented flow completely eliminated solid deposition on the channel walls during the lifetime of the device (∼5 hours), whereas analogous experiments under continuous flow showed clear deposition after only 20 minutes. 4.4.1.3

Lead Sulfide (PbS) and Lead Selenide (PbSe)

While Cd-based QDs have been shown to emit over the entire visible spectrum, their reach into the infrared spectrum is limited. Pb-based QDs are, in contrast, tunable across a broad region of the infrared (800–3000 nm) [101] and are attractive materials in various applications, such as solar cells, photodetectors, NIR lasers, and biological imaging [102–105]. In 2014, Lignos et al. at the Swiss Federal Institute of Technology in Zürich (ETH Zürich) [106] reported a segmented-flow

131

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

PFPE Oil

[S] PbS

[Pb] Droplet generation In-line fluorescence detection Heating zone

Normalized intensity

132

700 500 μm

(a)

PbSe

1200 1700 Wavelength (nm)

(b)

Figure 4.15 PbS QDs synthesis in a droplet-based capillary reactor with rapid in-line PL detection. (a) A schematic of the capillary microfluidic reactor, integrating an in-line NIR fluorescence detector; (b) tuning of the PL emission by variation of temperature for PbS and PbSe QDs. Source: Lignos et al. 2014 [106]. Reproduced with permission of American Chemical Society.

microfluidic system for PbS and PbSe QD synthesis using capillary tubing rather than the common chip formats discussed previously. Capillary reactors are typically easier to build, operate, and modify because of their modular construction using readily available components, while offering the ability of reaction scale-up [50]. In this work, the authors specifically used polytetrafluoroethylene (PTFE) tubing to define the microfluidic system (Figure 4.15a). Separate syringes loaded with the Pb precursor (lead oleate), S precursor (bis-(trimethylsily) sulfide), and the oil phase were used to deliver fluid into a polyether ether ketone (PEEK) cross, which served as a droplet generator. Formed droplets then traveled through the tubing to a heating zone, where the reaction proceeded at temperatures ranging between 80 and 155 ∘ C. Droplets containing synthesized PbS QDs were then cooled to room temperature and monitored using an in-line fluorescence detection system. Such PbS QDs exhibited narrow-size distributions (5–7%) and improved PL QYs (∼28%) versus comparable bulk-synthesized materials (∼12%). In-line monitoring of the PL of individual droplets allowed ready optimization of reaction parameters and thus excellent control over QD PL (765–1580 nm for PbS and 860–1600 nm for PbSe), as shown in Figure 4.15b. Furthermore, synthesis reactions could be performed using their microfluidic setup for periods between three and six hours. Accordingly, such a system can be tuned to produce kilogram quantities of QDs per day. 4.4.1.4

Perovskite QDs

Lead halide perovskite QDs have attracted tremendous research interest in the past five years because of their remarkable photophysical and photovoltaic properties [107–109]. Similar to standard QDs, their PL properties can be controlled through variations in size and/or composition [110]. For the chemist, this means

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots

that a rich and complex reaction parameter space must be accessed and explored to allow the synthesis bespoke materials with desirable properties. In this respect, segmented-flow microfluidic systems have been shown to be extremely effective tools, with a series of studies demonstrating fast parametric space mapping [111], shape evolution [112], and crystal growth mechanisms [113]. In a pioneering study, Kovalenko and coworkers at ETH Zürich [111] used a capillary-based microfluidic platform to synthesize CsPbX3 (caesium lead halide) perovskite QDs. A key feature of the reaction platform was the integration of online detectors for monitoring both absorbance and fluorescence during QD nucleation and growth (Figure 4.16). Different lead halide precursors (denoted as PbX2 and PbY2 ) were premixed using a T-junction with a variable molar ratio of X to Y, denoted as R2 . The PbXα Y(2−α) mixture was then delivered into a cross-mixer, where the Cs precursor (Cs oleate) was injected at a chosen ratio to Pb, denoted as R1 (Figure 4.16a). In this way, R1 and R2 could be tuned separately and continuously, generating droplets containing varying precursor ratios. The droplets were subsequently motivated in a coil around a heating rod (Figure 4.16b), with the residence time being controlled by the total flow rate. Using this setup, the authors were able to synthesize CsPbX3 QDs with real-time and continuous control of composition and PL (Figure 4.16b,c). Furthermore, ultrafast kinetic measurements (between 100 ms and 5 seconds after reaction initiation) and rapid parametric optimization experiments were performed. This study is not only an important demonstration of a new class of QDs synthesized within segmented On-line absorbance detection

CsPb(Br/Cl)3 CsPbBr3 CsPb(Br/l) 3 CsPbl 3

[Cs-Oleate]

Droplet generation

Oil [PbX2] [PbY2]

(a)

R1 R2

Motorized linear and rotation stage

Normalized PL intensity

On-line fluorescence detection

400 (b)

(c)

500

600

700

800

Wavelength (nm)

Figure 4.16 CsPbX3 perovskite QD synthesis in a droplet-based capillary reactor with integrated online absorbance and fluorescence detection. (a) A schematic of the droplet-based microfluidic platform. The Pb/Cs ratio (R1 ) and halide ratio (R2 ) are continuously varied to allow precise tuning of the chemical contents within the droplets. The online absorbance and fluorescence detection systems are used to probe the product at a range of reaction times, and as early as 0.1 second; (b) an image of generated droplets using different precursor ratios. The picture was taken of the coiled heating zone under UV excitation (405 nm); (c) the whole visible spectrum can be covered by CsPbX3 QD emission, as determined by online fluorescence measurements. Source: Lignos et al. 2016 [111]. Reproduced with permission of American Chemical Society.

133

134

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

flows but also a validation of such systems for exploring the reaction parameter space and performing rapid optimization. 4.4.2

Heterogenous Core/Shell Quantum Dots in Segmented Flow

4.4.2.1

Copper Indium Sulfide/Zinc Sulfide (CuInS2 /ZnS)

Although there have been many reports of homogenous core-type QD synthesis in segmented flow, examples of the synthesis of core/shell QDs via this approach are relatively rare. This is most likely because precise downstream dosing of droplets is challenging, which restricts somewhat the possibilities for multistage synthesis [114]. Nevertheless, Yashina et al. at ETH Zürich recently demonstrated the synthesis of CuInS2 /ZnS QDs in a two-stage droplet microfluidic platform that could independently and precisely dose droplets (Figure 4.17) [115]. Specifically, CuI (copper iodide) and In(OAc)3 (indium triacetate) were used as metal precursors, 1-dodecanethiol as the sulfur source, and PFPE as the carrier fluid. All precursors and carrier fluids were loaded into glass syringes connected with PTFE capillaries. These solutions were then mixed, injected into the setup, and segmented into droplets at a PEEK cross unit. The resultant droplets of the reaction mixture were transferred to a copper-heating rod (with the reaction tubing coiling around), where the reaction was initiated and CuInS2 QDs formed. After the first stage, Zn(OAc)2 (zinc diacetate) precursor was injected into the flow from a side channel. The reformed droplets were then transferred to a second heating rod for growth of the ZnS shell. To monitor the optical properties of the synthesized QDs and to optimize the reaction parameters, two in-line PL detectors were used. This two-stage microfluidic

Fluorescence detection LED To collection Stage 2 [Zn]

[Cu]

CulnS2/ZnS

[S] CulnS2

[In]

Oil

Stage 1

LED

Fluorescence detection

Figure 4.17 Core/shell CuInS2 /ZnS QDs synthesized in a two-stage droplet-based microfluidic platform. The first stage is used to synthesize the CuInS2 core, and the second stage is used for growing the ZnS shell. Fluorescence detectors were used to monitor the products from the first stage (CuInS2 core) and second stage (CuInS2 /ZnS core/shell QD). Source: Yashina et al. 2016 [115]. Reproduced with permission of Royal Society of Chemistry.

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots

Normalized emission

100°C 120°C 140°C 160°C 180°C 200°C 220°C 240°C

125 FWHM (nm)

CIS/ZnS

120 115 110 105

400

600

100

1000

Absorbance

150

200

250

Temperature (°C)

Wavelength (nm)

(a)

(b)

100–220°C

400 (c)

800

500

600

Wavelength (nm)

700

800

100°C

240°C

(d)

Figure 4.18 Core/shell CuInS2 /ZnS QDs synthesized in a two-stage droplet-based microfluidic platform. The effect of temperature within the first stage on the (a) emission spectra, (b) FWHM and (c) absorbance spectra of CuInS2 /ZnS QDs, and (d) CuInS2 /ZnS QDs (in toluene) synthesized at different temperatures within the first stage. The image was taken under UV excitation (405 nm). Source: Yashina et al. 2016 [115]. Reproduced with permission of Royal Society of Chemistry.

platform not only enabled precise dosing of reagents but also independently controlled the temperature for CuInS2 core and ZnS shell formation. This feature allowed for direct tuning of the optical properties of the CuInS2 /ZnS QDs by changing the diameter of CuInS2 cores at different reaction temperatures. As shown in Figure 4.18, PL emission maxima ranged from 580 to 760 nm and could be controlled by increasing the temperature of the first stage from 100 to 240 ∘ C. Figure 4.18d shows an image of CuInS2 /ZnS QDs synthesized at different temperatures in the first stage. The authors also investigated the effect of temperature on ZnS shell formation and subsequent PL properties of CuInS2 /ZnS QDs. Although they tested various temperatures between 150 and 230 ∘ C for the second stage, it was found that this did not affect the band gap of CuInS2 /ZnS QDs to an appreciable degree. 4.4.3

Multistep Synthesis of QDs in Segmented Flow

As an alternative to actively injecting new precursor solutions into existing droplets, it is possible to perform programmed merging of preformed droplets containing different reaction precursors. An example strategy involves the introduction of new droplets into a primary flow (via a side channel) followed by merging with incumbent droplets (Figure 4.19a). This approach requires special

135

136

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

Droplet fusion via electrocoalescence +V

–V (a)

Direct injection

(b)

Droplet phase

Droplet phase

Droplet phase

Gas Carrier phase (c)

Figure 4.19 Three methods for delivering reagents into flowing droplets: (a) New reagent droplets can be introduced into a main droplet flow from the side and then merged with the native droplets by electrocoalescence; (b) direct injection of new reagent into the droplets directly from a side channel using a continuous laminar stream; (c) three-phase direct injection, where an organic solvent, a fluorous carrier, and an inert spacer gas generate a three-phase flow, and the new reagent is injected into the droplets from a side channel with a continuous flow. Source: Nightingale et al. [114]. Reproduced with permission from Nature Publishing Group.

channel architectures to bring the two droplets together [66, 116] and external electric fields to lower the interfacial tension between droplets [117–119]. Direct injection (Figure 4.19b) is generally the preferred method for droplet dosing because it does not require special channel architectures. However, there are two key challenges associated with this technique. First, it can often be challenging to obtain uniform spacing between droplets, and second, the injection volume must be minimized to prevent the (undesired) formation of new droplets from the injection stream. Both these issues can restrict the practical application of direct injection. In an attempt to overcome the abovementioned issues, Nightingale and colleagues at Imperial College London [114] developed a three-phase droplet reactor (Figure 4.19c). The three-phase droplet flow was generated by pumping an organic solvent (ODE), a fluorous carrier (PFPE), and an inert spacer gas (argon) into a 3-to-1 flow junction, which was then fed into a PTFE capillary. Here, a

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots

gas phase was introduced to maintain uniform droplet spacing and eliminate the possibility of generating new droplets from the side stream. Additional reagents could be injected into the flowing droplets at a T-junction, which was formed by inserting a fused silica capillary into the PTFE tubing. The three-phase droplet reactor was used to synthesize CdSe QDs and ensure sustained particle growth (Figure 4.20). After initial droplet generation, the droplets traveled to a reaction zone for initial CdSe QD nucleation and growth within an oil bath at 233 ∘ C. The reaction stage incorporated four T-junctions for droplet dosing via direct injection. After each dosing, the three-phase stream was pumped into a 200 ∘ C heating zone to promote further growth. Once QD Reaction stage 1 QR1

2

Flow cell

3

QR2

4

5 QR5

QR4

QR3

QG QC 1 m coil 233 °C

2 m coils 200 °C

600 Normalized absorbance and emission intensity

Absorption (after final stage of multi-addition) λmax (nm)

Emission

25 570

Single addition (control)

20

560

(b)

500

550 600 Wavelength (nm)

600

650

700

0.8 Absorbance

3

570

2

3 Reaction stage

4

5

1

4

580

15 1

(c) 5

590 λmax (nm)

30

580

550

2

560

Single addition (control)

0.6 Multiaddition

0.4 0.2

550

(d)

35

Multi-addition

450

540

45 40

590

Calculated QD volume (nm3)

(a)

Multi-addition stage 1 0

30

60 Time (s)

0 400

90

(e)

450

500 550 Wavelength (nm)

600

650

Figure 4.20 CdSe QDs synthesized by multistep droplet dosing in a three-phase droplet microreactor. (a) A schematic of a five-stage reactor for CdSe QD synthesis; (b) normalized absorption and PL spectra of the QDs after four reagent doses; (c) peak emission wavelengths for multistep and single-step precursor addition syntheses; (d) peak emission wavelengths for each of the five reaction stages during multistep synthesis; and (e) absorption spectra of the QDs from both multistep and single-step precursor addition syntheses. Source: Nightingale et al. [114]. Reproduced with permission from Nature Publishing Group.

137

138

4 Microfluidic Tools for the Synthesis of Bespoke Quantum Dots

growth was complete for each dose, the stream was cooled to room temperature before the next dosing operation. As shown in Figure 4.20d, the peak emission wavelength of the CdSe QDs exhibited a red shift with each successive growth stage, with the peak wavelength shifting from 552 to 595 nm after four droplet dosing and growth steps. To compare their multistep synthesis approach to a single injection procedure, they conducted an experiment using the same total amount of reagents but only incorporating a single addition of precursors. As shown in Figure 4.20c, the peak emission wavelength from a single addition was also red-shifted, but to a significantly lesser degree (556–571 nm). 4.4.4

Nucleation and Growth Studies of Quantum Dots

In addition to their versatility in QD synthesis, droplet microfluidic platforms are also highly effective in mechanistic studies of QD nucleation and growth when integrated with in situ spectroscopic characterization tools. Such an approach offers vital insight into the underlying reaction mechanisms at both short and long timescales without interfering with the reaction [120]. The three most common in situ spectroscopic methods are fluorescence, absorbance, and X-ray scattering (Figure 4.21). In situ fluorescence spectroscopy is an effective and powerful characterization method for QDs as the emission peak wavelength and FWHM indicate the average size and size distribution, respectively, of the synthesized QDs [120]. Chan et al. [121] were the first to report a chip-based microfluidic reactor integrating a fluorescence detector, which they used to monitor CdSe QDs. Since this time, in situ fluorescence spectroscopy has commonly been integrated into microfluidic QD synthesis setups. Additionally, there are a few accounts describing the implementation of in situ spectroscopy to study QD nucleation and growth. An illustrative example is a report by Lignos et al. from ETH Zürich [50, 122], where they integrated both online fluorescence and absorbance spectroscopy into their droplet microfluidic platform. Figure 4.22a shows their microfluidic setup with integrated detection modules for performing online kinetic studies. The platform was fixed on an axle and a rotating movable stage, allowing measurements at different points along the reaction tubing. A light-emitting diode (LED) was used to PL

Abs

X-ray Nanocrystals

Figure 4.21 Three different in situ spectroscopic methods for semiconductor nanocrystal characterization. Source: Lignos et al. 2017 [50]. Reproduced with permission of American Chemical Society.

4.4 Segmented-Flow Microfluidic Synthesis of Quantum Dots PL optical system

Abs optical system

Reactor

Stages

(b)

Normalized intensity (a.u.)

1.2 1 0.8 0.6 0.4 0.2 0

(c)

800

1000 Wavelength (nm)

1200

t = 0.4 s t = 0.6 s t = 0.8 s t = 1.0 s t = 1.2 s t = 1.4 s t = 1.6 s t = 1.8 s t = 2.0 s t = 2.2 s t = 2.4 s t = 2.6 s t = 2.8 s t = 3.0 s t = 3.2 s t = 3.4 s t = 3.6 s t = 3.8 s

T = 120°C T = 125°C T = 130°C T = 135°C T = 140°C T = 145°C T = 150°C

Absorbance

(a)

600

700 800 900 1000 1100 Wavelength (nm)

(d)

Figure 4.22 Online kinetic studies of PbS QDs. (a) an image of a microfluidic platform incorporating online absorbance and PL modules; (b) a photograph of a heating rod wrapped in capillary tubing under LED excitation; (c) PL spectrum of PbS QDs at various time points after reaction initiation; and (d) absorption spectra of PbS QDs synthesized at various temperatures. Source: Lignos et al. 2017 [50]. Reproduced with permission of American Chemical Society.

excite QDs in droplets, with emission and absorption spectra being recorded at the same time. Such a microfluidic platform could be used for online kinetic analysis of ultrasmall PbS QDs (with absorption maxima between 650 and 750 nm). Figure 4.22c shows the real-time evolution of the emission spectra at 120 ∘ C. Interestingly, it could be seen that the position of the PL peak remained constant until 1200 ms, after which the PL peak shifted to longer wavelengths. This indicated that the size of the PbS QDs did not increase significantly during the first 1200 ms, with most of the growth occurring after this period. Figure 4.22d shows a series of absorbance spectra as a function of temperature, highlighting the maxima shift from 680 to 860 nm. The data demonstrated that the PbS QD diameters increased from 2.35 to 2.9 nm as the temperature increased from 120 to 150 ∘ C [123]. Overall, results from this real-time absorption and PL study indicated that the mechanism of PbS QD formation comprises two steps, the first step being nucleation, where there is fast formation of PbS NPs of constant size (90% of the original signals after two weeks. 6.4.3

Other Nanosensors

Besides the commonly used carbon and gold nanomaterials integrated on paper or paper hybrid devices, there are a few other metal-based nanosensors used for POC detection of infectious diseases. For example, the Liu group [64] fabricated a microfluidic paper-based origami platform (origami μPAD) for the detection of an HIV biomarker, p24 antigen. Zinc oxide nanowires (ZnO NWs) were integrated on the origami μPAD with electrochemical impedance spectroscopy (EIS) biosensing approach. Basically, the device consisted of two cellulose paper layers and was assembled via taping and origami, as shown in Figure 6.9. One of the paper layers contained a hydrophilic detection zone via wax printing and two electrodes via screen printing (i.e. a carbon CE and a silver/silver chloride RE). The other piece of paper comprised carbon WE, where the ZnO NWs were directly synthesized in situ via a hydrothermal process. Sample solutions were spotted on the WE with surface modification and incubated on the unfolded μPAD, allowing the immobilization of the probe antibody. The device was then folded, and the electron mediator solution was added to the test zones. With the binding of the target molecules, the efficiency of the electro–solution interface would be decreased and the increasing electron transfer resistor (Ret ) value was obtained as the readout signal in the EIS label-free immunoassays. The detection of HIV p24 antigen was achieved on the low-cost platform and the sensitivity

Wax Text zone CE (C) RE (Ag/AgCl) Contact (Ag/AgCl) WE (C) PDMS barrier ZnO NWs (a)

Folding line

Cookie paper Double-sided adhesive tape

(b)

ZnO NWs

1 cm

(i) Adding sample solution (c)

(ii) Washing and blotting

(iii) Peeling off cookie paper and folding

(iv) Adding electron mediator and testing

Figure 6.9 The origami μPAD for HIV biomarker detection. (a) Schematic of the components (left) and the assembly (right) of origami μPAD. (b) Photographs of an origami μPAD showing the top view and the inside view (before origami). (c) Schematic of the assay operations of the origami μPAD. Source: Adapted with permission from Liu and coworkers [64]. Copyright 2016 John Wiley & Sons.

202

6 Paper and Paper Hybrid Microfluidic Devices

was enhanced by the integration of ZnO NWs with large surface areas, resulting in the LOD of 300 fg/ml.

6.5 Summary and Outlook Because of its low cost and ease of fabrication, enormous paper-based microfluidic devices have been developed for infectious disease diagnosis, especially for resource-poor settings, such as rural areas and developing nations where people cannot afford costly instrumentation. It has been demonstrated that these microfluidic devices have emerged as promising diagnostic platforms to improve human health in low-resource settings. Although this paper has many advantages, it is not “perfect”. As mentioned in the Fabrication of Paper Hybrid Microfluidic Devices section, each device substrate has its own advantages and limitations. Since the advent of the concept of paper hybrid microfluidic devices, many new paper hybrid systems have been developed [42]. Researchers have demonstrated new benefits that used to be impossible in nonhybrid systems, from paper hybrid microfluidic platforms. In the near future, more paper hybrid devices will be developed and they will play an increasingly important role. Although paper-based devices are inexpensive, they need certain detectors that are often bulky and costly, which limits the applications of paper-based devices in low-resource settings. Several detection methods have been used in paper-based microfluidic analysis such as colorimetric, fluorescence, and electrochemical detection. Colorimetric detection is highly compatible with the nature of low-cost assay in low-resource settings, but the sensitivity and quantitation are often compromised. Electrochemical detection is also desirable for paper-based microfluidic devices, but the bulky and expensive potentiostat is a bottleneck. Although portable potentiostats are commercially available, the cost is still fairly high. In the last decades, smartphone technologies have dramatically developed. Powerful CPU chips have been used in smartphones. An increasing number of new applications and add-ons have enabled those smartphones to offer more and more functions for monitoring personal health status. Therefore, we believe that the combination of smartphone technologies or other inexpensive electronic readers (e.g. glucometers) with paper-based microfluidic devices will cause great impacts on infectious diseases diagnosis in the near future.

Acknowledgment We would like to acknowledge the financial support from the National Institute of Allergy and Infectious Disease of the NIH (R21AI107415), the U.S. NSF-PREM program (DMR 1827745), the Philadelphia Foundation, and the Medical Center of the Americas Foundation. Financial support from the National Institute of General Medical Sciences of the NIH (SC2GM105584), the NIH RCMI Pilot

References

grant, the University of Texas at El Paso for the IDR Program, and University of Texas (UT) System for the STARS award is also greatly acknowledged.

References 1 Yager, P., Domingo, G.J., and Gerdes, J. (2008). Point-of-care diagnostics for

global health. Annu. Rev. Biomed. Eng. 10 (1): 107–144. 2 Lozano, R., Naghavi, M., Foreman, K. et al. (2012). Global and regional

3 4 5 6 7 8 9 10

11 12 13

14

15 16

17

mortality from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 380 (9859): 2095–2128. Nelson, K.E. and Williams, C.M. (2013). Infectious Disease Epidemiology. Jones & Bartlett Publishers. Harris, M. and Reza, J. (2012). Global Report for Research on Infectious Diseases of Poverty. World Health Organization. Wild, D. (2013). The Immunoassay Handbook: Theory and Applications of Ligand Binding, ELISA and Related Techniques. Newnes. Innis, M.A., Gelfand, D.H., Sninsky, J.J., and White, T.J. (2012). PCR protocols: a guide to methods and applications. Academic Press. Jaye, D.L., Bray, R.A., Gebel, H.M. et al. (2012). Translational applications of flow cytometry in clinical practice. J. Immunol. 188 (10): 4715–4719. Urdea, M., Penny, L.A., Olmsted, S.S. et al. (2006). Requirements for high impact diagnostics in the developing world. Nature 444 (S1): 73–79. Hu, J., Wang, S., Wang, L. et al. (2014). Advances in paper-based point-of-care diagnostics. Biosens. Bioelectron. 54: 585–597. Shen, F., Li, X., and Li, P.C. (2014). Study of flow behaviors on single-cell manipulation and shear stress reduction in microfluidic chips using computational fluid dynamics simulations. Biomicrofluidics 8 (1): 014109. Li, X.J. and Zhou, Y. (2013). Microfluidic Devices for Biomedical Applications. Elsevier. Liang, J., Wang, Y., and Liu, B. (2012). Paper-based fluoroimmunoassay for rapid and sensitive detection of antigen. RSC Adv. 2 (9): 3878–3884. Li, X., Chen, Y., and Li, P.C. (2011). A simple and fast microfluidic approach of same-single-cell analysis (SASCA) for the study of multidrug resistance modulation in cancer cells. Lab Chip 11 (7): 1378–1384. Li, X., Ling, V., and Li, P.C. (2008). Same-single-cell analysis for the study of drug efflux modulation of multidrug resistant cells using a microfluidic chip. Anal. Chem. 80 (11): 4095–4102. Chen, H., Li, X., Wang, L., and Li, P.C. (2010). A rotating microfluidic array chip for staining assays. Talanta 81 (4–5): 1203–1208. Li, X., Valadez, A.V., Zuo, P., and Nie, Z. (2012). Microfluidic 3D cell culture: potential application for tissue-based bioassays. Bioanalysis 4 (12): 1509–1525. Jokerst, J.C., Adkins, J.A., Bisha, B. et al. (2012). Development of a paper-based analytical device for colorimetric detection of select foodborne pathogens. Anal. Chem. 84 (6): 2900–2907.

203

204

6 Paper and Paper Hybrid Microfluidic Devices

18 Yu, L., Li, C.M., Liu, Y. et al. (2009). Flow-through functionalized PDMS

19

20

21

22

23

24

25

26

27

28

29

30

31

32

microfluidic channels with dextran derivative for ELISAs. Lab Chip 9 (9): 1243–1247. Dou, M., Dominguez, D.C., Li, X. et al. (2014). A versatile PDMS/paper hybrid microfluidic platform for sensitive infectious disease diagnosis. Anal. Chem. 86 (15): 7978–7986. Dou, M., Sanjay, S.T., Dominguez, D.C. et al. (2017). Multiplexed instrument-free meningitis diagnosis on a polymer/paper hybrid microfluidic biochip. Biosens. Bioelectron. 87: 865–873. Lee, Y.-F., Lien, K.-Y., Lei, H.-Y., and Lee, G.-B. (2009). An integrated microfluidic system for rapid diagnosis of dengue virus infection. Biosens. Bioelectron. 25 (4): 745–752. Li, X., Xue, X., and Li, P.C. (2009). Real-time detection of the early event of cytotoxicity of herbal ingredients on single leukemia cells studied in a microfluidic biochip. Integr. Biol. 1 (1): 90–98. Li, X., Huang, J., Tibbits, G.F., and Li, P.C. (2007). Real-time monitoring of intracellular calcium dynamic mobilization of a single cardiomyocyte in a microfluidic chip pertaining to drug discovery. Electrophoresis 28 (24): 4723–4733. Medina-Sánchez, M., Miserere, S., Morales-Narváez, E., and Merkoçi, A. (2014). On-chip magneto-immunoassay for Alzheimer’s biomarker electrochemical detection by using quantum dots as labels. Biosens. Bioelectron. 54: 279–284. Zhang, X., Zou, Y., An, C. et al. (2014). A miniaturized immunosensor platform for automatic detection of carcinoembryonic antigen in EBC. Sens. Actuators, B 205: 94–101. Sardesai, N.P., Kadimisetty, K., Faria, R., and Rusling, J.F. (2013). A microfluidic electrochemiluminescent device for detecting cancer biomarker proteins. Anal. Bioanal. Chem. 405 (11): 3831–3838. Mohammed, M. and Desmulliez, M. (2014). Autonomous capillary microfluidic system with embedded optics for improved troponin I cardiac biomarker detection. Biosens. Bioelectron. 61: 478–484. Matos Pires, N.M. and Dong, T. (2013). Microfluidic biosensor array with integrated poly(2,7-carbazole)/fullerene-based photodiodes for rapid multiplexed detection of pathogens. Sensors 13 (12): 15898–15911. Jena, R.K., Yue, C., and Lam, Y. (2012). Micro fabrication of cyclic olefin copolymer (COC) based microfluidic devices. Microsyst. Technol. 18 (2): 159–166. Yu, H., Tor, S., and Loh, N. (2014). Rapid bonding enhancement by auxiliary ultrasonic actuation for the fabrication of cyclic olefin copolymer (COC) microfluidic devices. J. Micromech. Microeng. 24 (11): 115020. Su, M., Ge, L., Ge, S. et al. (2014). Paper-based electrochemical cyto-device for sensitive detection of cancer cells and in situ anticancer drug screening. Anal. Chim. Acta 847: 1–9. Hu, J., Wang, L., Li, F. et al. (2013). Oligonucleotide-linked gold nanoparticle aggregates for enhanced sensitivity in lateral flow assays. Lab Chip 13 (22): 4352–4357.

References

33 Liu, P., Li, X., Greenspoon, S.A. et al. (2011). Integrated DNA purification,

34

35 36

37 38 39

40 41

42

43 44 45

46 47

48

49

50

PCR, sample cleanup, and capillary electrophoresis microchip for forensic human identification. Lab Chip 11 (6): 1041–1048. Han, Y.L., Wang, W., Hu, J. et al. (2013). Benchtop fabrication of three-dimensional reconfigurable microfluidic devices from paper–polymer composite. Lab Chip 13 (24): 4745–4749. Wang, S., Inci, F., Chaunzwa, T.L. et al. (2012). Portable microfluidic chip for detection of Escherichia coli in produce and blood. Int. J. Nanomed. 7: 2591. Martinez, A.W., Phillips, S.T., Butte, M.J., and Whitesides, G.M. (2007). Patterned paper as a platform for inexpensive, low-volume, portable bioassays. Angew. Chem. Int. Ed. 46 (8): 1318–1320. Nie, Z., Nijhuis, C.A., Gong, J. et al. (2010). Electrochemical sensing in paper-based microfluidic devices. Lab Chip 10 (4): 477–483. Cheng, C.M., Martinez, A.W., Gong, J. et al. (2010). Paper-based ELISA. Angew. Chem. Int. Ed. 49 (28): 4771–4774. Abe, K., Kotera, K., Suzuki, K., and Citterio, D. (2010). Inkjet-printed paper fluidic immuno-chemical sensing device. Anal. Bioanal.Chem. 398 (2): 885–893. Nogi, M., Iwamoto, S., Nakagaito, A.N., and Yano, H. (2009). Optically transparent nanofiber paper. Adv. Mater. 21 (16): 1595–1598. Zhu, H.L., Fang, Z.Q., Preston, C. et al. (2014). Transparent paper: fabrications, properties, and device applications. Energy Environ. Sci. 7 (1): 269–287. Zuo, P., Li, X., Dominguez, D.C., and Ye, B.C. (2013). A PDMS/paper/glass hybrid microfluidic biochip integrated with aptamer-functionalized graphene oxide nano-biosensors for one-step multiplexed pathogen detection. Lab Chip 13 (19): 3921–3928. Li, X., Tian, J.F., and Shen, W. (2010). Thread as a versatile material for low-cost microfluidic diagnostics. ACS Appl. Mater. Interfaces 2 (1): 1–6. Lin, S.C., Hsu, M.Y., Kuan, C.M. et al. (2014). Cotton-based diagnostic devices. Sci. Rep. 4: 6976. Zhang, L., Zhang, Y., Wang, C.Y. et al. (2014). Integrated microcapillary for sample-to-answer nucleic acid pretreatment, amplification, and detection. Anal. Chem. 86 (20): 10461–10466. Li, X., Tian, J., Garnier, G., and Shen, W. (2010). Fabrication of paper-based microfluidic sensors by printing. Colloids Surf., B 76 (2): 564–570. Apilux, A., Dungchai, W., Siangproh, W. et al. (2010). Lab-on-paper with dual electrochemical/colorimetric detection for simultaneous determination of gold and iron. Anal. Chem. 82 (5): 1727–1732. Martinez, A.W., Phillips, S.T., Carrilho, E. et al. (2008). Simple telemedicine for developing regions: camera phones and paper-based microfluidic devices for real-time, off-site diagnosis. Anal. Chem. 80 (10): 3699–3707. Lu, Y., Shi, W., Qin, J., and Lin, B. (2009). Fabrication and characterization of paper-based microfluidics prepared in nitrocellulose membrane by wax printing. Anal. Chem. 82 (1): 329–335. Cheng, C.M., Martinez, A.W., Gong, J. et al. (2010). Paper-based ELISA. Angew. Chem. 122 (28): 4881–4884.

205

206

6 Paper and Paper Hybrid Microfluidic Devices

51 Dou, M., Sanjay, S.T., Benhabib, M. et al. (2015). Low-cost bioanalysis on

paper-based and its hybrid microfluidic platforms. Talanta 145: 43–54. 52 Bruzewicz, D.A., Reches, M., and Whitesides, G.M. (2008). Low-cost print-

53 54

55

56 57 58

59 60

61

62 63

64

65

66 67 68

ing of poly(dimethylsiloxane) barriers to define microchannels in paper. Anal. Chem. 80 (9): 3387–3392. Abe, K., Suzuki, K., and Citterio, D. (2008). Inkjet-printed microfluidic multianalyte chemical sensing paper. Anal. Chem. 80 (18): 6928–6934. Rosa, A.M., Louro, A.F., Martins, S.A. et al. (2014). Capture and detection of DNA hybrids on paper via the anchoring of antibodies with fusions of carbohydrate binding modules and ZZ-domains. Anal. Chem. 86 (9): 4340–4347. Dungchai, W., Chailapakul, O., and Henry, C.S. (2011). A low-cost, simple, and rapid fabrication method for paper-based microfluidics using wax screen-printing. Analyst 136 (1): 77–82. Olkkonen, J., Lehtinen, K., and Erho, T. (2010). Flexographically printed fluidic structures in paper. Anal. Chem. 82 (24): 10246–10250. Li, X., Tian, J., Nguyen, T., and Shen, W. (2008). based microfluidic devices by plasma treatment. Anal. Chem. 80 (23): 9131–9134. Magro, L., Jacquelin, B., Escadafal, C. et al. (2017). Paper-based RNA detection and multiplexed analysis for Ebola virus diagnostics. Sci. Rep. 7 (1): 1347. Xu, G., Nolder, D., Reboud, J. et al. (2016). Paper-origami-based multiplexed malaria diagnostics from whole blood. Angew. Chem. 128 (49): 15476–15479. Murdock, R.C., Gallegos, K.M., Hagen, J.A. et al. (2017). Development of a point-of-care diagnostic for influenza detection with antiviral treatment effectiveness indication. Lab Chip 17 (2): 332–340. Lei, K.F., Huang, C.H., Kuo, R.L. et al. (2015). Paper-based enzyme-free immunoassay for rapid detection and subtyping of influenza A H1N1 and H3N2 viruses. Anal. Chim. Acta 883: 37–44. Veigas, B., Jacob, J.M., Costa, M.N. et al. (2012). Gold on paper-paper platform for Au-nanoprobe TB detection. Lab Chip 12 (22): 4802–4808. Tsai, T.T., Shen, S.W., Cheng, C.M., and Chen, C.F. (2013). Paper-based tuberculosis diagnostic devices with colorimetric gold nanoparticles. Sci. Technol. Adv. Mater. 14 (4): 044404. Li, X. and Liu, X. (2016). A microfluidic paper-based origami nanobiosensor for label-free, ultrasensitive immunoassays. Adv. Healthc. Mater. 5 (11): 1326–1335. Carrilho, E., Martinez, A.W., and Whitesides, G.M. (2009). Understanding wax printing: a simple micropatterning process for paper-based microfluidics. Anal. Chem. 81 (16): 7091–7095. Li, X., Nie, Z., Cheng, C. et al. (2010). Paper-based electrochemical ELISA. Proceedings Micro Total Analysis Systems, pp. 1487–1489. Li, X., Tian, J., and Shen, W. (2010). Progress in patterned paper sizing for fabrication of paper-based microfluidic sensors. Cellulose 17 (3): 649–659. Li, X., Tian, J., and Shen, W. (2010). Quantitative biomarker assay with microfluidic paper-based analytical devices. Anal. Bioanal. Chem. 396 (1): 495–501.

References

69 Blazkova, M., Javurkova, B., Fukal, L., and Rauch, P. (2011). Immunochro-

70

71

72

73

74

75 76 77

78

79

80

81

82 83

84

matographic strip test for detection of genus Cronobacter. Biosens. Bioelectron. 26 (6): 2828–2834. Choi, J.R., Hu, J., Tang, R. et al. (2016). An integrated paper-based sample-to-answer biosensor for nucleic acid testing at the point of care. Lab Chip 16 (3): 611–621. Pereira, D.Y., Chiu, R.Y.T., Zhang, S.C.L. et al. (2015). Single-step, paper-based concentration and detection of a malaria biomarker. Anal. Chim. Acta 882: 83–89. Khan, M.S., Pande, T., and van de Ven, T.G.M. (2015). Qualitative and quantitative detection of T7 bacteriophages using paper based sandwich ELISA. Colloids Surf., B 132: 264–270. Fenton, E.M., Mascarenas, M.R., López, G.P., and Sibbett, S.S. (2008). Multiplex lateral-flow test strips fabricated by two-dimensional shaping. ACS Appl. Mater. Interfaces 1 (1): 124–129. Martinez, A.W., Phillips, S.T., and Whitesides, G.M. (2008). Three-dimensional microfluidic devices fabricated in layered paper and tape. Proc. Natl. Acad. Sci. U.S.A. 105 (50): 19606–19611. Xia, Y. and Whitesides, G.M. (1998). Soft lithography. Angew. Chem. Int. Ed. Engl. 37 (5): 550–575. Liu, X., Mwangi, M., Li, X. et al. (2011). Paper-based piezoresistive MEMS sensors. Lab Chip 11 (13): 2189–2196. Martinez, A.W., Phillips, S.T., Whitesides, G.M., and Carrilho, E. (2009). Diagnostics for the Developing World: Microfluidic Paper-Based Analytical Devices. ACS Publications. Dou, M., Sanjay, S.T., Dominguez, D.C. et al. (2017). A paper/polymer hybrid CD-like microfluidic SpinChip integrated with DNA-functionalized graphene oxide nanosensors for multiplex qLAMP detection. Chem. Commun. 53 (79): 10886–10889. Sanjay, S.T., Dou, M.W., Sun, J.J., and Li, X.J. (2016). A paper/polymer hybrid microfluidic microplate for rapid quantitative detection of multiple disease biomarkers. Sci. Rep. 6: 30474. Vilela, D., González, M.C., and Escarpa, A. (2012). Sensing colorimetric approaches based on gold and silver nanoparticles aggregation: chemical creativity behind the assay. A review. Anal. Chim. Acta 751: 24–43. Qu, W., Liu, Y., Liu, D. et al. (2011). Copper-mediated amplification allows readout of immunoassays by the naked eye. Angew. Chem. Int. Ed. 50 (15): 3442–3445. Wu, L. and Qu, X. (2015). Cancer biomarker detection: recent achievements and challenges. Chem. Soc. Rev. 44 (10): 2963–2997. Li, X., Ballerini, D.R., and Shen, W. (2012). A perspective on paper-based microfluidics: current status and future trends. Biomicrofluidics 6 (1): 011301. Gao, Z., Xu, M., Lu, M. et al. (2015). Urchin-like (gold core)@(platinum shell) nanohybrids: a highly efficient peroxidase-mimetic system for in situ amplified colorimetric immunoassay. Biosens. Bioelectron. 70: 194–201.

207

208

6 Paper and Paper Hybrid Microfluidic Devices

85 Fang, W.-F., Chen, W.-J., and Yang, J.-T. (2014). Colorimetric determination

86 87 88

89 90

91

92

93

94

95 96 97

98

99

100

of DNA concentration and mismatches using hybridization-mediated growth of gold nanoparticle probes. Sens. Actuators, B 192: 77–82. Xu, X., Akay, A., Wei, H. et al. (2015). Advances in smartphone-based point-of-care diagnostics. Proc. IEEE 103 (2): 236–247. Xu, X., Wang, X., Hu, J. et al. (2018). A smartphone-based on-site nucleic acid testing platform at point-of-care settings. Electrophoresis 40: 914–921. Wang, W., Wu, W.-Y., and Zhu, J.-J. (2010). Tree-shaped paper strip for semiquantitative colorimetric detection of protein with self-calibration. J. Chromatogr. A 1217 (24): 3896–3899. Connelly, J.T., Rolland, J.P., and Whitesides, G.M. (2015). “Paper machine” for molecular diagnostics. Anal. Chem. 87 (15): 7595–7601. Shetty, P., Ghosh, D., Singh, M. et al. (2016). Rapid amplification of Mycobacterium tuberculosis DNA on a paper substrate. RSC Adv. 6 (61): 56205–56212. Wang, X., Theodore, M.J., Mair, R. et al. (2012). Clinical validation of multiplex real-time PCR assays for detection of bacterial meningitis pathogens. J. Clin. Microbiol. 50 (3): 702–708. Maxwell, E.J., Mazzeo, A.D., and Whitesides, G.M. (2013). Paper-based electroanalytical devices for accessible diagnostic testing. MRS Bull. 38 (4): 309–314. Ahmed, M.U., Hossain, M.M., Safavieh, M. et al. (2016). Toward the development of smart and low cost point-of-care biosensors based on screen printed electrodes. Crit. Rev. Biotechnol. 36 (3): 495–505. Su, W., Gao, X., Jiang, L., and Qin, J. (2015). Microfluidic platform towards point-of-care diagnostics in infectious diseases. J. Chromatogr. A 1377: 13–26. Li, X., Scida, K., and Crooks, R.M. (2015). Detection of hepatitis B virus DNA with a paper electrochemical sensor. Anal. Chem. 87 (17): 9009–9015. Parolo, C. and Merkoci, A. (2013). Paper-based nanobiosensors for diagnostics. Chem. Soc. Rev. 42 (2): 450–457. Rohrman, B.A., Leautaud, V., Molyneux, E., and Richards-Kortum, R.R. (2012). A lateral flow assay for quantitative detection of amplified HIV-1 RNA. PLoS One 7 (9): e45611. Baptista, F.R., Belhout, S.A., Giordani, S., and Quinn, S.J. (2015). Recent developments in carbon nanomaterial sensors. Chem. Soc. Rev. 44 (13): 4433–4453. Thaxton, C.S., Georganopoulou, D.G., and Mirkin, C.A. (2006). Gold nanoparticle probes for the detection of nucleic acid targets. Clin. Chim. Acta 363 (1–2): 120–126. Zhao, W., Brook, M.A., and Li, Y. (2008). Design of gold nanoparticle-based colorimetric biosensing assays. ChemBioChem 9 (15): 2363–2371.

References

101 Magro, L., Escadafal, C., Garneret, P. et al. (2017). Paper microfluidics for

nucleic acid amplification testing (NAAT) of infectious diseases. Lab Chip 17 (14): 2347–2371. 102 Soo, P.C., Horng, Y.T., Hsueh, P.R. et al. (2006). Direct and simultaneous identification of Mycobacterium tuberculosis complex (MTBC) and Mycobacterium tuberculosis (MTB) by rapid multiplex nested PCR-ICT assay. J. Microbiol. Methods 66 (3): 440–448.

209

211

7 Biological Diagnosis Based on Microfluidics and Nanotechnology Navid Kashaninejad 1,2 , Mohammad Yaghoobi 3 , Mohammad Pourhassan-Moghaddam 1,4 , Sajad R. Bazaz 1,4 , Dayong Jin 2,4 , and Majid E. Warkiani 1,4,5 1 University of Technology Sydney, School of Biomedical Engineering, Faculty of Engineering and Information Technology, Broadway, Sydney, NSW 2007, Australia 2 University of Technology Sydney, School of Mathematical and Physical Sciences, Faculty of Science, Broadway, Sydney, NSW 2007, Australia 3 Sharif University of Technology, Department of Mechanical Engineering, Azadi, Tehran 11155, Iran 4 University of Technology Sydney, Institute for Biomedical Materials and Devices (IBMD), Faculty of Science, Broadway, Sydney, NSW 2007, Australia 5 Sechenov University, Institute of Molecular Medicine, Pyrogovskaya, Moscow 119991, Russia

7.1 Introduction Early diagnosis of diseases is of paramount importance, especially for managing chronic and asymptomatic illnesses. As such, the synergistic effect of combining nanotechnology with microfluidics can significantly improve biological diagnosis. Both organic and nonorganic nanoparticles have been used for clinical imaging and disease diagnosis [1]. Target molecules can interact with these nanoparticles for ultrasensitive detection of some in vivo and in vitro biological changes. The emergence of microfluidics with the capability of integrating various laboratory processes on a single chip, known as micro total analysis systems (μTAS) or lab-on-a-chip [2–11], has the potential to revolutionize the field of biological diagnosis. One strategy to diagnose a disease is to use biosensors for the aim of detecting biochemical analytes – small molecules, proteins, cardiac, or cancer biomarkers – in patients’ blood, urine, or other biological samples [12]. Numerous techniques can be used to create detectable signals for measuring the concentration or relative quantity of a specific bioanalyte [13]. For this purpose, optical, electrochemical, calorimetric, and piezoelectric transducers have been explored, extensively. Optical biosensors have gained attention in biosensing applications where they are used to study the different properties of targets/analytes and are useful in real-time and parallel detection assays with high precision [14, 15]. Among the numerous types of optical detection techniques, fluorescent nanomaterial-based approaches are broadly used in biological, medical, and drug discovery testing [16]. However, not all fluorescent nanomaterials provide Nanotechnology and Microfluidics, First Edition. Edited by Xingyu Jiang. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.

212

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

robust, stable, or long-term luminosity. They mostly suffer from photobleaching, low quantum yields (QYs), and narrow emission spectra. The main problems associated with fluorescence methods can be addressed by using semiconductor quantum dots (Qdots) possessing unique optical properties. Their surface is capable of being modified with various biomolecules making them selective to different analytes [17, 18]. They can be then used in bulk solution assays (as colloidal suspension) or solid-phase assays (as immobilized on the surface). Upconversion nanoparticles (UCNPs) are another type of fluorescent nanomaterials that have been used in photoluminescent (PL) detection. UCNPs can be used for bioimaging, biosensing, and drug delivery [19]. The higher excitation wavelength of these nanoparticles adds a multitude of advantages to this type of nanoparticle. The last generation of fluorescent nanomaterials, namely carbon-based Qdots or simply “biodots,” has generated interest because of their exceptional properties that are useful in the biomedical imaging [20]. Their properties include high QY, water bioavailability, multicolor wavelength emission, and exceptional biocompatibility. Innovative approaches of nanobiosensing using fluorescent nanomaterials are performed in microfluidic devices. Moreover, because of the movement of flow, the nonspecific deposition of these nanoparticles is reduced, and their reaction rates with targets are increased in microfluidic-based assays because of the nanoscale mixing [21]. Microfluidics is an enabling technology that can be employed in numerous biological studies because of its high-throughput, precision, sample efficiency, portability, and low-cost production [13, 21–26]. Three types of microfluidics used in bioanalysis are continuous, droplet-based, and digital microfluidics (DMFs) [17, 27], each of which has its advantages and disadvantages. Integrated microfluidic devices combined with multiplexed immunoassays provide an opportunity for this technology to facilitate point-of-care (POC) diagnostics [28] for diseases such as cancer [29], malaria [30], or some other abnormalities [31]. Herein, we first describe the functions of Qdots, UCNPs, and biodots; afterward, the applications of nanoparticles-based microfluidics for disease diagnosis will be highlighted. Further, we expand upon the new trends in biological diagnosis platforms by discussing two emerging formats of microfluidics: digital and paper-based microfluidics.

7.2 Quantum Dot-Based Microfluidic Biosensor for Biological Diagnosis Qdots are nano-sized particles with binary compounds of elements with II–IV, III–V, and IV–VI number of valence electrons and semiconductor materials that present many advantages over fluorescent proteins and organic dyes. Their emission spectra are dependent on their size and excitation wavelength, ranging from ultraviolet to infrared. It is possible to excite multiple Qdots only by single wavelength [32], which makes them suitable for multiplexed detection assays. In other words, Qdots absorption spectra cover a broad span of wavelengths. They have a high QY and long fluorescent lifetime suitable for long-term biological

7.2 Quantum Dot-Based Microfluidic Biosensor for Biological Diagnosis

assays. Photostability and resistance to photobleaching are also other essential characteristics of Qdots [33]. Moreover, they can be used as donors and sometimes as acceptors in energy transfer-based assays because of their unique optical properties. Fluorescence resonance energy transfer (FRET) is a method for measuring distances as small as 1 nm up to 20 nm. In this phenomenon, the nonradiative energy transfer between two fluorescent components (called donor and acceptor) is sensitive to the distance between (d) them where this energy is inversely proportional to the d6 of this distance. The donor needs to be excited by an external light source, and acceptor’s absorption and donor’s emission range should have an overlap. Beyond the range of 1–20 nm, no energy transfer occurs [17]. Also, there are several other methods of energy transfer such as chemiluminescence resonance energy transfer (CRET), which have been described in the literature [32]. Qdots can serve as donor or acceptor to improve the efficiency of energy transfer, implementing all aforementioned optical characteristics [34]. Nevertheless, the distance itself is not always the case, especially in biorecognition events. It has been mentioned that if the distance is less than 20 nm, energy transfer will take place. In this range, a fluorescent-labeled bioanalyte is coupled to a biofunctionalized Qdot. Therefore, the bioanalyte is detected with an optical detector. This optical detection is the most common application of Qdots in biosensing. 7.2.1

Qdot-Based Disease Diagnosis Using Microfluidics

In the bioanalysis with Qdots, these nanoparticles are conjugated with biomolecules for the detection of analytes. In the case of a disease, the target analytes can be viruses, antigens, and cancer biomarkers, whose concentration is also of interest. As such, in a study, Hu et al. produced stable antibody-conjugated-Qdots in the aqueous phase and used them in a microfluidic chip for multiplexed detection of carcinoma embryonic antigen (CEA) and α-fetoprotein (AFP) in both sandwich and reversed-phase immunoassays (Figure 7.1) [35], indicating an excellent capability for POC cancer diagnosis. In addition to small-volume usage of sample and reagent as well as reducing the time-to-result provided by microfluidics, this technology allows for better manipulation of flow in order to enhance the capture efficiency of targets onto the biofunctionalized reaction sites. This opportunity is used by Ng et al. [36] for the detection of QD-labeled SkBr3 (breast cancer) or fluorescein isothiocyanate (FITC)-labeled Colo205 (colon cancer) cells by capturing them onto both nanoporous silica anti-EpCAM (epithelial cell adhesion molecule)-coated substrate and the polydimethylsiloxane (PDMS)-stamped areas of this plate inked with antibody solution (Figure 7.2). Cancer is not the only field of applicability of the Qdot-based microfluidic assays. Infectious disease can also be detected or monitored using such assays. Klostranec et al. [20] proposed an IgG (goat antimouse)-conjugated Qdot barcode microbead sandwich assay for the detection of HBV (hepatitis B virus), HCV (hepatitis C virus), and human immunodeficiency virus (HIV) antigens. A spiked human serum with corresponding antibodies was incubated with the

213

4

3

5

6

200

Detection of CEA by QD-IgG Detection of AFP by QD-IgG Detection of CEA by Cy3-IgG

Fluorescent intensity (a.u.)

2

Fluorescent intensity (a.u.)

250

1

150 100 50 0

150 100 50 0

101 102 103 104 105 106 107 108 Antigen concentration (fM)

(c)

Detection of CEA in human serum by QD-IgG

200

(d)

101 102 103 104 105 106 107 108 109 CEA concentration (fM)

(a)

(e) 1

2

3

4

5

6

CEA

Sandwich immunoassay red-colored Qds

(b)

Reverse-phase immunoassay CEA

AFP

AFP

green-colored Qds

anti-CEA mAb

anti-AEP mAb

goat anti-mouse IgG

anti-CEA pAb

anti-AEP pAb

CEA + AFP

(f)

Figure 7.1 (a) Illustration of the process of making a two-dimensional array of immobilized protein layers. (b) Schematic of sandwich and reversed-phase immunoassay with CEA and AFP antibodies and antigens. (c) The dose–response calibration curve for CEA and AFP based on the QD-IgG probes and microfluidic network chip. The measured concentration ranges from 25 fM to 25 nM with limit of detection (LOD) as low as 250 fM (S/N > 3) for both targets. (d) CEA concentration in human serum ranging from 25 fM to 250 nM and LOD was 2.5 pM. The presence of serum albumin slightly hinders antibody–antigen reaction, but still, the tests show better sensitivity in comparison to PBS buffer detection. (e) Three different types of Qdots with green, orange, and red colors excited only under UV illumination for a CEA concentration of 25 pM. (f ) The multiplexed ability of Qdots with red and green emission colors. Horizontal channels associate with CEA, AFP, and CEA-AFP mixtures from top to bottom, respectively. The perpendicular microchannels 1-3 are filled with red-colored emitting QD-IgG and 4–6 with green-colored emitting QD-IgG. Source: Hu et al. 2009 [35]. Reproduced with permission of American Chemical Society.

Inlet

Inking

PDMS microfluidic channel 20 mm 3 mm 5 mm

PDMS stamp

Labeling

100 μm

z PDMS stamp

Stamp

Stamping

x

(d) y

Nanoporous silica substrate

20 mm

Nanoporous silica

SkBr3 or Colo205 cell

silica su

bstrate

x

Nanoporous silica

(e) Shake

Cell suspension

Transferred antibodies

Nanoporous silica

(a) Vortex

5 mm

14

250

12

200

10

6

Stamped antiEpCAM

(b)

(b) Optimal

150 100 50

Stamped antiEpCAM

10 mm

5 mm

8

Outlet

Grooves

1.6 x10–3 350 x10–4 16 300

z (μm)

Porous

y

Block

18 mm

1% BSA

Inlet

x = 0.3 cm x = 0.8 cm

PDMS stamp

200 μm

PDMS stamp

Outlet

QD625-labeled anti-HER2 or FITC-labeled anti-CK Anit-EpCAM

Cell suspension

Antibody solution

4

Capture Nanoporous silica

(c)

0

(f)

0 –50

x = 3469 μm

2

–100

0

–150 3100 3150 3200 3250 3300 3350 3400 3450 3500 3550 3600 3650 3700 3750 3800

x (μm)

Figure 7.2 (a) Schematic of stamped nanoporous manufacturing process (microcontact printing). (b) Top view of the resulting substrate. (c) Immunofluorescence schematic procedure of detection of cancer cells labeled with Qdots or fluorescent dye. The detection of cells whose nuclei were stained with DAPI was performed by fluorescent microscopy. (d) Side view of the grooved microfluidic device is used to exploit the ability of microfluidics to create optimal sites for labeled cancer cells and (e) top view of the same device. (f ) Optimal site for capturing cells are located where the vertical component of velocity is downward and significantly enough to push them toward the reaction area. This area is shown in green color. Source: Ng et al. 2013 [36]. Reproduced with permission of Elsevier.

216

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

biofunctionalized Qdots. This assay (Figure 7.3) suggests a fast (less than an hour) diagnosis with 50 times greater sensitivity compared to FDA-approved ELISA (enzyme-linked immunosorbent assay) technique, which has many disadvantages, including high costs and lower multiplexing capability. As the polystyrene Qdot barcode flow through the focused laser point, a detection signal is recorded. Then, this output barcode signal is interpreted. The green signal is associated with the AlexaFluor-488, and the peaks indicate the passage of every single bead, while red, yellow, and their combination specify the type of antigen. Among infectious diseases, influenza virus detection is of paramount importance because of its severe health complications and the challenges in the virus classifications. Therefore, a wealth of research has been undertaken to address these issues. Zhang et al. [37] proposed a simultaneous detection of H1N1, H3N2, and H9N2, as well as subtyping of them, using integrated micromagnet field microfluidic chip within 80 min interval and LOD of 10–20 nM concentration of the viruses’ genetic material. The capture probe DNAs (CP-DNAs) were immobilized on the surface of supermagnetic beads (SMB). These conjugated probes were then injected into the integrated microfluidic device to react with the target cDNAs (T-DNAs or DNAs of influenza A viruses) injected from another side. Each sample was injected through different neighboring channels that made it possible to analyze the genetic samples of all three viruses simultaneously. The assay is composed of a sandwich structure composed of SMB-(CP-DNA), T-DNA, and (RP-DNA)-biotin-(SA-Qdots) serving as the substrate, target, and detector units, respectively. An ITO heating electrode controlled the temperature of the assay in order to keep the reaction time of DNA as low as possible. In this assay, the normalized fluorescent intensities of T-DNAs in 1 μM solutions were optimized under 45 ∘ C, and the DNA hybridization reaction time was 25 minutes. The concentration of H1N1 and H9N2 was a linear function of fluorescent intensity corresponding to 1–150 nM of the target, while H3N2 concentration was linear between 5 and 150 nM of the target (Figure 7.4). Although there are other studies of influenza, virus detection using Qdots and other microplatforms with more sensitivity and lower LOD [38], low sample usage, and ultrafast manipulation of targets are still exclusive to microfluidics. Another application of magnetic beads and Qdots (CdSd/CdZnS) was demonstrated by Kim et al. [39] for detection of Plasmodium falciparum histidine-rich protein 2 (pfHRP2), the most common biomarker of malaria, with the aid of microdroplets. The pfHRP2 antibody was conjugated on the surface of Qdots and microbeads (MB). The conjugates were then incubated with human serum containing pfHRP2. After separation of MB-Ab-pfHRP2-Ab-Qdot and elution of Qdots, fluorescence measurements were performed to determine the concentration of HRP2. This process was carried out in an array of six consecutive wells in a PDMS-based microchip (Figure 7.5) and was compared to a standard vial-based assay. Because of economic reasons, animal health is also monitored through the diagnosis of their diseases, mainly using novel technologies such as microfluidics. For instance, subclinical ketosis (SCK), a disease that affects the cows’ products

7.2 Quantum Dot-Based Microfluidic Biosensor for Biological Diagnosis

Intensity (a.u.)

1500

QdotB1 Virus: HBV Antigen: HBsAg

1000

500

0 475 525 575 625

675

Wavelength (nm) 2000 Intensity (a.u.)

QdotB2 Virus: HCV

(a)

Antigen: Fluorescence intensity (a.u.)

NSP4

1500 1000 500 0 475 525 575

625 675

Wavelength (nm) 1500 Intensity (a.u.)

QdotB3 Virus: HIV Antigen:

1000 500

gp41 500 550 600 650 Wavelength (nm)

(b)

0 475 525 575

(c)

625 675

Wavelength (nm)

HCV

Intensity (a.u.)

Intensity (a.u.)

HBV

HlV 0

(d)

2

4

6 Time (s)

8

10

2.53 2.55 2.57 2.59 Time (s)

Figure 7.3 (a) A sample of QdotBs with three different emission colors and fluorescent emission spectrums. (b) The orange QdotB is composed of a combination of red and yellow color Qdots, and the bimodal distribution is a sign of that. (c) The three different QdotB microbeads were conjugated with antigens. Antibody-conjugated dye was added to provide a fluorescence detection peak at 520 nm. The other peak in the diagram of intensity–wavelength is related to the corresponding Qdots. (d) An example of the data detected by a photodetector over 10 seconds intervals for green, red, and yellow channels. An HBV antigen-conjugated QdotB crossing the laser spot leads to the peak of green and yellow fluorescent intensity, and red channel signal remains unaffected. Source: Klostranec et al. 2007 [20]. Reproduced with permission of American Chemical Society.

217

7 Biological Diagnosis Based on Microfluidics and Nanotechnology SMBs-CP-H1 Heated area

N

S S

30°

RP-DNA S

SA-QDs

(b)

Channel 2

Channel 1 SMB

10 nM

5 nM

1.2 Normalized intensity

y = 0.1521 + 0.00362x R2 = 0.990

0.6 0.4 0.2

(e)

0

40

80

120

160

Concentration of H1N1 (nmol/l)

T–H1 T–H3 T–H9

40 nM

0.4

(f)

0

40

80

100 nM 0.8

y = 0.24133 + 0.00475x R2 = 0.994

0.8

0.0

RP–H1 RP–H3 RP–H9

80 nM Normalized intensity

(d) 1 nM

Channel 3

CP–H1 CP–H3 CP–H9

SA-QDs

(c)

0.0

Copper wire

MBGM

Channel

(a)

PDMS

ITO heating electrode

Glass

SMBs-CP-H9

0.8

N

T-DNA

N

SMBs-CP-H3

Normalized intensity

218

120

160

Concentration of H1N1 (nmol/l)

0.6

150 nM

y = 0.20264 + 0.00228x R2 = 0.987

0.4 0.2 0.0

(g)

0

40

80

120

160

Concentration of H1N1 (nmol/l)

Figure 7.4 (a) Integrated microfluidic schematic diagram and style of different sample loading. (b) Side view of the microfluidic device. (c) Schematic configuration of virus detection and subtyping in the three juxtaposed sandwich assays. (d) The multiplexed detection of the T-DNA solutions and concentration evaluation for different concentrations of H1, H2, and H9 viruses. (e, f ) linear dependence of H1 and H9 fluorescent intensity on a concentration within 1–150 nM solutions and (g) linear fluorescent intensity dependence on H2 concentration in 5–150 nM range. Source: Zhang et al. 2018 [37]. Reproduced with permission of Elsevier.

and health, is mainly diagnosed by measuring β-hydroxybutyrate (βHBA) concentration in their blood, urine, or milk. Current diagnostic methods are time-consuming and expensive. On the contrary, a microfluidic device coupled with biofunctionalized Qdots developed by Weng et al. [40] showed sensitive and on-farm detection of SCK. The Qdots were modified with adenine dinucleotide (NAD+ ). NAD+ in the presence of βHBA, as a result of an enzymatic reaction, was converted to NADH that increased the fluorescence intensity of Qdots (Figure 7.6). The serum and pretreated milk samples were tested after 3 min

7.3 Upconversion Nanoparticles

Target protein in serum

Add MB-Ab and QD-Ab

Incubate with target protein

Isolate beads with magnet

Discard supernatant

Resuspend Elute QD and beads measure PL

MB MB

MB MB

(a)

Washing wells

PDMS glass bottom

Fuorescence intensity (RFU)

(b)

100

10

0.01

Fuorescence intensity (RFU)

(c)

0.1 1 10 100 HRP2 concentration (ng/ml)

300 250 200 150 100

(d)

0 ng 0.1 ng 0.3 ng 1 ng HRP2 concentration

3 ng

(e)

(f)

Figure 7.5 (a, b) Schematic illustration of viral-based and microfluidic MB-based assays for detection of pfHRP2 proteins. (c) The vial-based assay results regarding the HRP2 concentration in human serum. The fluorescence intensity and concentration show a linear correlation between 0.1 and 10 ng/ml. (d) The results of MB-based assay show increased RFU for the same concentration of HRP2. (e) Simple agitation/mixing process by changing current direction and (f ) translocation of MBs within different wells using PCB. Source: Kim et al. 2017 [39]. Reproduced with permission of American Chemical Society.

of incubation in the microfluidic device. The results of serum-based detection showed a better detection limit compared to those of milk samples.

7.3 Upconversion Nanoparticles Another type of nanomaterial that has been used in photoluminescence (PL)-based nanobiosensing approached is UCNP. These nanoparticles have to

219

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

OH OH QDs O bHBA dehydrogenase

bHBA O

OH

O acetoacetate QDs

High fluorescence

1.2 Serum 1.0 0.8 0.6 0.4 0.2 0.0 30

(a)

Normalized photodiode output

Low fluorescence

Normalized photodiode output

220

(c)

60

120

300

Concentration (
M)

Milk

1.0 0.8 0.6 0.4 0.2 0.0

500

0

(d)

50

250

500

1000

Concentration (
M)

LED 470 nm single band band-pass filter

Mixing channel 200
m

Sensing well

Inlet

Microfluidic chip

ϕ80
m

655 nm single band band-pass filter

Incubation channel

ϕ2.5 mm Outlet Depth: 60
m

Overall size: 50 mm × 75 mm

Photodiode

(b)

Data acquisition

Figure 7.6 (a) Conjugation of NAD+ on the surface of Qdots was performed by immobilization of them on the 3-aminophenylboronic acid monohydrate (APBA)-conjugated Qdots. NAD+ reduces the amount of fluorescent intensity of Qdots through an energy transfer (ET) process. (b) Illustration of microfluidic device and photodiode-equipped platform for diagnosis of SCK. (c) The normalized photodiode intensity versus different concentration of βHBA in serum and (d) milk samples. Source: Weng et al. 2015 [40]. Reproduced with permission of Elsevier.

be excited with a relatively higher wavelength, thus proposing some additional advantages over Qdots or conventional organic dyes. UV is typically used for excitation of PL reagents, which inflict damage on living cells in long-term assays. Moreover, the signal-to-background ratio is usually low in the Qdot- or organic dye-based assays because of scattering of light by living tissues. Furthermore, Qdots made by heavy metals may have a toxic effect on cells. This can limit their application in biological diagnostics. UCNPs, on the other hand, are excited by near-infrared (NIR) emission in which cells show an optical transparency window. UCNPs also produce a lower level of phototoxicity and light scattering [41]. Similar to Qdots, UCNPs can be coupled with FRET technology, thus improving the sensitivity and selectivity for detection of bioanalytes [42]. UCNPs can also be conjugated with different moieties or antibodies to be prepared for biological assays. In one study, Li et al. [43] were able to detect prostate-specific antigen (PSA) in human serum using luminescence resonance energy transfer (LRET) between UCNPs and gold nanoparticles (GNPs). The UCNP was conjugated with an anti-PSA antibody (donor) to perform a sandwich assay in which the fluorescent emission from the UCNP was quenched after binding with antibody-conjugated GNP (acceptor) (Figure 7.7).

7.4 Fluorescent Biodots 980 nm laser

CMOS

Green fluorescence

Tube lens Single-particle counting

Emission filter

980 nm Laser

Interfering molecules

Dichroic mirror

Beam expander Target protein Quenched fluorescence

40X NA 0.65

Sample on coverslip

UCNPs

Single-particle counting

GNPs

Ab1

Ab2

PSA

Interfering molecules

(a)

(b)

(c) 0.6

0.8

y = 0.0095x + 0.013 R2 = 0.98

0.6 Ratio

Ratio

0.4 0.4

0.2 0.2

0

0 0

(d)

200 400 Concentration of PSA (pM)

600

0

(e)

20 40 Concentration of PSA (pM)

60

Figure 7.7 (a) Schematic illustration of PSA detection using LRET quenching between UCNPs and GNPs sandwiching specific antigens. The selectivity of the test was examined in the presence of interfering proteins such as IgG and HSA. (b, c) Fluorescence of antibody-conjugated UCNPs on a glass slide and fluorescence quenching after addition of antigen and antibody-conjugated GNPs. (d) The ratio of quenched UCNPs and number of total UCNPs versus the concentration of PSA with LOD of 1.0 pM. The correlation between this ratio and PSA concentration is found to be linear in the range of 0–60 pM concentration of PSA as shown in part (e). Source: Li et al. 2018 [43]. Reproduced with permission of American Chemical Society.

7.4 Fluorescent Biodots Many investigations are currently underway concerning the imaging of the biological molecules and cellular components using fluorescent nanomaterials. The primary trend is to apply such materials in the precise imaging of the

221

222

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

biological phenomena, which holds promise in the diagnosis and tracking of various biomolecules. The last generation of fluorescent nanomaterials, namely carbon-based Qdots or simply “biodots,” has attracted considerable interest because of their exceptional properties that are useful in the biomedical imaging [44]. Their properties, including high QY, water bioavailability, multicolor wavelength emission, and exceptional biocompatibility make them attractive alternatives for the current toxic Qdots. In other words, biodots possess many of the properties of Qdots except for their toxicity, which extends the applicability of these fluorescent nano-sized dots toward in vivo environments, including sensing, imaging, drug delivery, and theranostics [45]. Biodots were discovered as a by-product in research work on the preparation of carbon nanotubes [46]. Later, a group dubbed these fluorescent nanostructures as carbon dots (c-dot). The term biodots was later given to carbon dots prepared from biological sources of carbon such as plant extracts [47]. The primary driver for the transition from c-dot to biodots was that many nonbiocompatible compounds were needed to be used in the synthesis of c-dots. Therefore, green approaches were introduced to minimize the environmental risk and to maximize the biocompatibility of c-dots. Therefore, green c-dots are called biodots because of their eco-friendly origin [48]. They are prepared from various natural sources such as plant materials, animal materials, human hair, or biofluids. The source may also include pure biomaterials such as vitamins, carbohydrates, nucleic acids, and proteins [49]. The main feature of biodots is their exceptional optical properties, including high resistance to the photobleaching, high QY, and multicolor excitation emission as an agent for multiplex imaging [49]. From the chemistry point of view, biodots are mainly composed of oxygen, carbon, hydrogen, and nitrogen atoms. Nevertheless, these atoms are organized in such a way that they can absorb light, be excited, and emit the light in a longer wavelength, hence demonstrating fluorescent properties. Regarding the source material for synthesis, biodots show superior properties, as they are synthesized using natural materials that are cheap and rich in multiple chemical groups, which provide enough passivation of the surface not to require any doping modifications [50]. Several approaches have been devised to prepare different types of biodots. These methods are mainly classified into two strategies: “top-down” and “bottom-up” [51, 52]. In top-down approaches, the biodots are created by breaking down the bulk sources such as multiwalled carbon nanotubes or graphite materials under harsh physical or chemical conditions [47, 53, 54]. Whereas in the bottom-up approaches, biodots are synthesized by the association of small organic molecules, which exist in biological materials such as carbohydrates, under various conditions. In most bottom-up methods, heat is recruited as the primary tool for the formation of the fluorescent molecular structures, where hydrothermal heating, microwave pyrolysis, and ultrasonication are among frequently used techniques [55–57]. The most suitable method for preparation of biodots is hydrothermal treatment, which is affordable and straightforward and uses nontoxic material. The source material is converted to biodots through a multistage process including dehydration, polymerization, carbonization, and passivation [58–60] (Figure 7.8).

7.5 Digital Microfluidic Systems for Diagnosis Detection

CH3 CH3 CH3

OH CH3

CH3

HOH HO

HO H

H OH OH OH

Figure 7.8 Various natural resources for the synthesis of fluorescent biodots. Biodots are synthesized from various carbon sources including nucleic acids, proteins, plant material, graphene, and small molecules. The linear carbon atoms produce rings. The fluorescence properties of biodots originate from the rings that can absorb light and emit it in higher wavelengths.

7.5 Digital Microfluidic Systems for Diagnosis Detection One of the main aims of microfluidics is to be utilized as an easy-to-use POC system for rapid diagnosis and disease detection [61]. DMFs is one of the growing fields within the microfluidic community with great promise for translation into the practice [62]. Discrete droplets are manipulated within DMF platforms by electrostatic forces. Numerous operations can be carried out for droplets in DMF systems such as mixing, merging, transport, dispensing, and splitting [63]. These systems do not require any pumps or valves and can handle several droplets while specific processes are performed, independently. In addition, these devices do not require any predesigned inlets and outlets, and a specific flow path is not defined for them. These salient features enable scientists to integrate multistep immunoassays into a single microfluidic-based POC device [64]. Generally, DMF systems are categorized into two classes: open and closed systems. In the open classification, droplets are placed on top of a plate where electrodes are installed at the bottom layer of the device, and each of them can be controlled individually. Jain et al. investigated the effect of electrode geometry on the velocity of droplets in an open DMF system. In this study, electrode patterns with the shape of zigzag-flat, zigzag, interdigital, and square were evaluated both numerically and experimentally. They realized that the configuration of the zigzag-flat electrode led to rapid transportation of droplets [65]. Open DMF devices are characterized to have access to each droplet from the top side, facilitating the delivery of droplets to different locations within the device. A closed system is the one where droplets are surrounded between the top and bottom plate. As the top layer exists, friction also exists that reduces the droplet

223

224

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

velocity. Therefore, these devices have a lower droplet velocity compared to open ones [66]. Unlike the open DMF systems, in closed ones, the grounding is always the top layer. In the beginning, the detection mechanism for DMF immunoassay was mainly optical, where chemiluminescence or fluorescence was used for detection [67]. Miller et al. developed a DMF device that can directly perform heterogenous immunoassay on the surface of the device [68]. As the heterogenous immunoassay was performed on the surface of the device, directly, the need for BSA (for blocking the surface), beads or magnets, and carrier oil for suspending the media and movement of droplets were eliminated. The schematic workflow of the device is illustrated in Figure 7.9a–c. The detection method was based on the fluorescent signal where FITC-labeled anti-IgG nanoparticles were used as the detector, and a fluorescence plate reader was utilized for measuring the fluorescence. By a combination of magnetic force and DMF, A.H.C. Ng et al. proposed a new platform for particle-based immunoassay that relied upon chemiluminescent detection [69]. In their platform, they eliminated the usage of carrier oil, where the reagent volume, as well as processing time, was reduced to 100 and 10 times, respectively. The mechanism and schematic illustration of the device is shown in Figure 7.1d–f. In another work, Choi et al. proposed an automated DMF platform for magnetic particle-based immunoassays based on chemiluminescent detection [70]. Their automated platform contained a 90 Pogo pin interface for DMF control, an integrated, motorized detector for detection, and a magnetic lens assembly that can facilitate simultaneous (up to eight) DMF magnetic separation. Furthermore, their platform enabled investigators to gain the efficiency of design of experiment protocol for optimization of immunoassay performance. Recently, a DMF immunoassay was developed based on electrochemical detection [67]. As illustrated in Figure 7.9g, gold sensing and silver counting/pseudoreference electrodes were added to the upper plate. Then, stimulating thyroid hormone (TSH) was captured by the primary antibody (mounted on top of magnetic beads), and the captured TSH was recognized by a horseradish peroxidase enzyme (HRP)-labeled anti-TSH antibody. Afterward, the positive signal was detected through oxidation of 3,3′ ,5,5′ -tetramethylbenzidine by HRP. Loop-mediated isothermal amplification (LAMP) is an emerging method for diagnosis and detection because it can be operated at a constant temperature while maintaining its high efficiency. Thus, the requirements of reaction facilities are simplified, dramatically. The combination of DMF with LAMP can be considered as the next step toward molecular detection. Wan et al. proposed a DMF–LAMP system for sequence-specific pathogen detection [71]. In their system, they improved their technique in several aspects, including a considerable reduction in reagent consumption, increasing the sensitivity of the technique through a short incubation time, and real-time monitoring. Also, Coelho et al. proposed a specific DMF platform for LAMP-based amplification of c-Myc oncogene as a cancer biomarker [72]. Their proposed device had a retrieving reservoir in which reagent transport, LAMP reaction, and product withdrawal occur.

Capture antibody Droplet

Counter electrode

te pla ell ltiw Mu

Surrounding medium

Capture antibody (Fc-specific anti-IgG)

Antigen (IgG)

Driving electrodes

e

vic

Insulator

De

(b)

(a)

(c)

Electrode array

Probe points

Magnetic Particle

FITC-labeled detection antibody (Fab-specific anti-IgG)

Analyte Glass Electrode Hydrophobic Dielectric Movable magnet

Antibody Enzyme

ve etiti

p com

Non

Top plate

Com

(d)

Droplet

peti

tive

Air

Bottom plate

Magnet Magnetic particles

Wash buffer

Waste

Waste buffer

Conjugate

(e) DMF counter-electrode

2.25 mm Magnet

Au WE Ag CE/RE

Sample 1 Substrate 1 Waste

(f)

Assay Region

Contact pads

Au WE

4.5 mm Substrate 2 Sample 2

(g)

Figure 7.9 Digital microfluidics for biological diagnosis. (a) Driving electrodes lead to movement of droplets. (b) Illustration of the 10-platform device. The DMF device is placed on top of a multiwell plate for the aim of fluorescence detection. (c) Schematic illustration of the detection method. The top plate is first decorated with the captured antibody. Then, binds with IgG that binds FITC-labeled anti-IgG. (d) Schematic depicting of competitive and noncompetitive immunoassay in the presence of magnetic particles. (e) Isometric and (f ) top view of the proposed design by Ng et al.. A movable magnet was mounted at the bottom of the device for immobilization of particles. Source: Ng et al. 2012 [69]. Reproduced with permission of American Chemical Society. (g) Schematic illustration of the top plate from top view. The top plate was patterned with indium tin oxide to contain gold sensing and silver counter/pseudo-reference electrodes.

226

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

7.6 Paper-Based Diagnostics 7.6.1

Structure and Chemistry of Paper

Paper, a planar thin material, is prepared through processing of wood-based resources such as trees where the cellulose fibers of pulp are squeezed together. Because of the capillary wicking of water in the paper, it is used as a substrate for the fabrication of channels and reservoirs in which fluid can flow. Furthermore, the paper is a cheap material, and microfluidic structures can easily be implemented by using simple desktop printers and heaters. In other words, the main feature of paper-based microfluidic devices is their exceptional potential for mass production [73–75]. Therefore, many methods are being developed to fabricate paper-based microfluidic devices, also known as microfluidic paper-based analytical devices (μPADs). Nevertheless, the core technique is to print the desired pattern, such as channels and reservoirs, from a hydrophobic ink on the paper substrates, followed by curing the printed structure [75, 76]. Different types of papers are used as the device substrate to meet specific requirements. For instance, papers with smaller pore sizes transport the fluid more quickly. The difference in the chemical structure is another factor that affects the selection of paper substrates, where for instance, nitrocellulose paper is used for the detection of biomolecular analytes [77, 78]. As mentioned earlier, paper-based diagnostic devices have more potential than non-paper-based devices for upscaling and commercialization. Nevertheless, some disadvantages may hamper their use for accurate detection of diseases. The main issue is that paper has low mechanical durability, and its porous structure may interfere with a homogenous biorecognition reaction that may result in nonreproducible results. Furthermore, it is not a suitable device for continuous analyses where multiple steps are needed to complete the biorecognition reaction. Another issue with paper-based (bio)-sensing is that paper-based readout signals are not as robust as the standard diagnostic assays. Different strategies are recruited to address some of the above issues. In order to obtain the maximum efficiency, modification of the texture and surface of the paper is undertaken to improve the functionality of the device [79]. For example, in order to increase the speed of the liquid in the device, glutaraldehyde is used for cross-linking the paper texture. This cross-linking produces smaller pores in the paper texture and leads to a more rapid liquid mass transportation [78, 80]. Another strategy for improvement of the physicochemical properties of paper-based devices is to blend them with other materials such as polymeric materials including chitosan and plastic. This blending renders more flexibility and durability to the paper while keeping its desirable physicochemical properties [80, 81]. In order to increase the signal intensity, numerous approaches are being applied, where incorporation of signal amplification steps is the most common strategy. These signal amplification steps can be biological, nanomaterial-based, or a combination of these methods. A typical example of biological signal amplification is the application of enzymes, such as HRP, in the detection

7.6 Paper-Based Diagnostics

process. The main reason for the application of enzymes is that they transduce a single biorecognition event into a detectable signal through the conversion of several substrate molecules into detectable products [82, 83]. In other strategies, bimolecular targets, such as nucleic acids, are amplified by using in vitro nucleic acid amplification methods, where the amount of target nucleic acid is increased to a detectable level [84–86]. Apart from biological entities, nanomaterials, particularly nanoparticles, serve as new platforms of signal amplification. For this purpose, nanomaterials can be used to carry thousands of signal molecules where the recognition event can trigger the release of these signal molecules, which are traceable as detectable readouts [83, 87]. In addition to their signal-carrying capabilities, nanomaterials can serve as artificial enzymes that mimic the function of natural enzymes and can catalyze a signal amplification reaction similar to that of natural enzymes. Furthermore, nanomaterials of metallic origin are widely used as microelectrodes to facilitate the transduction of the electrochemical signals into the readout zone of paper-based devices [88]. In recent strategies, a combination of biological and nanomaterial-based signal amplification methods are used, where the devices can be used for ultrasensitive detection. A nanobiomaterial of QDs and antibody was used to detect the presence of Aflatoxin B1, in which application of QD-antibody conjugates was able to amplify the positive signal 2 orders of magnitude more than GNP–antibody conjugates [89–91].

7.6.2

Applications of Paper-Based Devices in the Diagnostics

Paper holds promise for fabrication of diagnostic devices owing to the cheap source, biocompatibility, easy surface functionalization, and potential for upscaling. Therefore, paper is best suited for application as a substrate for fabrication of POC diagnostic devices. The application of paper-based devices as POC solutions adds values to existing POC devices in several ways. Facile and almost equipment-free mass production of POC devices is the main contribution of paper-based substrates in the diagnostics industry, which results in the production of very cheap diagnostic kits. In other words, these characteristics facilitate dissemination of these diagnostic devices into remote and underdeveloped areas, which in turn leads to the minimization of disease burden in such locations through affordable diagnosis [76, 92–94]. The most challenging step in the production of paper-based POC diagnostic devices is the integration of in vitro diagnostic approaches, such as biosensing methods, into such devices. Based on the readout signal, mainly optical and electrical biosensing methods are integrated into paper-based POC diagnostic devices. Optical biosensing is rather frequent owing to its simplicity, ease of integration, and compatibility with multiplex readouts. Nevertheless, electrical readouts are shown to be more sensitive and can be used in reagent-free detection systems [90, 95–100]. Paper-based biosensing is classified into labeled and label-free categories:

227

228

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

7.6.2.1

Labeled Biosensing

A label is a moiety that functions as a signal molecule while a biorecognition event occurs. The main reason for using a label molecule is that neither detection target nor the detection ligand produces a signal during their interaction. Therefore, a variety of label molecules is conjugated to the ligand, target, or both. Because of the incorporation of labels in the detection process, labeled biosensing tends to give a more sensitive signal. However, as the sensing counterparts (target and ligand) need to be modified, labeling may negatively affect the activity of the target and the ligand. Furthermore, the labeling contributes to a higher fabrication cost. Depending on the chemical properties of the sensing counterparts and the required accuracy, a variety of label molecules is applied. Optical label molecules are the most commonly used moieties, which are detected visually or by using a reader instrument such as UV–Vis spectrophotometer or fluorescence reading machine. The optical labels range from small chemical molecules to biomacromolecules, where a variety of chemistries are applied to acquire the biorecognition signal [101]. 7.6.2.2

Label-Free Biosensing

In label-free biosensing, no label is used to produce the signal. Instead, the signal is produced through changes in the physicochemical properties of either sensor surface, receptor, ligand, or a combination of these elements. For instance, the recognition event may lead to changes in light reflection, electron transfer, heat transfer, etc. [101, 102]. In paper-based microfluidic devices, the paper matrix is used to deposit the sensing materials including nanomaterials, to produce a target-responsive matrix. For example, for performing electrochemical analysis, conductive materials are deposited as electrodes on the paper substrate where the signal transduction takes place. Lei et al. has deposited a layer of carbon nanotubes, as nanoelectrodes, on a paper device to perform a sensitive electrochemical label-free immunoassay of biotin and streptavidin interaction with a detection limit of 25 ng/ml [103]. The main advantage of label-free methods is their simplicity, where there is no need for labeling. Moreover, the fabrication cost of the device is affordable, and more importantly, the labeling does not destroy the biosensing structure. 7.6.3 Integration of Nanoparticles with Paper-Based Microfluidic Devices With regard to the exceptional properties of nanomaterials, they are widely integrated into paper-based devices. In this section, we overview the application of a variety of nanomaterials and their roles in these devices. 7.6.3.1

Gold Nanomaterials

Gold nanomaterials are mainly used in paper-based device detection. These nanomaterials possess various exceptional optical properties that originate from the oscillation or resonance of their surface plasmon (SPR). Because the surface plasmon is sensitive to the refractive index change of the surrounding

7.6 Paper-Based Diagnostics

media, it is used to design various plasmonic-based nanobiosensors. When an object, regardless of its origin, binds to the surface of gold nanomaterials, the surface plasmon is changed, which gives rise to a change in the optical signal. The mentioned object can be neighboring nanoparticles or direct deposition of chemical layers on gold nanomaterials. In its most simple configuration, the change in SPR can lead to a color change that can be applied for the visual detection of targets in paper-based devices. In this context, the aggregation and seed-mediated growth of GNPs are used to detect the analyte. Aggregation of GNPs induces a color change; similarly, the seed-mediated growth leads to the amplification of specifically surface-bound GNPs that serve as nuclei of the gold nanomaterials deposition [104]. As an example of gold nanomaterial-based visual detection, GNPs were used to detect the presence of tuberculosis bacterium (TB) DNA. The study showed that TB DNA could induce aggregation of GNPs, which leads to a change in color in the GNP solution, and the intensity of this color change can be analyzed after transferring the complete reaction to wax-printed microwells on paper. Later, the image of each reaction zone is captured by a mobile phone and then analyzed to quantify the extent of signal intensity. In this work, the paper serves as an adsorbent to remove the fluid of the reaction (Figure 7.10a) [105]. In addition to infectious disease biosensing, GNP-based signal amplification was recruited to detect cancer protein biomarkers. In this detection method, authors report on the application of a gold nanobioconjugate that carries multiple HRP signal molecules and the recognition antibody. In this work, a nitrocellulose paper substrate was used to blot the sample, followed by the addition of the as-prepared nanobioconjugate. In the next step, detection is performed based on the HRP–TMB reaction. Compared to conventional ELISA, this method was shown to lead to more intense color development, indicating the signal amplification of the nanobioconjugate (Figure 7.10b) [83]. In addition to optical detection chemistries, gold nanomaterial-based electrodes have been integrated into paper-based devices to improve the limit of the detection figure of merits. In some configurations, the electrode was printed on the paper substrate using screen-printing technology, and microfluidic structures were fabricated using photolithography. This device has been used to detect several bioanalytes using oxidase enzyme reactions, in which the electrode was used to transfer the produced electrons to signal processing parts [107]. 7.6.3.2

Fluorescent Nanomaterials

The high sensitivity of fluorescent nanotags has led to the design of ultrasensitive, yet simple paper-based devices that are used to detect minute quantities of analytes. Nevertheless, in the readout of the nanobiosensing process, a device is needed that can excite and read the fluorescence signals. This issue, however, has been partially addressed by analysis of the signals by using smartphone-based fluorescence detectors. Qdots [108] and UCNPS [109] are the main classes of the nanomaterials that have been embedded into paper-based devices. As explained in Section 7.2, Qdots are tiny semicrystal fluorescent nanoparticles. These nanoparticles have

229

230

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

TB negative

TB positive

Culture of TB

Staining

TB 123 bp

(a)

Gel electrophoresis

Fluorescence

Phycocyanin

FL

Colorimetric signaling

FL

Quenching

Wavelength (nm) Paper cellulose

(b)

QDs

Recognition site

Wavelength (nm) PC

Quenching QDs

Excitation light

(c)

Figure 7.10 Integration of nanoparticles with paper-based microfluidic devices for biological diagnosis. (a) The visual paper-based detection of TB DNA based on the aggregation of gold nanoparticles. Source: Tsai et al. 2017 [105]. Reproduced with permission of American Chemical Society. (b) The paper-based signal amplification for the detection of cancer biomarker. Source: Huang et al. 2018 [83]. Reproduced with permission of American Chemical Society. (c) The principle of the 3D paper device integrated with Qdots for phycocyanin detection. Source: Li et al. 2017 [106]. Reproduced with permission of American Chemical Society.

been used in a variety of paper-based biosensing configurations. In one configuration, an origami-based 3D paper device was used to detect phycocyanin using molecularly imprinted Qdots. When the phycocyanin migrates toward the Qdots and binds to them, the emission of the Qdot is quenched compared with the unbound Qdots, leading to an “on-off” paper-based device (Figure 7.10c) [106]. As discussed in Section 7.3, UCNPs are another class of the fluorescent nanomaterials which have found many applications in ultrasensitive cellular and molecular tracking systems. Therefore, they have been applied to detect targets in paper-based devices because of their exceptionally high QY and prolonged decay. In a study by He et al., a portable UCNP-based paper device has been designed to detect cocaine, where an aptamer-based fluorescence-quenching detection system was applied. In this detection system, GNPs, which had been preconjugated to the free ends of the cocaine aptamer and UCNPs, were immobilized on the paper surface. In its unbound form, the GNP was not located in the vicinity of the UCNP. However, binding of the cocaine to the

Acknowledgment

aptamer induced the conformational change of the aptamer, which culminated its two ends. This culmination led to quenching of the UCNP signal by the GNP. Therefore, with increases in the concentration of cocaine, the fluorescence decreased in the microwells of the paper-based device [110].

7.7 Conclusion and Future Perspective Early biological diagnosis means detecting a marker or markers that manifest a particular disease before clinical symptoms. Early detection is crucial for managing chronic diseases such as cancer before further complications. Thus, an efficient diagnostic assay should possess high sensitivity, high selectivity, reliable limit-of-detection, and other analytical metrics. The combination of microfluidics and nanotechnology can lead to the development of highly efficient platforms for effective biological diagnosis. Microfluidics can hugely affect this field by high-throughput production and screening of nanoparticle-based diagnosis, making nanotechnology more appealing for clinical applications. Qdots, UCNPs, and biodots are three principal nanoparticles that are used for biosensing and POC applications. In this chapter, we discussed the integrations of these nanoparticles with microfluidic platforms for biological diagnosis. Furthermore, we evaluated the applications of two emerging formats of microfluidics, i.e. DMFs and paper-based microfluidics, for biological diagnosis. The application of signal quantification methods, such as image-processing procedures, can be adopted as an efficient strategy to cope with the challenges in signal quantification, which are generated in the microfluidic devices. Currently, the integration of nanomaterials and mobile-based image processing technologies in microfluidic-based diagnostics have opened up a tremendous opportunity for end user scale implementation of analytically robust and fully automated low-cost devices. The less-developed part of microfluidic-based biosensors is the absolute quantification of signals, broadening the dynamic range and the multiplicity of the devices. It is evident that addressing the above challenges requires further efforts in the development of novel readout devices that can analyze the signals in reliable levels and be comparable to standard analytical instruments.

Conflicts of Interest The authors declare no conflicts of interest.

Acknowledgment This research work is supported by the National Health and Medical Research Council via the career development fellowship (APP1143377) and the Australian Research Council through discovery project grants (DP170103704 and DP180103003).

231

232

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

References 1 Chen, H., Zhen, Z., Todd, T. et al. (2013). Nanoparticles for improving 2

3

4

5

6

7

8

9

10 11

12 13

14 15

cancer diagnosis. Mater. Sci. Eng., R 74 (3): 35–69. Nguyen, N.-T., Shaegh, S.A.M., Kashaninejad, N., and Phan, D.-T. (2013). Design, fabrication and characterization of drug delivery systems based on lab-on-a-chip technology. Adv. Drug Deliv. Rev. 65 (11): 1403–1419. Nguyen, N.-T., Hejazian, M., Ooi, H.C., and Kashaninejad, N. (2017). Recent advances and future perspectives on microfluidic liquid handling. Micromachines 8 (6): 186–205. Kashaninejad, N., Nikmaneshi, R.M., Moghadas, H. et al. (2016). Organ-tumor-on-a-chip for chemosensitivity assay: a critical review. Micromachines 7 (8): 130–153. Moshksayan, K., Kashaninejad, N., Warkiani, M.E. et al. (2018). Spheroids-on-a-chip: recent advances and design considerations in microfluidic platforms for spheroid formation and culture. Sens. Actuators, B 263: 151–176. Kashaninejad, N., Shiddiky, M.J.A., and Nguyen, N.-T. (2018). Advances in microfluidics-based assisted reproductive technology: from sperm sorter to reproductive system-on-a-chip. Adv. Biosyst. 2 (3): 1700197. Dinh, T., Phan, H.-P., Kashaninejad, N. et al. (2018). An on-chip SiC MEMS device with integrated heating, sensing, and microfluidic cooling systems. Adv. Mater. Interfaces 5 (20): 1800764. Mollajan, M., Bazaz, S.R., and Mehrizi, A.A. (2018). A thoroughgoing design of a rapid-cycle microfluidic droplet-based PCR device to amplify rare DNA strands. J. Appl. Fluid Mech. 11 (1): 21–29. Moghadas, H., Saidi, M.S., Kashaninejad, N., and Nguyen, N.-T. (2018). A high-performance polydimethylsiloxane electrospun membrane for cell culture in lab-on-a-chip. Biomicrofluidics 12 (2): 024117. Kashaninejad, N., Chan, W.K., and Nguyen, N.-T. (2011). Fluid mechanics of flow through rectangular hydrophobic microchannels, (44632), pp. 647–655. Moghadas, H., Saidi, M.S., Kashaninejad, N., and Nguyen, N.-T. (2018). Challenge in particle delivery to cells in a microfluidic device. Drug Deliv. Transl. Res. 8 (3): 830–842. Vashist, S.K., Luppa, P.B., Yeo, L.Y. et al. (2015). Emerging technologies for next-generation point-of-care testing. Trends Biotechnol. 33 (11): 692–705. Luka, G., Ahmadi, A., Najjaran, H. et al. (2015). Microfluidics integrated biosensors: a leading technology towards lab-on-a-chip and sensing applications. Sensors 15 (12): 30011–30031. Perumal, V. and Hashim, U. (2014). Advances in biosensors: principle, architecture and applications. J. Appl. Biomed. 12 (1): 1–15. Tereshchenko, A., Bechelany, M., Viter, R. et al. (2016). Optical biosensors based on ZnO nanostructures: advantages and perspectives. A review. Sens. Actuators, B 229: 664–677.

References

16 Strianese, M., Staiano, M., Ruggiero, G. et al. (2012). Fluorescence-based

17 18

19

20

21

22 23 24

25

26 27

28

29 30

31

32

biosensors. In: Spectroscopic Methods of Analysis, Methods in Molecular Biology, vol. 875 (ed. W. Bujalowski), 193–216. Springer. Vannoy, C.H., Tavares, A.J., Noor, M.O. et al. (2011). Biosensing with quantum dots: a microfluidic approach. Sensors 11 (10): 9732–9763. Li, Z., Wang, Y., Wang, J. et al. (2010). Rapid and sensitive detection of protein biomarker using a portable fluorescence biosensor based on quantum dots and a lateral flow test strip. Anal. Chem. 82 (16): 7008–7014. Tan, G.-R., Wang, M., Hsu, C.-Y. et al. (2016). Small upconverting fluorescent nanoparticles for biosensing and bioimaging. Adv. Opt. Mater. 4 (7): 984–997. Klostranec, J.M., Xiang, Q., Farcas, G.A. et al. (2007). Convergence of quantum dot barcodes with microfluidics and signal processing for multiplexed high-throughput infectious disease diagnostics. Nano Lett. 7 (9): 2812–2818. Moghadas, H., Saidi, M.S., Kashaninejad, N. et al. (2017). Fabrication and characterization of low-cost, bead-free, durable and hydrophobic electrospun membrane for 3D cell culture. Biomed. Microdevices 19 (4): 74. Ghafar-Zadeh, E. (2015). Wireless integrated biosensors for point-of-care diagnostic applications. Sensors 15 (2): 3236–3261. Kashaninejad, N., Chan, W.K., and Nguyen, N.-T. (2012). Eccentricity effect of micropatterned surface on contact angle. Langmuir 28 (10): 4793–4799. Kashaninejad, N., Nguyen, N.-T., and Chan, W.K. (2012). Eccentricity effects of microhole arrays on drag reduction efficiency of microchannels with a hydrophobic wall. Phys. Fluids 24 (11): 112004–112021. Barisam, M., Saidi, S.M., Kashaninejad, N., and Nguyen, N.-T. (2018). Prediction of necrotic core and hypoxic zone of multicellular spheroids in a microbioreactor with a U-shaped barrier. Micromachines 9 (3): 94–112. Bazaz, S.R., Mehrizi, A.A., Ghorbani, S. et al. (2018). A hybrid micromixer with planar mixing units. RSC Adv. 8 (58): 33103–33120. Lashkaripour, A., Mehrizi, A.A., Goharimanesh, M. et al. (2018). Size-controlled droplet generation in a microfluidic device for rare dna amplification by optimizing its effective parameters. J. Mech. Med. Biol. 18 (1): 1850002. Luppa, P.B., Bietenbeck, A., Beaudoin, C., and Giannetti, A. (2016). Clinically relevant analytical techniques, organizational concepts for application and future perspectives of point-of-care testing. Biotechnol. Adv. 34 (3): 139–160. Tothill, I.E. (2009). Biosensors for cancer markers diagnosis. Semin. Cell Dev. Biol. 20 (1): 55–62 Ragavan, K., Kumar, S., Swaraj, S., and Neethirajan, S. (2018). Advances in biosensors and optical assays for diagnosis and detection of malaria. Biosens. Bioelectron. 105: 188–210. Lee, W.G., Kim, Y.-G., and Chung, B.G. (2010). Nano/microfluidics for diagnosis of infectious diseases in developing countries. Adv. Drug Deliv. Rev. 62 (4–5): 449–457. Zhou, J., Yang, Y., and Zhang, C.-Y. (2015). Toward biocompatible semiconductor quantum dots: from biosynthesis and bioconjugation to biomedical application. Chem. Rev. 115 (21): 11669–11717.

233

234

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

33 Petryayeva, E., Algar, W.R., and Medintz, I.L. (2013). Quantum dots in bio-

34

35

36

37

38

39

40

41

42

43

44 45 46

47

48

analysis: a review of applications across various platforms for fluorescence spectroscopy and imaging. Appl. Spectrosc. 67 (3): 215–252. Chou, K. and Dennis, A. (2015). Förster resonance energy transfer between quantum dot donors and quantum dot acceptors. Sensors 15 (6): 13288–13325. Hu, M., Yan, J., He, Y. et al. (2009). Ultrasensitive, multiplexed detection of cancer biomarkers directly in serum by using a quantum dot-based microfluidic protein chip. ACS Nano 4 (1): 488–494. Ng, E., Hoshino, K., and Zhang, X. (2013). Microfluidic immunodetection of cancer cells via site-specific microcontact printing of antibodies on nanoporous surface. Methods 63 (3): 266–275. Zhang, R.-Q., Hong, S.-L., Wen, C.-Y. et al. (2018). Rapid detection and subtyping of multiple influenza viruses on a microfluidic chip integrated with controllable micro-magnetic field. Biosens. Bioelectron. 100: 348–354. Lee, N., Wang, C., and Park, J. (2018). User-friendly point-of-care detection of influenza A (H1N1) virus using light guide in three-dimensional photonic crystal. RSC Adv. 8 (41): 22991–22997. Kim, C., Hoffmann, G., and Searson, P.C. (2017). Integrated magnetic bead–quantum dot immunoassay for malaria detection. ACS Sensors 2 (6): 766–772. Weng, X., Chen, L., Neethirajan, S., and Duffield, T. (2015). Development of quantum dots-based biosensor towards on-farm detection of subclinical ketosis. Biosens. Bioelectron. 72: 140–147. Chen, G., Qiu, H., Prasad, P.N., and Chen, X. (2014). Upconversion nanoparticles: design, nanochemistry, and applications in theranostics. Chem. Rev. 114 (10): 5161–5214. Wang, L., Yan, R., Huo, Z. et al. (2005). Fluorescence resonant energy transfer biosensor based on upconversion-luminescent nanoparticles. Angew. Chem. Int. Ed. 44 (37): 6054–6057. Li, X., Wei, L., Pan, L. et al. (2018). Homogeneous immunosorbent assay based on single-particle enumeration using upconversion nanoparticles for the sensitive detection of cancer biomarkers. Anal. Chem. 90 (7): 4807–4814. Lim, S.Y., Shen, W., and Gao, Z. (2015). Carbon quantum dots and their applications. Chem. Soc. Rev. 44 (1): 362–381. Peng, Z., Han, X., Li, S. et al. (2017). Carbon dots: Biomacromolecule interaction, bioimaging and nanomedicine. Coord. Chem. Rev. 343: 256–277. Xu, X., Ray, R., Gu, Y. et al. (2004). Electrophoretic analysis and purification of fluorescent single-walled carbon nanotube fragments. J. Am. Chem. Soc. 126 (40): 12736–12737. Sun, Y.-P., Zhou, B., Lin, Y. et al. (2006). Quantum-sized carbon dots for bright and colorful photoluminescence. J. Am. Chem. Soc. 128 (24): 7756–7757. Shi, L., Li, Y., Li, X. et al. (2015). Facile and eco-friendly synthesis of green fluorescent carbon nanodots for applications in bioimaging, patterning and staining. Nanoscale 7 (16): 7394–7401.

References

49 Baker, S.N. and Baker, G.A. (2010). Luminescent carbon nanodots: emergent

nanolights. Angew. Chem. Int. Ed. 49 (38): 6726–6744. 50 Sharma, V., Tiwari, P., and Mobin, S.M. (2017). Sustainable carbon-dots:

51

52

53

54

55

56

57

58

59

60

61 62

63

recent advances in green carbon dots for sensing and bioimaging. J. Mater. Chem. B 5 (45): 8904–8924. Yang, Z., Li, Z., Xu, M. et al. (2013). Controllable synthesis of fluorescent carbon dots and their detection application as nanoprobes. Nano-Micro Lett. 5 (4): 247–259. Miao, P., Han, K., Tang, Y. et al. (2015). Recent advances in carbon nanodots: synthesis, properties and biomedical applications. Nanoscale 7 (5): 1586–1595. Zhou, J., Booker, C., Li, R. et al. (2007). An electrochemical avenue to blue luminescent nanocrystals from multiwalled carbon nanotubes (MWCNTs). J. Am. Chem. Soc. 129 (4): 744–745. Zheng, X.T., Ananthanarayanan, A., Luo, K.Q., and Chen, P. (2015). Glowing graphene quantum dots and carbon dots: properties, syntheses, and biological applications. Small 11 (14): 1620–1636. Li, H., He, X., Liu, Y. et al. (2011). One-step ultrasonic synthesis of water-soluble carbon nanoparticles with excellent photoluminescent properties. Carbon 49 (2): 605–609. Wang, C., Xu, Z., Cheng, H. et al. (2015). A hydrothermal route to water-stable luminescent carbon dots as nanosensors for pH and temperature. Carbon 82: 87–95. Choi, Y., Thongsai, N., Chae, A. et al. (2017). Microwave-assisted synthesis of luminescent and biocompatible lysine-based carbon quantum dots. J. Ind. Eng. Chem. 47: 329–335. Yuan, M., Zhong, R., Gao, H. et al. (2015). One-step, green, and economic synthesis of water-soluble photoluminescent carbon dots by hydrothermal treatment of wheat straw, and their bio-applications in labeling, imaging, and sensing. Appl. Surf. Sci. 355: 1136–1144. Liu, Y., Zhao, Y., and Zhang, Y. (2014). One-step green synthesized fluorescent carbon nanodots from bamboo leaves for copper(II) ion detection. Sens. Actuators, B 196: 647–652. Roy, P., Chen, P.-C., Periasamy, A.P. et al. (2015). Photoluminescent carbon nanodots: synthesis, physicochemical properties and analytical applications. Mater. Today 18 (8): 447–458. Srinivasan, V., Pamula, V.K., and Fair, R.B. (2004). Droplet-based microfluidic lab-on-a-chip for glucose detection. Anal. Chim. Acta 507 (1): 145–150. Srinivasan, V., Pamula, V.K., and Fair, R.B. (2004). An integrated digital microfluidic lab-on-a-chip for clinical diagnostics on human physiological fluids. Lab Chip 4 (4): 310–315. Yafia, M., Emran, B.J., and Najjaran, H. (2018). Digital microfluidic systems: fundamentals, configurations, techniques, and applications. In: Microfluidics: Fundamentals, Devices, and Applications (eds. Y. Song, D. Cheng and L. Zhao), 175–209. Wiley.

235

236

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

64 Sun, J., Xianyu, Y., and Jiang, X. (2014). Point-of-care biochemical assays

65

66 67 68

69

70

71

72

73

74

75

76

77

78

79

using gold nanoparticle-implemented microfluidics. Chem. Soc. Rev. 43 (17): 6239–6253. Jain, V., Devarasetty, V., and Patrikar, R. (2017). Effect of electrode geometry on droplet velocity in open EWOD based device for digital microfluidics applications. J. Electrostat. 87: 11–18. Banerjee, A.N., Qian, S., and Joo, S.W. (2011). High-speed droplet actuation on single-plate electrode arrays. J. Colloid Interface Sci. 362 (2): 567–574. Shamsi, M.H., Choi, K., Ng, A.H., and Wheeler, A.R. (2014). A digital microfluidic electrochemical immunoassay. Lab Chip 14 (3): 547–554. Miller, E.M., Ng, A.H., Uddayasankar, U., and Wheeler, A.R. (2011). A digital microfluidic approach to heterogeneous immunoassays. Anal. Bioanal.Chem. 399 (1): 337–345. Ng, A.H.C., Choi, K., Luoma, R.P. et al. (2012). Digital microfluidic magnetic separation for particle-based immunoassays. Anal. Chem. 84 (20): 8805–8812. Choi, K., Ng, A.H., Fobel, R. et al. (2013). Automated digital microfluidic platform for magnetic-particle-based immunoassays with optimization by design of experiments. Anal. Chem. 85 (20): 9638–9646. Wan, L., Chen, T., Gao, J. et al. (2017). A digital microfluidic system for loop-mediated isothermal amplification and sequence specific pathogen detection. Sci. Rep. 7 (1): 14586. Coelho, B.J., Veigas, B., Águas, H. et al. (2017). A digital microfluidics platform for loop-mediated isothermal amplification detection. Sensors 17 (11): 2616. Martinez, A.W., Phillips, S.T., Wiley, B.J. et al. (2008). FLASH: a rapid method for prototyping paper-based microfluidic devices. Lab Chip 8 (12): 2146–2150. Faustino, V., Catarino, S.O., Lima, R., and Minas, G. (2016). Biomedical microfluidic devices by using low-cost fabrication techniques: a review. J. Biomech. 49 (11): 2280–2292. Akyazi, T., Basabe-Desmonts, L., and Benito-Lopez, F. (2018). Review on microfluidic paper-based analytical devices towards commercialisation. Anal. Chim. Acta 1001: 1–17. Yamada, K., Henares, T.G., Suzuki, K., and Citterio, D. (2015). Paper-based inkjet-printed microfluidic analytical devices. Angew. Chem. 54 (18): 5294–5310. Hosseini, S., Vázquez-Villegas, P., and Martínez-Chapa, S.O. (2017). Paper and fiber-based bio-diagnostic platforms: current challenges and future needs. Appl. Sci. 7 (8): 863–885. Sher, M., Zhuang, R., Demirci, U., and Asghar, W. (2017). Paper-based analytical devices for clinical diagnosis: recent advances in the fabrication techniques and sensing mechanisms. Expert Rev. Mol. Diagn. 17 (4): 351–366. Credou, J., Faddoul, R., and Berthelot, T. (2014). One-step and eco-friendly modification of cellulose membranes by polymer grafting. RSC Adv. 4 (105): 60959–60969.

References

80 Wu, T., Du, Y., Yan, N., and Farnood, R. (2015). Cellulose fiber networks

81

82

83

84

85

86

87

88 89

90

91

92

93

94

reinforced with glutaraldehyde–chitosan complexes. J. Appl. Polym. Sci. 132 (33): 42375–42383. Morsy, F.A. and El-Sherbiny, S. (2004). Mechanical properties of coated paper: influence of coating properties and pigment Blends. J. Mater. Sci. 39 (24): 7327–7332. Murdock, R.C., Shen, L., Griffin, D.K. et al. (2013). Optimization of a paper-based ELISA for a human performance biomarker. Anal. Chem. 85 (23): 11634–11642. Huang, J.Y., Lin, H.-T., Chen, T.-H. et al. (2018). Signal amplified gold nanoparticles for cancer diagnosis on paper-based analytical devices. ACS Sensors 3 (1): 174–182. Seok, Y., Joung, H.A., Byun, J.Y. et al. (2017). A paper-based device for performing loop-mediated isothermal amplification with real-time simultaneous detection of multiple DNA targets. Theranostics 7 (8): 2220–2230. Altundemir, S., Uguz, A.K., and Ulgen, K. (2017). A review on wax printed microfluidic paper-based devices for international health. Biomicrofluidics 11 (4): 041501. Jauset-Rubio, M., Svobodova, M., Mairal, T. et al. (2016). Ultrasensitive, rapid and inexpensive detection of DNA using paper based lateral flow assay. Sci. Rep. 6: 37732. Ge, L., Yu, J., Ge, S., and Yan, M. (2014). Lab-on-paper-based devices using chemiluminescence and electrogenerated chemiluminescence detection. Anal. Bioanal.Chem. 406 (23): 5613–5630. Han, K.N., Choi, J.-S., and Kwon, J. (2017). Gold nanozyme-based paper chip for colorimetric detection of mercury ions. Sci. Rep. 7 (1): 2806. Ren, M., Xu, H., Huang, X. et al. (2014). Immunochromatographic assay for ultrasensitive detection of aflatoxin B(1) in maize by highly luminescent quantum dot beads. ACS Appl. Mater. Interfaces 6 (16): 14215–14222. Silveira, C.M., Monteiro, T., and Almeida, M.G. (2016). Biosensing with paper-based miniaturized printed electrodes-A modern trend. Biosensors (Basel) 6 (4): 51–67. Ruecha, N., Yamada, K., Suzuki, K., and Citterio, D. (2017). (Bio)Chemical sensors based on paper. In: Materials for Chemical Sensing (eds. T.R.L. Cesar Paixão and S.M. Reddy), 29–74. Cham: Springer International Publishing. Tenda, K., van Gerven, B., Arts, R. et al. (2018). Paper-based antibody detection devices using bioluminescent BRET-switching sensor proteins. Angew. Chem. 57 (47): 15369–15373. Hamedpour, V., Leardi, R., Suzuki, K., and Citterio, D. (2018). Fabrication of paper-based analytical devices optimized by central composite design. Analyst 143 (9): 2102–2108. Yamada, K., Shibata, H., Suzuki, K., and Citterio, D. (2017). Toward practical application of paper-based microfluidics for medical diagnostics: state-of-the-art and challenges. Lab Chip 17 (7): 1206–1249.

237

238

7 Biological Diagnosis Based on Microfluidics and Nanotechnology

95 Alhogail, S., Suaifan, G.A.R.Y., and Zourob, M. (2017). Development of rapid

96

97

98 99

100 101 102 103

104 105

106

107 108 109 110

and low-cost paper based sensing platform for bacterial detection. Procedia Technol. 27: 146–148. Liu, B., Du, D., Hua, X. et al. (2014). Paper-based electrochemical biosensors: from test strips to paper-based microfluidics. Electroanalysis 26 (6): 1214–1223. Sin, M.L.Y., Mach, K.E., Wong, P.K., and Liao, J.C. (2014). Advances and challenges in biosensor-based diagnosis of infectious diseases. Expert Rev. Mol. Diagn. 14 (2): 225–244. Ulep, T.-H. and Yoon, J.-Y. (2018). Challenges in paper-based fluorogenic optical sensing with smartphones. Nano Converg. 5 (1): 14. Shafiee, H., Asghar, W., Inci, F. et al. (2015). Paper and flexible substrates as materials for biosensing platforms to detect multiple biotargets. Sci. Rep. 5: 8719. Yang, Y., Noviana, E., Nguyen, M.P. et al. (2017). Paper-based microfluidic devices: emerging themes and applications. Anal. Chem. 89 (1): 71–91. Sang, S., Wang, Y., Feng, Q. et al. (2016). Progress of new label-free techniques for biosensors: a review. Crit. Rev. Biotechnol. 36 (3): 465–481. Zanchetta, G., Lanfranco, R., Giavazzi, F. et al. (2017). Emerging applications of label-free optical biosensors. Nanophotonics 6 (4): 627–645. Lei, K.F., Yang, S.I., Tsai, S.W., and Hsu, H.T. (2015). Paper-based microfluidic sensing device for label-free immunoassay demonstrated by biotin-avidin binding interaction. Talanta 134: 264–270. Zhou, W., Gao, X., Liu, D., and Chen, X. (2015). Gold nanoparticles for in vitro diagnostics. Chem. Rev. 115 (19): 10575–10636. Tsai, T.-T., Huang, C.-Y., Chen, C.-A. et al. (2017). Diagnosis of tuberculosis using colorimetric gold nanoparticles on a paper-based analytical device. ACS Sensors 2 (9): 1345–1354. Li, B., Zhang, Z., Qi, J. et al. (2017). Quantum dot-based molecularly imprinted polymers on three-dimensional origami paper microfluidic chip for fluorescence detection of phycocyanin. ACS Sensors 2 (2): 243–250. Dungchai, W., Chailapakul, O., and Henry, C.S. (2009). Electrochemical detection for paper-based microfluidics. Anal. Chem. 81 (14): 5821–5826. Xu, Y., Wang, X., Zhang, W.L. et al. (2018). Recent progress in two-dimensional inorganic quantum dots. Chem. Soc. Rev. 47 (2): 586–625. Wen, S., Zhou, J., Zheng, K. et al. (2018). Advances in highly doped upconversion nanoparticles. Nat. Commun. 9 (1): 2415. He, M., Li, Z., Ge, Y., and Liu, Z. (2016). Portable upconversion nanoparticles-based paper device for field testing of drug abuse. Anal. Chem. 88 (3): 1530–1534.

239

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses Dong Wang 1 , Ho N. Chan 1 , Zeyu Liu 1 , Sean Micheal 1 , Lijun Li 1 , Dorsa B. Baniani 2 , Ming J. A. Tan 2 , Lu Huang 1 , Jiantao Wang 1 , and Hongkai Wu 1,2 1

The Hong Kong University of Science and Technology, Department of Chemistry, Kowloon, Hong Kong, China

2 The Hong Kong University of Science and Technology, Division of Biomedical Engineering, Kowloon,

Hong Kong, China

8.1 Introduction Microfluidics is the science and technology of systems that process or manipulate small (10–9 to 10–18 l) amounts of fluids, using channels with dimensions of tens to hundreds of micrometers [1]. After first applied to the field of chromatography and electrophoresis [2, 3], microfluidics technology has remarkable results in different areas such as drug screening, single-cell analysis, tissue engineering, and disease diagnosis [4–11]. In the past few decades, diagnostic tests for disease-related biomarkers (cells, nucleic acids, proteins, and small molecules) are still actualized in central laboratories with benchtop equipment and carried out by experienced technicians. As a result of that, patients need to wait for several days to a few weeks to obtain the results, and the state of the illness could deteriorate more during the waiting period. Also, in low- and middle-income countries (LMICs), disease diagnosis has further limitations owing to inadequate resources and lack of skilled persons. Point-of-care testing (POCT), referred to a laboratory assay that can be carried out outside of the key laboratory, shows great promise for application in diagnostic tests [12–15]. Moreover, the ability for microfluidic chips for processing samples with small volume and obtain analytical signal within a few minutes makes it perfectly suitable for POCT. In the past few years, microfluidic-based POCT devices emerge as a powerful technique that shows a great hope to satisfy the ASSURED criteria (Affordable, Sensitive, Specific, User-friendly, Rapid and robust, Equipment-free, and Deliverable to end users) outlined by the World Health Organization (WHO) [16]. As a powerful technique, microfluidic systems play an essential role in the POCT device. The achievement in biotechnology, including single-cell analysis, novel isothermal nucleic acid amplification, and enzyme-linked immunosorbent assay (ELISA), has provided patients both sensitive and specific diagnostic Nanotechnology and Microfluidics, First Edition. Edited by Xingyu Jiang. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.

240

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

methods [17–20]. These types of POCT devices are capable of detecting cell behavior, nucleic acid, protein, and small molecules by utilizing microfluidic chip as a core subunit [13, 15]. In this chapter, microfluidic-based POCT devices based on the biomarkers they respond to will be briefly introduced. Their applications in disease diagnose will also be covered.

8.2 Cell Cell-based POCT diagnostics play a vital role in hematology laboratory practices. This contains counting of specific cell types from whole blood for the diagnosis, assessment, and monitoring of diseases. To replace traditional complex techniques such as flow cytometry, an increasing number of microfluidic devices for point-of-care (POC) cellular analysis emerge, for example, whole blood counting, characterization of CD64 expression for sepsis, enumeration of CD4+ T lymphocytes for HIV monitoring, and circulating tumor cells isolation and analysis for cancers. In the following, the aspect mentioned above will be discussed. 8.2.1

Blood Cell Counting

Microfluidic devices have also been widely exploited in cell-based diagnosis, among which blood cell counting is one of the most important applications. Blood cell counting refers to the process of enumerating different types of hematological cells, including erythrocytes, leukocytes, platelets, and different subtypes of leukocytes, which is closely correlated with types and stages of many diseases and provides valuable information for disease diagnosis and personalized treatment [21]. Current methods for cell counting are mainly based on hemocytometers and flow cytometry, which require laborious manual enumeration and specialized instrumentation, respectively, and are limited in POCT. Compared with those conventional approaches, miniaturized chips for cell counting are portable, highly compact, user-friendly, and cost-effective, which fit well with the requirements of POC diagnosis. To date, a number of microfluidic devices based on different cell-counting principles have been developed, which can be divided into two categories, impedimetric and optical detection [21]. Impedance-based approaches are generally label free, which can differentiate different types and subtypes of blood cells by measuring cellular impedance and enumerate cells by electrical signals [22–26]. For example, Morgan and coworkers reported a microfluidic device that enabled blood sample processing and impedimetric cell counting, as shown in Figure 8.1a, which demonstrated good performance and promising utility for POC diagnosis [25]. Optics-based methods typically utilize light signals for quantification such as scattered light, fluorescence, and absorbance [27–30]. For instance, Land and coworkers developed a microfluidic cartridge allowing automatic blood sample preparation and visual cell counting, as shown in Figure 8.1b [27]. It showed results comparable to those obtained by flow cytometry and the hemocytometer, exhibiting potential use in POCT.

8.2 Cell

Blood Lysis

2

Quench

1

waste Impedance chip

Electronic detection

Microfluidic system

(a) (1) Blood sample introduction and inlet plugging

(3) and (4) Push/pull of air to transfer fluid to mixing chamber

(5) Magnetic stirring for mixing

(2) Blister pack compression and reagent release

(3) and (4) Push/pull of air to transfer fluid to visualization chamber

(b)

Figure 8.1 (a) Schematic illustration of the impedance-based microfluidic leukocyte analysis device. Source: Reproduced with permission from Han et al. [25]. Copyright 2012, American Chemical Society. (b) Working principles of the microfluidic cartridge. Source: Reproduced with permission from Smith et al. [27]. Copyright 2017, SAGE Publications.

8.2.2

Characterization of CD64 Expression

The upregulated expression level of CD64 on neutrophils is reported to be an important biomarker for sepsis [31, 32]. Despite significant progress in its diagnosis and treatment, sepsis remains one of the leading causes of death. Currently, the quantification of CD64 expression level is mainly achieved by flow cytometry, which requires cumbersome instrumentation and technical

241

242

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

Whole blood CD64 + cell depletion Lysing buffer Entrance count RBCs CD64 + cells Other WBCs

(a)

Exit count

Quenching buffer

CD 69

Main channel CD 64

as

Gas valve

CD 25

G

va

lve

Inl Ou

tle

(b)

Side channel inlet

t

et

CD54+ CD25+ CD69+ Other cells

Side channel outlet CD 64

cap

ture

Figure 8.2 (a) Schematic illustration of the microfluidic biochip for CD64 quantification. Source: Reproduced with permission from Hassan et al. [33]. Copyright 2017, Springer Nature. (b) Schematics of the multiregion microfluidic chip, surface modification of the main channel, and cell capture in one antibody region. Source: Reproduced with permission from Zhang et al. [34]. Copyright 2018, American Chemical Society.

experience. Lately, many efforts have been made to promote the POC diagnosis of sepsis. Among them, the microfluidic chip has become one of the most effective approaches. In general, the antibody-modified surface is utilized for the capture of cells with correspondence antigen. The portion of positive cells can thus be quantified. For example, Bashir and colleagues have reported a POC device that can measure the number of leukocytes and CD64 expression levels on neutrophils by integrating an anti-CD64-coated cell capture chamber, as depicted in Figure 8.2a [33]. Similarly, Pappas and coworkers developed a multiparameter affinity microchip, which integrated several cell capture regions modified by different types of antibodies, exhibiting improved performance for sepsis diagnosis, as shown in Figure 8.2b [34]. 8.2.3

Enumeration of CD4+ T Lymphocytes for HIV Monitoring

By 2017, there are more than 36 million people currently living with HIV. A significant method for monitoring and assessment of HIV status is counting CD4+ T lymphocytes numbers in the blood, which provides valuable clinical information for the diagnosis and therapy of HIV. As a gold standard, flow cytometry is the first choice. However, flow cytometry is very expensive, complicated, and requires experienced personnel to operate, which restricts its application. Therefore, over recent years, there have been a number of laudable developments in the field of rapid, convenient, and low-cost CD4+ T lymphocyte enumeration. The approaches can be divided into two categories: one way is counting the lysate of CD4+ T lymphocytes, and the other way is counting the CD4+ T lymphocytes directly. In the following, these two categories will be briefly discussed.

8.2 Cell

In terms of counting the lysate of CD4+ T lymphocytes, which usually contains several steps including separation, chemical lysing, and detecting the lysed cells signal. For example, Professor Xianbo Qiu et al. developed an easy-to-use microfluidic system for CD4+ T lymphocyte enumeration, which consists of disposable microfluidic chips and a low-cost companion actuation module (Figure 8.3a) [35]. There were three separate chambers for reaction, detection, and waste storage, respectively, in the enumeration system, which can detect CD4 automatically. A single polycarbonate bead modified with CD4 antibody was used to capture the CD4 antigen in the lysed sample. Then, the chemiluminescence intensity was measured for CD4+ T lymphocyte counting. Similarly, Professor Landers and coworkers reported on a cost-effective microfluidic platform integrating separation and enumeration via DNA-induced bead aggregation (Figure 8.3b) [36], which adopted a two-stage immunocapture microdevice. Firstly, monocytes were subtracted using anti-CD14 magnetic beads; afterward, CD4+ T-cells were captured by anti-CD4 magnetic beads. Secondly, the CD4+ T-cell lysate interacted with silica-coated magnetic beads to form aggregates. Thirdly, the resulting aggregate images were captured and processed to measure the mass of DNA, which was used to back-calculate the CD4+ T-cell enumeration. On the contrary, the other way is to isolate and detect the single CD4 cell based on its physical, optical, or electrical properties, and then the CD4 cell enumeration is measured by recognizing every CD4 lymphocyte one by one. For instance, Macdara T. Glynn et al. reported a low-cost microfluidic chip operated by dual-force CD4+ cell magnetophoresis (Figure 8.3c) [37]. CD4+ cells were separated by magnetophoresis, then the concentration of CD4+ cells was detected by bright-field inspection in the capture chamber. Another example is that Professor Beck and coworkers fabricated microfluidic chambers for cell counting based on fully printing methods, which can deposit hydrogel layers with fluorophore-labeled antibodies. This system can integrate on-chip sample preparation and reagent storage (Figure 8.3d) [38].

8.2.4

Circulating Tumor Cell (CTC) Isolation and Analysis

Since the discovery of circulating tumor cells (CTCs), they have been used as a “liquid biopsy” for peripheral blood analyses and an early biomarker for the diagnosis and prognosis of various cancers. Many studies were focused on the isolation and analysis of CTCs recently, and CTC separation methods can be divided into two major categories: antibody-independent or antibody-dependent. Antibody-independent methods allow the separation of CTCs based on the size, deformability, and electrical properties without antibodies. For example, Peng Li et al. reported an acoustic separation microfluidic system (Figure 8.4a) [39], which improved separation throughput and efficacy without antibodies. Moreover, this system was applied for the clinical blood samples of cancer patients successfully. Another example is that Tae-Hyeong Kim et al. developed an ultrafast and user-friendly centrifugal microfluidic technology (Figure 8.4b) [40], which can achieve highly sensitive size-based isolation of CTCs. As they

243

Vertical linear motor Reaction module

Magnetic module

Horizontal linear motor

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

Rubber pin

(b)

Waste storage module Vent hole

(a) 70 mm

P2

36 mm

Sample input

12 mm

=50 μm

Separation chamber 0.1 mm

600

9k

300

4.5k

0

0k

Fluorescence (CPS)

CD4 capture

Layer thickness (nm)

P1 press

76 mm

P1

(c)

26 mm

Image area 7.8 mm × 5.2 mm

Magnet

P1 release

244

0.1 mm

500 μm

(d)

Figure 8.3 (a) Integrated microfluidic system based on chemiluminescence. Source: Reproduced with permission from Qiu et al. [35]. Copyright 2018, Springer Nature. (b) Illustration of the cell enumeration chip via DNA-induced bead aggregation. Source: Reproduced with permission from Liu et al. [36]. Copyright 2016, The Royal Society of Chemistry. (c) Microfluidic chip operates by dual-force CD4+ cell magnetophoresis. Source: Reproduced with permission from Glynn et al. [37]. Copyright 2014, The Royal Society of Chemistry. (d) Fully printed microfluidic CD4 counting chips. Source: Reproduced with permission from Wasserberg et al. [38]. Copyright 2018, Elsevier.

showed, this method is a robust and cost-effective system for POC isolation of CTCs. In contrast, antibody-dependent methods use tumor-specific antibodies to identify CTCs. In these approaches, CTCs can be separated from other blood cells using magnetic or fluorescence markers. For instance, Myoung-Hwan Park et al. utilized a herringbone chip to isolate cancer cells from whole blood, on which a thiolated ligand exchange reaction with gold nanoparticles (AuNPs) was integrated (Figure 8.4c) [41]. In their system, the antibody-coated AuNPs were chemically assembled onto the chip directly. The whole process of this approach has slight impact on cell viability. Tina Saberi Safaei et al. also reported a microfluidic cell capture system to detect CTCs in whole blood sensitively, which was based on electrochemical ELISA method (Figure 8.4d) [42]. This approach is antibody dependent, which provides the clinically relevant specificity and sensitivity needed for a POC assay.

8.3 Nucleic Acid

Sheath flow

Sample

Sheath flow Battery

Tilted IDTs

Pressure sensor

Data acquisition board Bluetooth CTCs WBCs (a)

Waste

Collection

PC (c)

CTC

before GSH Cell release

Redox reporter

Blood

CTC (d)

Captured cancer cell

in 3 mins (b)

Figure 8.4 (a) Illustration of the acoustic chip for separation of CTCs. Source: Reproduced with permission from Li et al. [39]. Copyright 2015, National Academy of Sciences. (b) Schematic diagram and a photo image of a fluid-assisted separation technology (FAST) disk for the wireless measurement. Source: Reproduced with permission from Kim et al. [40]. Copyright 2017, American Chemical Society. (c) A herringbone chip using a thiolated ligand exchange reaction with AuNPs to isolate and release cancer cells from whole blood. Source: Reproduced with permission from Park et al. [41]. Copyright 2017, American Chemical Society. (d) Electrochemical ELISA method integrated within a microfluidic cell capture system. Source: Reproduced with permission from Safaei et al. [42]. Copyright 2015, American Chemical Society.

8.3 Nucleic Acid DNA or RNA amplification is one of the essential parts in nucleic acid detection. Nonisothermal amplification and isothermal amplification can be used to amplify and detect the target template that people are interested in. In order to get convincing and reliable target amplification results, there are some key points in POCT of nucleic acids such as reactant storage, reactant mixing, and temperature controlling [13]. These parts can be integrated into a single microfluidic platform. 8.3.1

Nonisothermal Amplification

The most successful nonisothermal amplification is polymerase chain reaction (PCR). PCR is a method that mimics the DNA replication system in the organism

245

246

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

to amplify a specific DNA/RNA target–template sequence from one or a few copies to a large number of copies with the same DNA sequence. From the past decades, there were some cheap, fast, handheld, and user-friendly PCR microfluidic devices with an accurate temperature cycling control module developed by researchers that met requirements of POCT diagnostics [43]. A microfluidic platform with unique ability to amplify PCR template in a low-cost thermoplastic format with cycle time as low as 14 seconds and full reaction in 8.5 minutes has been developed as shown in Figure 8.5a. There was accurate final melt analysis combined in this microfluidic platform [44]. A disk-formed microfluidic device by incorporating individually addressable diaphragm valves and a wireless heating system including laser irradiation combined with a small aluminum plate has been developed for PCR amplification. This device uses centrifugal force to pump liquids and achieves a fully automated manner to demonstrate the robust, reversible, leak-free, and thermally stable actuation of the valves during PCR amplification [47]. Some scientists developed an ultrafast rtPCR system that included microfluidic thermalization and indicated that the time of whole PCR reaction was lower than eight minutes as shown in Figure 8.5b. The method is based on the circulation of preheated liquids in a microfluidic chip that thermalizes the PCR chamber by diffusion and ultrafast flow switches [45]. According to the high-speed thermalization, this system could not only run PCR reaction in a few minutes but also perform sharp melting curve analyses. The scientists believe that this system with microfluidic thermalization can be quickly developed as a POCT system because of its high efficiency and lower cost. An all-thermoplastic integrated sample-to-answer centrifugal microfluidic lab-on-disk system (LoD) for nucleic acid analysis was presented by Roy et al. [46]. Their system included lysis, clarification, PCR amplification, exonuclease reaction, microarray hybridization, and washing functions as shown in Figure 8.5c. Some researchers also presented a POC nucleic acid device based on a droplet magnetofluid platform [48]. Using this droplet magnetofluid platform, it is easy to miniaturize and automate some chemical reaction and biological assay in the laboratory to the POCT by utilizing functionalized magnetic particles. The combination of novel thermoformed disposable cartridge as shown in Figure 8.6a and a portable multiaxial magnetofluid instrument as shown in Figure 8.6c makes it possible to detect and quantify the target nucleic acid in different samples by real-time PCR assays sensitively and specifically. Furthermore, this device is friendly for the end user. The user injects the interested sample into a cartridge and then load it into the device as shown in Figure 8.6b. 8.3.2

Isothermal Amplification

Even though a large number of PCR assay-based POCT microfluidic device has been presented, an accurate thermal cycling controller and additional external heater are still significant problems in the POCT device, especially in some developing countries. In order to deal with these problems, scientists developed some nucleic acid amplification methods without thermal cycle that are called isothermal amplification such as loop-mediated isothermal

8.3 Nucleic Acid

Heater electrodes

Thermistor voltage output

Air vent Hydrophilic valve

Thermistor current input

Reaction chamber

Sample in (a) Temperature (°C) 96

To + 24 °C

72 T0 0

20

40

60

80

100

120

Time (s) To + 24 °C

To (b)

t=0s

t = 1.6 s

Figure 8.5 (a) Schematic view of the two-layer thermoplastic PCR chip, revealing the detailed geometry of the thin film Cr/Au thermistor aligned to the center axis of the reaction chamber and the hydrophilic valve used to control sample loading. Source: Reproduced with permission from Sposito et al. [44]. Copyright 2016, The Royal Society of Chemistry. (b) The temperature during thermocycling for ultrafast rtPCR and temperature homogeneity of microfluidic thermalization during the heating phase. Source: Reproduced with permission from Houssin et al. [45]. Copyright 2016, The Royal Society of Chemistry. (c) Schematic of the disposable plastic device including nucleic acid parallel tests for point of care NA diagnostic. Source: Reproduced with permission from Roy et al. [46]. Copyright 2015, The Royal Society of Chemistry.

amplification (LAMP), recombinase polymerase amplification (RPA), nucleic acid sequence-based amplification (NASBA), helicase-dependent amplification (HDA), rolling circle amplification (RCA), strand displacement amplification (SDA), and so forth. A fixed temperature is required in these isothermal amplification methods, and these methods show significant potential in POCT device especially in some poor energy area. The LAMP is the reaction that includes an autocycling strand displacement DNA synthesis catalyzed by DNA polymerase with high strand displacement activity [49, 50]. Some scientists presented practical and low-cost polydimethylsiloxane (PDMS)/paper hybrid microfluidic device integrated with LAMP, which could detect bacteria without instrument rapidly and sensitively [51]. Compared

247

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses Exonuclease reagent reservoir

Deformable pinch valves areas

Lysing/clarification and metering chamber Mixing droplet based unit

Self-venting Waste

Overflow channel

Exonuclease reaction chamber PCR chamber Loading ports Hybridization microlluidic unit Hybridization salt resuspension

DNA microarray plastics parts

Washing reservoir and syphon-crest gate

PCR buffer reservoirs

Bottom CD

(c)

bu ash ffer

PCR b u

2 mm PMMA +PSA (Lasercut)

m

6m

Fluorescence

0.2 mm PMMA (Thermoformed)

m

40 m

(a)

(ii)

Thermal cycling melt

(+) HCV RNA

(–) HCV RNA Time

(b)

(i)

Temperature

buffer ing

r ffe

0.75 mm PMMA (Lasercut) 0.25 mm PTFE (Lasercut)

W

Figure 8.5 (Continued)

Bin d

248

(iii)

(c)

Figure 8.6 Droplet magnetofluid platform. (a) Fabrication of cartridge. (b) Brief idea of nucleic acid detection reagents storage and real-time PCR assay in cartridge wells. (c) Magnetic particle transfer in a portable multiaxial magnetofluid instrument. Source: Reproduced with permission from Shin et al. [48]. Copyright 2018, Springer Nature.

with the paper-free microfluidic system, the stability of LAMP amplification increases significantly during a longtime reaction in PDMS/paper hybrid microfluidic device as shown in Figure 8.7a. Furthermore, this device can be used on pathogenic microorganisms without preprocessing of the sample so that it can be used in a resource-limited area with no large-scale instrument requirement. This hybrid microfluidic device shows great potential in POC diagnosis. A

8.3 Nucleic Acid

microfluidic in-gel loop-mediated isothermal amplification (gLAMP) chip was developed [52]. This chip preloads the LAMP reagents with gel in a microfluidic chip and then stores in 4 ∘ C before sample loading as shown in Figure 8.7b. By integrating the LED array, this chip can detect pathogenic nucleic acid in near POC settings [52]. Nowadays, 3D printing is also a common method to fabricate microfluidic chip. An inexpensive full 3D printed microfluidic chip integrated with nucleic acid isolation membranes as shown in Figure 8.7c was reported [53]. Figure 8.7d shows a polymer/paper hybrid microfluidic biochip device integrated with LAMP for multiplexed instrument-free diagnosis of these three major types of bacterial meningitis [54]. This rapid, instrument-free, and highly sensitive microfluidic approach has great potential for POC diagnosis of multiple infectious diseases simultaneously, especially in resource-limited settings. A paper/poly(methyl methacrylate) (PMMA) hybrid CD-like microfluidic SpinChip integrated with DNA probe-functionalized graphene oxide (GO) nanosensors was developed for multiplex quantitative LAMP detection (mqLAMP) shown in Figure 8.7e [55]. This approach can simply and effectively address a major challenging problem of multiplexing in current LAMP methods. Combining this full 3D-printed microfluidic chip with LAMP reaction, low-cost system is pretty suitable to use in the resource-limited area. Except for LAMP, RPA and RCA are also widely used nucleic acid amplification methods in POC diagnosis. A solid-phase RPA-based silicon biophotonic-based detection sensor was developed [56]. This sensor can be used to detect Mycobacterium tuberculosis (MTB) rapidly and is called MTB isothermal solid-phase amplification/detection (MTB-ISAD) as shown in Figure 8.8A. This MTB-ISAD sensor is label-free and in a real-time manner, and it is potentially adaptable for better diagnosis across various clinical applications [56]. A rotary microfluidic device as shown in Figure 8.8B, which can detect single-nucleotide polymorphism (SNP) and point mutations using ligation-rolling circle amplification (L-RCA) reaction, was presented [57]. Five mutation points related to cancer prognosis can be identified according to this rotary microfluidic device. This valve-free microfluidic device is more portable and convenient to detect biomedical diagnostics in the resource-limited area. Also, other POC detection platform based on aptasensors were reported recently. A portable multiplexed bar chart SpinChip (MB-SpinChip) integrated with nanoparticle-mediated magnetic aptasensors was developed for visual quantitative instrument-free detection of multiple pathogens (shown in Figure 8.8C) [58]. This multiple important features of the MB-SpinChip are appealing as a universal POC platform for the multiplexed detection of pathogens and other biochemicals. In conclusion, an ideal device for multiplexed POCT should offer a high sensor performance, such as high sensitivity and multiplexing capability, as well as short turnaround times, at low system complexity, including low-cost fabrication and minimized user intervention. The future technology challenges will be the standardization and further miniaturization of the system components for the most effective use of this tool as a part of infectious diseases surveillance programs. It seems that the more widely used microfluidic devices will bring diagnostic testing closer to patients and will be a driving force for a shift from a centralized model to a decentralized patient-centered approach [59].

249

(a)

(b) Glass PDMS

PC

NC

Inlet N. meningitidis

Oil phase

Agarose solution

Degassing

LAMP zones

Reagent loading

Purge with oil

Incubation for 1 h

Purge with oil

Outlet Sample

Channel

Middle PDMS layer Chromatography paper

Agarose gel Sample loading

(c)

Alignment guiding structures

Part ll

LAMP zone Microchannel Top PDMS layer

NA membranes

Part l

Paper

3D printed parts

Packing NA isolation membranes (Qiagen TMSilica) Out

Photo polymer Plastic support

Align and attach UV light

Glass slide 3D printed reactor array for NAATs

Sample in

Permanent leak proof bonding via photo polymerization

Figure 8.7 (a) PDMS/paper hybrid microfluidic device combined with LAMP. Source: Reproduced with permission from Dou et al. [51]. Copyright 2014, American Chemical Society. (b) Workflow of self-contained microfluidic in-gel LAMP chip. Source: Reproduced with permission from Chen et al. [52]. Copyright 2017, Elsevier. (c) Schematic illustration of the assembly of multifunctional 3D-printed reactor array for LAMP amplification with leak-proof bonding. Source: Reproduced with permission from Kadimisetty et al. [53]. Copyright 2018, Elsevier. (d) Schematic of the PDMS/paper hybrid microfluidic biochip for multiplexed bacterial meningitis diagnosis. Source: Reproduced with permission from Dou et al. [54]. Copyright 2017, Elsevier. (e) Schematic of the PMMA/paper hybrid microfluidic SpinChip for mqLAMP detection. The top figure shows the 3D schematic of the exploded view of the SpinChip. Left figure at the bottom is the illustration of detection principle based on the interaction among the DNA probe-functionalized graphene oxide (GO) nanosensors, ssDNA probes, and target LAMP products. Right figure at the bottom is a photograph of the assembled PMMA/paper hybrid microfluidic SpinChip. Source: Reproduced with permission from Dou et al. [55]. Copyright 2017, The Royal Society of Chemistry.

(d)

(e)

Top plate

Inlet

Bottom plate LAMP zone

1 cm PDMS Glass

N. meningitidis Inlet S. pneumoniae Detection well

NC Hib

Inlet Screw

LAMP zone

Detection well

4 mm

Cy3-probe

Paper

LAMP zones Channel Outlet

GO Paper

Figure 8.7 (Continued)

Complementary LAMP product

252

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

(A)

On-chip MTB-ISAD (20 min)

(a) Mixing sample

(b) Sample loading

Primer ll

Input

(3) Amplification/detection Output

37 °C

RT

Light input

PMMA chamber Polymerase

Output Silicon microring resonators

Recombinase

(B)

(C) (a) DNA-probe immobilization

(b) Sample recognition

(c) Catalysis amplification 2H2O2

Magnet beads N

PtNPs

S

Multiplexed visual quantitation

Food Dye with different color

H2O2 Sample inlet

Connect and inject Capture sequence

Aptamer sequence

2H2O + O2

Spin and seal Platinum nanoparticles

Pathogen

Shake and read Control

S. enterica

E. coil

L. monocytogenes

Figure 8.8 (A) MTB-ISAD assay for fast MTB detection. Source: Reproduced with permission from Shin et al. [56]. Copyright 2015, Elsevier. (B) A photograph of the portable L-RCA rotary microdevice. Source: Reproduced with permission from Heo et al. [57]. Copyright 2016, Elsevier. (C) Schematic of the multiplexed bar chart SpinChip (MB-SpinChip) procedure. (a) DNA probe immobilization: H2 O2 substrate solutions (light blue region), food dye solutions (yellow, dark blue, red, and green circle), and DNA probes (light gray circle) are, respectively, prestored in the MB-SpinChip. Herein, magnetic DNA probes are immobilized in different sample recognition microwells by a magnetic field (dark gray circle). (b) Sample recognition: pathogens specifically combine with PtNPs–aptamers to form complexes that are then released into sample solutions (purple circle). (c) Catalysis amplification: sample solutions and H2 O2 solutions are mixed to generate O2 with the pressure increase inside, resulting in the internal pressure increase that further leads to the food dye to move into channels to form different bar chart signals for visual multiplexed pathogen detection. Source: Reproduced with permission from Wei et al. [58]. Copyright 2018, American Chemical Society.

8.4 Protein

8.4 Protein The ability to molecularly detect proteins remains the pinnacle technique that has historically propelled us into the modern age of medical diagnostics. Ever since Paul Ehrlich first coined the term “antibody” more than a century ago [60], technologists and medical professionals have fully harnessed this tool borrowed from immunology and designed an entire spectrum of diagnostic tests built upon the concept of antibody–antigen interactions [61]. From Coombs tests and radioimmunoassays to sandwich immunoassays and gold-amplified pregnancy tests, these inventions have all been created to allow us to “see” into the world of proteins. One such invention, ELISA, serves as the current clinical gold standard for informing medical decisions for a large majority of diseases. Therefore, it is unsurprising that since the origination of microfluidics around three decades ago, countless developments have been put forth marrying the myriad of protein detection methodologies with the advantages of miniaturization, thus transforming a technique that is traditionally confined in wet lab spaces, into a lab-on-a-chip format [1]. However, despite many state-of-the-art works that have detailed elegant and commercially promising microfluidic-based platforms for POC protein detection in resource-limited settings, many weaknesses and limitations remain [62–65]. Firstly, despite the reliability and robustness of on-chip immunoassays in improving the overall detection characteristics (e.g. sensitivity, specificity, and reaction time) compared to conventional methods, microfluidic detection of ultralow analyte concentrations remains a key challenge, especially owing to the reduced sample volumes and the lack of integrated antigen pre-enrichment modalities. Next, mainstream microfluidic chips are predominantly made of PDMS, which has not been wholeheartedly accepted by the medical community, as its material properties differ from those that are currently FDA-verified for use in in vitro diagnostic purposes. Furthermore, another challenge in microfluidics is that chips typically require connections to peripheral fluidic regulatory elements (e.g. pumps, pressure sources, and valves) to be able to function, especially for performing multistep diagnostic tests, such as immunoassays, rendering them confined as a “chip-in-a-lab,” rather than a truly standalone device that can be readily utilized for diagnosis at the POC. In this chapter, we aim to provide a brief account of some seminal works in microfluidic-based protein detection that have adopted highly promising strategies in overcoming some of the challenges previously mentioned. Although nowhere near an exhaustive list, we posit that these are key areas that would be particularly significant in the development of next-generation POC devices for carrying out immunodiagnostics tests. 8.4.1

Novel Chemistry and Nanomaterials

One of the limiting factors with POC diagnostics is the lower signal-to-noise ratio for simple assays compared to other more complex analytical techniques. Device sensitivity is especially important in cases where early disease detection and diagnosis are critical factors in effective treatment and patient prognosis. In this section, we will discuss novel chemistries and nanomaterial techniques

253

254

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

that have increased the signal-to-noise ratio for protein analytical techniques within the past few years. These techniques include signal amplification using novel labeling techniques and concentrating targets through capturing or volume reduction. Novel labeling techniques, such as quantum dots [66], enzyme amplification [67], and silver precipitation, have dramatically increased the signal and sensitivity. Quantum dots are advantageous in that they have a narrow emission spectrum, broad excitation, are photostable, and have a high quantum yield and have been used to increase the assay sensitivity by 4 orders of magnitude [68]. Recent enzyme techniques have included using polydopamine (PDA) [69] and streptavidin–biotin–peroxidase [70] for signal amplification. Silver precipitation has also been shown to help signal amplification in electrochemical readout devices [71, 72]. Next, protein capture devices using zinc oxide nanowires [73, 74], silicon nanowires [75], and carbon nanotubes [76], magnetic beads [77, 78], and membranes [79]. Concentration devices are advantageous in that they can capture a large amount of the target analyte but have limitations in that clogging can occur easily, and for nonlabeled filtration, it is size dependent and can trap other contaminants. We will discuss two novel works related to signal amplification via labeling and concentration via filtration. One technique is to increase the signal by using novel chemistry for labeling. A promising technique called enzyme-accelerated signal enhancement (EASE) is a chemical-based method that uses PDA to increase the bioconjugation of reporter molecules for signal amplification [69]. In particular, they showed that using PDA to coat targets before adding horseradish peroxidase (HRP) increased the limit of detection (LOD) by 3 orders of magnitude (1217 times greater than using standard ELISA) and had a sensitivity of 3 fg/ml. The PDA has high reactivity with amine, sulfhydryl, and phenol groups in proteins that helps to locally coat the target complex. The sensitivity and LOD were demonstrated using a standard sandwich ELISA for mouse IgG with a tetramethylbenzidine (TMB) substrate in a well format as shown in Figure 8.9a. Without the EASE technique, the color was detectable in the range of 10−7 to 10−8 g/ml; however, when they included the EASE technique, the sensitivity increased to 10−12 g/ml. Furthermore, without EASE, the LOD was 108 pg/ml and improved to 85.3 fg/ml with the addition of EASE (1266 times). Next, the technique was used to investigate four disease biomarkers: HIV capsid antigen p24 (HIV p24), kallikrein 3 (KLK3), c-reactive protein (CRP), and vascular endothelial growth factor (VEGF). The following assays had LODs of 2.87 fg/ml, 0.31 pg/ml, 0.24 pg/ml, and 11.5 fg/ml, respectively, with an average LOD improvement of 1217-fold with the EASE ELISA technique over their standard ELISA equivalents. Finally, the technique was converted to a lateral flow assay (LFA) with the HIV p24 target and enabled detectability down to 10 pg/ml with the naked eye (1000× improvement). This signal amplification is versatile and has the potential to greatly increase the sensitivity and LOD of POCT, enabling earlier detection in an easily accessible format. Another method is to increase the signal by capturing and enriching targets. Yin Ting Yeh et al. demonstrated a device using a carbon nanotube-sized tunable enrichment microdevice (CNT-STEM) to capture viruses and enrich them to increase diagnostic sensitivity [76]. Nitrogen-doped multiwalled carbon nanotubes (N-MWCNT) were grown on a silicon substrate and integrated

8.4 Protein

(a)

l

na

io

nt

e SE nv EA Co

Sandwich detection 10–3

EASE Concentration Detectability

Conventional

10 ng/ml

10 pg/ml

10 ng/ml

10 pg/ml

+

+

+

–

10–4 10–5 EASE HR P

10–6 Signal development

10–7 HIV p24 test line

10–8 10–9 –10

Capture Ab HRP

10 10–11 10–12 10–13 0

Detection Ab Target

(b)

Sample in

(g/ml)

Original (EID50) Enriched (EID50)

CNT-STEM enrichment

Filtrate out

104 103 102

104 103 102

PDMS

Flow

TEM

SEM Silicon

Virus

Aligned N-MWCNT

Contaminants

Figure 8.9 Novel signal amplification and analyte enrichment strategies applicable to microfluidic devices. (a) Polydopamine-based enzyme-accelerated signal enhancement, or EASE, used to improve the detection sensitivity of conventional immunoassays. Source: Reproduced with permission from Chin et al. [65]. Copyright 2011, Springer Nature. (b) Carbon nanotube size-tunable enrichment microdevice or CNT-STEM for size-based capture or viruses. SEM image (scale bar, 100 nm): H5N2 AIV virions trapped inside the aligned N-MWCNTs (nitrogen-doped multiwalled CNT). Dark-field TEM image (scale bar, 100 nm): enriched H5N2 AIV after the aligned N-MWCNTs structures were retrieved from the CNT-STEM. The right figure is the Dot-ELISA detection of H5N2 AIV after virus cultivation in embryonated chicken eggs. Source: Reproduced with permission from Guo et al. [72]. Copyright 2015, Royal Society of Chemistry.

with a PDMS device for confining the fluid flow through the filtration unit as shown in Figure 8.9b. The internal spacing between N-WMCNT was accurately tuned by varying the thickness of iron catalyst thin films (1–12 nm) deposited on the substrate, which resulted in a decrease in iron particle density and increases in intratubular spacing for increased iron catalyst thicknesses. Using this technique, they were able to tune the spacing between N-MWCNTs from 17 ± 6 to 325 ± 56 nm with average diameters of 17–99 nm. The CNT-STEM device filtration performance was compared by trapping 20, 50, 100, 140, 400, and 1000 nm sized fluorescent nanoparticles within 25, 95, and 325 nm intratubular-sized CNT-STEM devices and comparing the original and filtrate solution fluorescence. Next, the enrichment performance was tested using low pathogenic avian influenza strain (LPAIV) (A/duck/PA/02099/2012(H11N9)), which had a size of 93 ± 35 nm. Capture efficiency was determined using both an indirect fluorescent antibody (IFA) assay for qualitative analysis and real-time reverse transcriptase polymerase chain reaction (rRT-PCR) for quantitative

255

256

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

analysis and were determined to be 96.5 ± 0.5%, 88.0 ± 0.3%, and 57.5 ± 0.4% for 25, 95, and 325 nm intratubular-sized CNT-STEM devices, respectively. The device also improved the detection sensitivity by 2 orders of magnitude via the enrichment process. Therefore, this technique has promise in increasing the detection sensitivity of low-abundance targets. 8.4.2

3D-Printed Microfluidic Devices

Recent developments in 3D-printed milli-, submilli-, and microfluidics have welcomed much promise in revolutionizing disease diagnostics [80, 81]. In contrast to existing PDMS-based microfluidic chips and commercial fluidic systems, stereolithographic (SLA) and fused deposition modeling (FDM) 3D-printing of devices enables rapid design-test cycles, low-cost cleanroom-free in situ fabrication, freeform embedding of 3D elements on-chip, and utilization of industrially validated chip materials [82]. These are critical parameters that render these platforms as ideal candidates for enabling POC diagnostics. As such, owing to the advantages of 3D printing, over the past few years, we have seen an enormous spur in academically developed 3D-printed platforms. Many seminal works that have been published detail 3D-printed devices incorporated with various on-chip components, such as pumps [83–85], mixers [86, 87], membranes [88, 89], ,and electrodes [16, 90]. The first generation of 3D-printed devices typically showed demonstrations of simple homogenous assays, such as enzymatic glucose detection [91, 92]. Our group has demonstrated total protein quantification on-chip using finger-powered components [93]. However, as mentioned earlier, complex heterogenous diagnostic assays, such as sandwich immunoassays and ELISA tests, strictly involve multiple wash/incubation and development steps, making them much more difficult to implement in a POC setting, even in current microfluidic devices. We will see here how 3D-printing has infused much potential for solving such problems by enabling the construction and integration of new features and architectures. One formative example of 3D-printed microfluidic devices for immunosensing was demonstrated by Rusling’s group [94]. Using SLA 3D-printing, they fabricated a multicomponent fluidic device that consisted of reagent storage, passive mixer, and detection chamber (which interfaced with an antibody array printed on a glass substrate) components. Having serpentine channel, reagent reservoirs allowed multiple immunoreagents and washed buffers to be discretely stored in series and pumped sequentially by using a programmable syringe pump. Aimed toward POC detection of cancer-related biomarkers (platelet factor 4 [PF-4] and prostate-specific antigen [PSA]), the operation of their device was designed to be simple and automated, only requiring the user to pipette the sample into the first reagent reservoir and turning on the setup. Detection limits for both proteins on their device were 0.5 pg/ml. Furthermore, the freeform fabrication capabilities offered by 3D-printing allowed the entire device to be fabricated monolithically with three-dimensional complex architectures (e.g. an embedded 3D mixer network). Chemiluminescent readout was performed by using a simple camera for imaging the glass array after the completion of the assay (∼30 minutes). Using this concept as a platform for developing further devices with enhanced sensitivity for

8.4 Protein

testing multiple antigens simultaneously, Rusling and coworkers further explored the integration of such a 3D-printed serpentine-format flow layer for multiplexed immunosensing based on nanomaterial-enhanced immunointeractions (echoing our earlier discussion on nanomaterials and signal enhancement) as shown in Figure 8.10a [95]. Instead of interfacing with a glass slide array, they replaced the detection substrate with patterned graphite sheet with printed capture antibody spots. With a more complex detection involving generation of electrochemiluminescent signals using RuBPY-silicon nanoparticle-conjugated detection antibodies, up to seven cancer biomarkers (IGF-1, PSA, PF-4, CD-14, VEGF-D, GOLM-1, PSMA, and IGFBP-3) were quantified down to concentrations as low as 78–110 fg/ml in diseased human serum samples. Essentially, this format easily allows different reagents to be stored in different chips depending on the application, thus allowing rapid integration of reagents without the need for excessive tubing and fluidic connections. Additionally, multiplexed protein detection using the device can be conducted in parallel by simply scaling the number of devices and external pump connections [96]. However, owing to the serial nature of the fluidic pathway within individual devices, one disadvantage of this technique is that successive reagents that flow along the same region may be contaminated if there is any liquid residue left over after dispensing the previous liquid segment. Another SLA 3D-printed fluidic architecture that has enabled the translation of protein detection assays was demonstrated by Tanner’s group [97]. Utilizing the rapid prototyping capability of 3D-printing, they were able to demonstrate a workflow verifying two (syringe-type versus chamber-type) device designs simultaneously. Furthermore, instead of relying on antibodies as in traditional immunoassays, their design included the use of novel aptamer-based assay, termed “aptamer-tethered enzyme capture,” or APTEC. By utilizing aptamer-conjugated paper disks or magnetic microbeads, they could detect concentrations of a malaria biomarker, Plasmodium falciparum lactate dehydrogenase (PfLDH), down to 5 and 50 ng/ml, respectively. Practically, the device enables users to easily manipulate the beads to be dispensed into discrete chambers containing successive washing and color development reagents by only using a handheld magnet-containing sheath. Thus, their setup negates the need for any bulky or expensive fluid handling equipment and peripherals (e.g. pipettes, syringe pumps, compressed gas, and liquid handling robots), which represents one of the primary pitfalls associated with performing diagnostics on conventional 96-well plates or microfluidic chips. Recently, Tanner et al. have also shown that a similar device setup that exploits the laminar flows in a microfluidic geometry to operate a device with one continuously adjoined chamber but with different reagent zones as shown in Figure 8.10b [98]. Thus, the APTEC assay can be carried out with the minimal sample and solution volumes, and in an enclosed manner without the need for physically removing the magnetic beads from the device, minimizing contamination and exposure risk to the end user. Testing their new device with patient samples, the authors were able to establish promising sensitivity (90%) and specificity (90%) levels. Although the field of 3D-printed microfluidic assays has focused on SLA 3D-printers owing to their innate ability to fabricate leakproof devices with resolutions around 100–200 μm (Nordin and coworkers have even developed

257

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

Micropump inlet

30 mm

Tripropylamine

Reagent chamber Wash buffer with inlet port RuBPYVent holes SiNP-Ab2 Detection chamber

Counter electrode (platinum)

Wash buffer Sample (serum)

Detection platform

Reference electrode (Ag/AgCl) PSA

CD-14

GOLM-1

Outlet

BSA Analyte Ab1

ECL

IGFBP-3

SWCNT

IGF-1

PF-4 PSMA VEGF-D Pyrolytic graphite sheet-detection platform with patterned microwells

TrPA + 1.0 V

RuBPY-SiNP-Ab2

(a) Microfluidic APTEC biosensor

Cross section of biosensor 25 mm

40 mm

Primary layer Secondary layer

2 mm

500 μm 25 mm

258

Detection of P.falciparum from 5 μl of blood PfLDH positive

PfLDH negative

(b)

Figure 8.10 3D-printed microfluidic devices for the detection of protein biomarkers. (a) 3D-printed serpentine flow chamber coupled with graphite sheet detection array for multiplexed nanoparticle-enhanced immunoassays. Source: Reproduced with permission from Comina et al. [91]. Copyright 2015, John Wiley & Sons, Inc. (b) 3D-printed aptamer-tethered enzyme capture (APTEC) device for malaria diagnosis using magnetic microbeads. Source: Reproduced with permission from Tang et al. [94]. Copyright 2017, Royal Society of Chemistry.

an advanced system capable of making 20 μm × 20 μm channels [99]), they are expensive and the use of proprietary photocurable resins raises questions about surface biocompatibility and long-term use in tandem with reactive chemicals. On the other hand, FDM 3D-printers are much cheaper, accessible, and easy to adopt for nontrained individuals and organization. Additionally, they use commonly found thermoplastic materials such as polylactic acid (PLA) and acrylonitrile butadiene styrene (ABS). Hence, it is unsurprising that some groups have also demonstrated integrated immunoassay devices using FDM-printed platforms. Kulinsky and coworker published his work detailing a fully automated ELISA platform for the qualitative immunodetection of Plasmodium-specific antibodies [100]. Assembly of their device consisted of a disposable 3D-printed cartridge with depressible elastomeric domes as reagent reservoirs and a reusable control frame with Arduino-controlled servomotors and batteries. Although this current device is not as “miniaturized” compared to SLA-based devices, it would worthwhile to explore how the accessibility of FDM 3D-printers can

8.4 Protein

fulfill niche applications whereby aspects of miniaturization (e.g. sensitivity and portability) are not as vital as compared to the wide distribution and availability (e.g. large-scale automated screening in remote hospitals). 8.4.3

Digital and Droplet Microfluidics

Digital microfluidics (DMFs) is another platform that shows promise for POC diagnostics, which consists of discrete droplets in an open format. Various techniques include sessile droplet arrays, electrowetting-on-dielectric devices, and magnetic–fluidic droplets. The categories can be further classified into passive and active methods based on the external forces used for fluid manipulations. Active devices are typically self-contained, automated, and can use external forces such as electrowetting-on-dielectric [101, 102], surface acoustic waves [103], thermomechanical [104], magnetic fields [105], and mechanical motion [106]. Passive devices, on the other hand, are densely packed arrays of droplets but usually rely on fluid handling robots or manual pipetting. Common detections methods are ELISA and chemiluminescence, but a few novel methods include concentrating droplet arrays via evaporation [107], coffee ring effect [108], length-based measurement [108], and measuring electrical impedance [109]. Advantages to DMFs include being able to perform complex multistep assays (dispensing, combining, splitting, and washing), adaptable format, and automation. Some disadvantages include needing external equipment and fluidic handling, still needing basic resources (power, cold storage for reagents), and needing trained personnel to setup and perform the test. Despite these disadvantages, there has been some work recently to make open source platforms (DropBot [110] and OpenDrop [111]) as well as to make the systems more portable [112]. Wheeler’s group demonstrated a DMF platform for performing stereological immunoassays in remote settings [112]. The platform, termed the Measles-Rubella box (MR box), consists of a DMF cartridge, controller electronics, high-voltage amplifier, magnetic lens, light source, optics, and photomultiplier tube (PMT) for signal detection. All the components fit into a 25 cm × 20 cm × 28 cm box as shown in Figure 8.11a, weighed 4 kg, cost less than $2500, and could be powered off a 12 V laptop power supply. The DMF cartridges were fabricated by printing silver conductive ink using an inkjet printer, applying a dielectric layer using chemical vapor deposition (CVD) and spin coating a hydrophobic layer. The active units were then sandwiched between a normal glass slide (bottom) and indium tin oxide (ITO) glass slide (top). The cost of each DMF cartridge cost was $6. The DMF cartridge was used to perform four ELISAs to detect rubella immunoglobulin G (IgG) and measles IgG simultaneously and the assays were completed within 35 minutes on diluted whole blood samples. Paramagnetic particles coated with the target antigen and were used for capturing the antimeasles and antirubella IgG. Subsequent steps were used to wash and amplify the signal using antihuman IgG HRP conjugate and chemiluminescence readout. The platform was tested in the laboratory and had LOD and limit of quantification (LOQ) of 0.14 and 0.55 mIU/ml for measles IgG and 0.15 and 0.20 IU/ml for rubella IgG, respectively. Later, when performing a field test in northern Kenya (Kakuma refugee camp), the device sensitivity was

259

260

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

3

1

Hydrophobic layer Dielectric layer Inkjet-printed silver Flexible printing media Adhesive tape

ITO glass Droplet Glass slide

2

4 1

2

Magnetic particles

4

5 mm

Magnetic particles

Blood sample

5

3

Immobilized particles

Absorbent waste wick

Blood droplet

6

Particles in wash buffer

Conjugate H2O2

7

8

Luminol

9 Detection zone

(a)

Figure 8.11 Drop-based microfluidic devices for point-of-care immunoassays. (a) Field-tested digital microfluidic setup, termed Measles-Rubella box (MR box) for performing magnetic microparticle-based ELISAs on diluted whole blood samples. Source: Reproduced with permission from Zhang et al. [108]. Copyright 2018, Royal Society of Chemistry. (b) Active droplet array (ADA) for multithroughput multistep assays using magnetic microbeads. Here, the footnotes “0” and “1” represent the inactive and active status of the interaction zones (T, thermal; M, magnetic; and O, optical). The linear dynamic range of sample-to-answer CL immunoassay for CRP (c-reactive protein) and PCT (procalcitonin) are shown in Figures 8.4 and 8.5. Source: Reproduced with permission from Mok et al. [109]. Copyright 2014, National Academy of Sciences.

determined to be 86% (measles) and 81% (rubella) with the specificity of 80% (measles) and 91% (rubella) when compared with commercial tests in a laboratory setting. Therefore, this platform offers the ability to perform complex testing of immunoassays in a controlled and automated way in limited resource settings. Liu’s group developed a water-in-oil droplet device, termed active droplet array (ADA), which was used to perform multistep assays by moving magnetic particles through aqueous droplets [113]. The device had an open top with parallel units of microslits connected to tandem wells as shown in Figure 8.11b. Each unit was preloaded with the required reagents and encapsulated in oil. The device could be oscillated to perform mixing, raise and lower a magnet to

8.4 Protein

Blood serum

Mineral oil

1

MB-Abs

2

3

HRP-Abs

Washing buffer

Step 1

Step 2

Step 3

T1 M0 O0

T0 M1 O0

T0 M1 O0

Reaction

Substrate Step n

T0 M0 O1

Detection

Washing

λ = 425 nm

O0

O0

M1

M0

Target protein

Y = 6365*X – 1844

1.2 × 1006

R2

= 0.993

8.0 × 1005 4.0 × 1005

HRP-Abs

3.2 × 1006 CL intensity (a.u.)

5

0

Y = 46655*X + 52863 2.4 × 1006

R2 = 0.993

1.6 × 1006 8.0 × 1005 0

0

(b)

M0 MB-Abs

1.6 × 1006 CL intensity (a.u.)

4

O1

50

100

150

CRP concentration (ng/ml)

200

0

15

30

45

60

PCT concentration (ng/ml)

Figure 8.11 (Continued)

move magnetic beads through subsequent wells, and had optical imaging well for quantification. The microslits allowed the magnetic beads to pass through to the next well but confined the aqueous droplet to the previous well. To test the device, CRP and procalcitonin (PCT) targets were tested from serum samples using a sandwich immunoassay. First, the serum sample was added into a well with magnetic bead antibodies and HRP-conjugated detection antibodies and oscillated to mix. Next, the magnetic beads were passed through two subsequent wells to perform wash steps for removing unbound HRP. Finally, the magnetic beads were moved to the well with HRP substrate droplets and the chemiluminescence was recorded with a PMT. The CRP sample (3 μl) had a

261

262

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

linear detectable range (LDR) of 0.8–200 ng/ml with an LOD of 0.31 ng/ml while PCT sample (30 μl) had an LDR of 0.5–60 ng/ml and an LOD of 0.09 ng/ml. This device demonstrated a parallel platform (15 units) that could perform sequential steps (up to 5) in a simple format.

8.5 Metabolites and Small Molecules Products of metabolism that produce energy, process wastes, and nutrients or decompose and renew the tissues in the body are called metabolites. These can be treated with ionic blood chemicals such as H+ , K+ , Cl− , Na+ , HCO3 − , etc., and small molecules such as nonprotein hormones such as cortisol, epinephrine, and peptide hormones for POC assays. In fact, levels of hormones, metabolites, as well as blood-borne chemicals can be used as diagnostic disease indicators. Glucose, cholesterol, creatinine, triglycerides, urea, lactate, and ammonia are the most common metabolites that have been used for POC diagnostic purposes [5]. In this part, recent developments in microfluidics used for POC diagnostic applications that utilize small molecules as analytical targets would be explained. Sweat is an interesting biofluid that can be collected in a noninvasive manner and contains important biomarkers such as electrolytes, small molecules, as well as proteins [114]. Measuring the in situ concentration of such biomarkers provides important information on physiological health status (such as hydration state) as well as for the diagnosis of disease (such as cystic fibrosis) [115, 116]. However, existing systems for sweat collection of whole-body have been limited to the laboratory because of the need for composition analysis of samples using standard chemical analysis technologies [117]. Thus, attempts have been made to detect and collect sweat at the same time by direct contact with sensors (paper-based substrates) on the skin for electrochemical and/or optical evaluations to detect chemical components such as sodium ions and lactate in real [118, 119]. Also, sweat pH can be achieved via colorimetric responses in functionalized porous substrates and also quantitative assays can be performed using systems that are capable of capturing high-quality digital images including smartphones [120–122]. Besides, sweat generation rate can be measured by the integration of radio-frequency identification systems with porous materials. However, inaccuracy of the total sweat rate and volumetric loss may occur as the sweat gland density, and overall areas are typically unknown [123]. On the basis of the above improvements in achieving information from sweat, a thin and soft microfluidic system could be fabricated, which is capable of direct and reliable sweat harvest from the skin surface as shown in Figure 8.12. This device used the collected sweat for multiparametric sensing of biomarkers and had options for wireless interfaces to external devices for image capture and analysis [114]. Because of its capability to be mounted at different locations on the body without any irritation because of its biocompatibility, flexible and stretchable properties, as well as watertight interfaces, it can be considered as an important development in wearable microfluidic platforms that could measure the total sweat loss, pH, lactate, chloride, and glucose concentrations using colorimetric detection via wireless data transmission [114].

(A)

(B)

Microfluidic system (middle)

Top side

Back side

Image process marker Black reference

Adhesive Outlet

NFC electronics

Chloride

Lactate White reference

(C)

Glucose

pH (a)

Cover PDMS

(b)

Inlet / harvesting area

Water detection (b)

(a)

Colorimetric dye

PDMS microfluidic channel

(D) NFC chip

Top PI

Adhesive layer

Cu bridge

Middle PI

Epidermis

Cu coll Bottom PI

1 cm (dia. 3 cm and ~700 μm thickness)

Figure 8.12 (A) Schematic design of the microfluidic sweat monitoring device with the integrated system of near-field communication (NFC) (inset). (B) Presentation of the top, middle, and back sides of the microfluidic device. The white and black colors (reference color markers) are located on the top side, with the NFC electronics. In the middle, the channels with colorimetric assay reagents such as water, chloride, lactate, PH, and glucose are positioned. The bottom side contains an adhesive uniform layer that is bonded to the bottom surfaces of the channels enclosed with PDMS, with some openings that provide sweat access as well as inlets that connect to the mentioned channels. (C) Cross-sectional diagrams of the cuts defined by the dashed lines (a) and (b) shown in the top side illustration in (b). (D) Optical image of a fabricated device mounted on the forearm. Source: Reproduced with permission from Koh et al. [114]. Copyright 2017, American Association for the Advancement of Science.

264

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses (a) 1 cm

(b)

Exploded view

Top view

Chamber

Microfluidic channel Bottom layer Adhesive Layer

Epidermis

Inlet

Outlet hole

Figure 8.13 (a) Schematic pictures and optical images of soft, thin microfluidic devices for sweat chrono-sampling. (b) Exploded view design of the device and its interface with skin. (c) Top view sketch of microfluidic channels that are filled with blue-dyed water. Source: Reproduced with permission from Choi et al. [124]. Copyright 2017, John Wiley & Sons, Inc.

Another important development in this field has been achieved recently by fabrication of a thin, soft, “skin-like” microfluidic system for quantitative determination of biomarkers via colorimetric analysis as shown in Figure 8.13 [124]. This device has extra features in chrono-sampling of sweat, for specification of variations in biomarkers with time. It also provides collection chambers for sweat extraction and ex situ analysis [124]. In this work, the pressure from sweat glands due to natural differences of osmolality between plasma and sweat has been utilized to actively driving flow into networks of microchannel and microreservoirs. Besides, to guide flow in the microfluidic device, a collection of carefully designed capillary bursting valves (CBVs) are used by which the flow of sweat can be directed to fill a collection of microservers sequentially, which leads to a precise sampling capability. Results confirmed that such a skin-compatible platform allows efficient means for storage and final extraction of discrete sweat samples. Human studies demonstrated applications of this device in the precise chemical analysis of lactate, sodium, and potassium concentrations as well as their temporal variations [124].

8.5 Metabolites and Small Molecules

(a)

1 cm

(b)

1 cm

(c)

1 cm

Microfluidic channel layer

SAP layer

Cover layer

Adhesive layer

(A)

Colorimetric assay papers

SAP

Epidermis

(B)

Figure 8.14 (A) The exploded schematic design illustrates the device structure. (B) Optical image of the microfluidic platform (a) and its mechanical flexibility under bending (b) as well as twisting (c). Source: Reproduced with permission from Kim et al. [125]. Copyright 2018, John Wiley & Sons, Inc.

Another achievement in this field is the utilization of valves based on superabsorbent polymers in skin-mounted soft microfluidics for time-sequenced discrete sampling and chloride analysis by colorimetric chemistries [125]. Using this microfluidic as shown in Figure 8.14, it is possible to precisely get chronographic information on sweat chloride as well as sweat loss. The superabsorbent polymer, which is used in valve construction, is on the basis of sodium polyacrylate. Using such water-actuated valves as well as patterned regions of hydrophilic and hydrophobic channel surfaces in a multilayer format, the device is capable of time-sampling by selective isolation of sweat in individual reservoirs in which there is no backflow of sweat but air ventilation is allowed. This approach can be considered as a suitable substitute for passive CBV methods. In addition, for stabilization of color development and providing accurate quantitative analysis on chloride concentration, noninteracting ions and surfactants were used in this platform. Because of the in vivo results, it is claimed that the mentioned microfluidic device is useful for real practical usage in many application fields such as screening for cystic fibrosis as well as exercise physiology [125]. Another important issue in the fabrication of POC diagnostic devices is the readout mechanism that would be used in such devices. Development of cheap quantitative assays that can be used by everyone everywhere without using any supplementary constituents such as specialized readers is a critical challenge in the area of POC as well as point-of-use diagnostic. Fabrication of a quantitative readout, which necessitates no instruments, is still a big challenge as most of the common readouts such as fluorescent, colorimetric, as well as electrochemical assays need some electronic devices to quantify the intensity of the results

265

266

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

from the assays. To address this need, measurements of time requires for the appearance of the color has been utilized as a novel readout for a quantitative paper-based assay for active enzymes. Therefore, time-based methods can be considered as suitable approaches to obtain readouts without using any external devices for output quantification [126]. The mentioned time-based method for readout was used for assessment of small molecules such as adenosine and inorganic ions such as Hg2+ and Pb2+ for measuring heavy metal contamination in water [127]. As aptamers were used in this research, this achievement can be used for evaluation of new classes of analytes. Moreover, by a simple exchange of the reagents in the platform, the device can be used for other target analytes and also the sensitivity as well as a dynamic range of the assay(s) can be tuned conveniently. Another important development in this device is the capability of this platform to be used for different assays by just adding a single sample aliquot onto the device and receiving the results within one-step sample testing. The assay platform contains the required reagents for analyte detection as well as signal amplification, and there is no need for additional washing or any other manipulations. Finally, the device measures the sample volume for quantitative readout. The paper-based platform includes hydrophobic regions that determine where the sample moves by capillary action via hydrophilic areas. By visualizing the green color in the hydrophilic area of the microfluidic chip after adding the sample in the center part of the device, the user should start measuring the time needed for the stop region to turn red. This time reveals the analyte concentration in the sample [127]. Another important improvement in the readout mechanism is utilizing distance-based detection that was developed in a paper analytical device for metal (Ni, Cu, and Fe) quantification purposes [128]. This is an impressive achievement as it eliminates the need for using external optical instruments such as camera or scanner for colorimetric analyses. Although visual detection can be performed via a color/intensity comparator, the perception of brightness and color type would be different from person to person, and thus, distance-based detection approach can be a suitable substitute for such limited techniques. In this method, the length related to a colored reaction product through a paper channel would be read simply. The device includes a sample reservoir as well as a channel that is patterned by a colorimetric indicator for the analyte as shown in Figure 8.15. When the analyte flows through the channel, the indicator precipitate and the analyte would react and produce precipitate with a specific color. The analyte content is proportional to the length of this color band that can be detected by a ruler printed near this channel. It was shown that the device could be used for simultaneous measuring of Ni, Fe, and Cu from different sources of metal particulates with suitable detection limits [128]. Later, scientists utilized a cost-effective and easy-to-use forehead thermometer as a readout for quantitative POCT [129]. This thermometer can be used to read the temperature signal that is generated by a specific analyte–aptamer interaction. The device includes a channel for fluid flow with a hydrogel microvalve modified with an aptamer that is able to bind to analytes such as Hg2+ and Pb2+ at the inlet of the capillary channel. As a result, microvalve would open because of the hydrogel volume reduction that leads to enhancement of the capillary flow

8.5 Metabolites and Small Molecules

Figure 8.15 Schematic illustration of distance-based detection in a multilayer device. Buffers as well as colorimetric and masking reagents are inkjet printed in the related zones. On the back side of the device, a colorimetric indicator was printed as a passive timer. When a metal–ligand complex precipitated on the substrate, a color band would be created and the amount of present metal is proportional to this length and by this way the mass of the analyte can be quantified. Source: Reproduced with permission from Cate et al. [128]. Copyright 2015, The Royal Society of Chemistry.

Pretreatment zone Sample zone

Detection zone

Timer

Print reagents Frontside

Backside

Add analyte Flow

Ni Fe Cu

Measure distance

Ni Fe Cu 23 mm = 15 μg Cu

rate of the liquid sample that goes to the exothermic reservoir. Then, NaOH would be dissolved by the sample that results in heat production that can be detected by forehead thermometer and this value relates to the analyte concentration [129]. Another important application of POC devices is a quantification of drug abuse that is considered as a major public health problem worldwide. Most of the available platforms are inaccurate and a long time is needed for the assays that require external instruments for readouts and finally they provide low throughputs. Therefore, fabricating an inexpensive, noninvasive device for fast quantitative detection of drug abuse is quite important even for assessing the medications efficacies, which needs numerous explorations on patient samples. To address this need, Li et al. [130] fabricated an integrated competitive volumetric-bar chart chip (CV-chip) for multiple drug target analysis, which is capable of quantitative readout with a low assay time of about 10 minutes and

267

268

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

also with simplified ELISA procedure. This sensitive device is able to detect different complex samples such as serum, urine, and whole blood and the results were consistent with liquid chromatography–mass spectroscopy. Therefore, the mentioned portable platform that is fast and can provide accurate results using visible readout is an interesting development for detection of drugs abuse such as cocaine and amphetamine in serum, urine, and whole blood that can be used with least invasive manner with suitable sensitivity (0.94) and specificity (1.00) [130]. Recently, the coffee ring effect was utilized for transduction of color-based test result into length, and the device was designed for quantitative detection of glucose and lactate [131]. The coffee ring is considered as a pattern that is left after drying a drop of coffee on one substrate. Because of the fact that the liquid located near the droplet edge evaporates more quickly compared to the center part, a capillary flow would be created that leads to the transportation of the solute from the droplet center toward its edge. Subsequently, a ring-shaped stain would be generated on the substrate by liquid evaporation. In this work, the test reagents (e.g. GOx that is used for catalyzing the oxidation of glucose by oxygen and results in the formation of H2 O2 ) would be preloaded to the starch-iodide paper strip and the sample that is in liquid form would be introduced onto it. Then, the reagents and the sample would be reacted in a dry space for 10 minutes (the resulting H2 O2 oxidizes the iodide to iodine that was transformed to triiodide anion and leads to formation of a complex by starch with strong purple/blue color) and the resulting color can be transformed to the width of the color ring, which is easy to be detected by a ruler and correlates with the glucose concentration [131]. This approach is also a promising achievement for quantitative POCT under resource-limited situations. In recent years, microfluidic paper-based devices (μPAD) have extracted huge attention to detect small molecules in terms of its simple, fast, portable, disposable features for POC analytical applications [132]. However, effective POC analytical methods need stringent requirements and challenges for biosensor development [133]. Increasing groups try to functionalize μPAD [134] with a wide variety of biosensors such as colorimetric indicators, fluorescent/chemiluminescent immunoassays, microfluidic arrays, origami immunosensors, and electrochemical biosensors. For example, Pappa et al. fabricated a compact biosensing device that is made up of multiplexed organic electrochemical transistors (OECTs) for the simultaneous detection of biomarkers as shown in Figure 8.16. This device embeds organic electronic design with finger microfluidics, which immobilizes high selective and sensitive proteins, namely glucose oxidase (GOx), lactate oxidase (LOx), and cholesterol oxidase (ChOx) for the respective detection of glucose, lactate, and cholesterol from saliva. The sensing mechanism of this platform base on the enzyme/mediator complex functionalized gate electrode. Change of electric signal from the gating of the channel is proportional to the concentration of the analyte allowing its quantification. Figure 8.16c shows the salivary metabolite levels of five healthy volunteers as measured by this device. By collecting samples from two healthy volunteers before and after intense physical exercise, the relative variations of those metabolites are examined as shown in Figure 8.16d.

8.5 Metabolites and Small Molecules

“Finger-powered microfluidics” Multianalyte sensing platform GOx ChOx LOx 3 G S

BSA 2

1

D

Gluc

(a)

Concentration (mM)

Concentration (mM)

Ctrl

(b) 2.5 2.0 1.5 1.0 0.5 0.0

(c)

Chol Lact

Glucose

Lactate

Cholesterol

2.0

After exercise Glucose Lactate Cholesterol

After exercise

1.5 1.0 0.5 0.0

(d)

Person 1

Person 2

Figure 8.16 Selective multianalyte detection in complex media using the OECT array. (a) Schematic illustration of the biosensing multiplatform with the embedded “finger-powered” PDMS microfluidic showing the (1) activation “button,” (2) the liquid reservoir, and (3) the punched inlet. (b) Photograph of the actual device used for the measurements, showing a red-colored solution that was pressure-driven from the inlet through the sensing areas, as indicated by the arrow. (c) Salivary metabolite levels of five healthy volunteers as measured with our setup (the marked areas represent the physiological ranges of concentrations for each analyte), (d) relative salivary metabolite variations of two healthy volunteers before and after intense physical exercise. Source: Reproduced with permission from Pappa et al. [135]. Copyright 2016, John Wiley & Sons, Inc.

Nanoscience and nanotechnology have been one of the most rapidly fast research fields in recent years. Point of care with microfluidic paper-based device integrated with nanostructured materials with characteristics such as large surface area, tunable surface charge, and chemical stability [136] will assist in the sensitive detection of small molecules. The integration of nanomaterials has been of vital importance to improve the performance of sensing devices such as detection limits, sensibility, selectivity, reproducibility, and miniaturization [16]. Figure 8.17 shows an example designed by Choi et al. in order to detect thyroid-stimulating hormone (TSH) that is responsible for regulating the body metabolism. Based on surface-enhanced Raman scattering (SERS) immunoassay technique, they prepare a lateral flow immunoassay (LFIA) test strip with a sample pad, an absorbent pad, nitrocellulose (NC) membrane with a plastic backing card and a conjugate pad, in which AuNPs are used as SERS detection probes. The antigens interact with the detection antibody-conjugated AuNPs when they are passing through the conjugate pad after clinical fluids containing target antigens migrate toward the absorption pad by capillary force. Then, highly accurate quantitative measurement will be achieved by detecting the SERS signals that originate from the Raman reporter-labeled AuNPs that accumulated in the test line of the LFIA strip.

269

270

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

Conventional TSH LFIA

T

C

Hypothyroid Naked eye

Qualitative analysis T

(a)

C

Normal

T C

SERS-based TSH LFIA

T

C

Hyperthyroid Raman spectroscopy

Quantitative analysis T

Normal or hyperthyroid (only with SERS measurement)

T C

(b)

NC membrane

On

Test line Control line Absorbent pad

Off Flow

Flow T

C

T

C

TSH

AuNP

MGITC

C

Control Capture Detection antibody antibody antibody (anti-TSH IgG 1) (anti-TSH lgG 2) (chicken IgY)

PVP

BSA

(c)

Figure 8.17 Schematic illustration of (a) conventional naked eye LFIA platform and (b) SERS-based LFIA platform. Both hyperthyroidism and hypothyroidism can be detected by measuring the SERS intensity, even for concentrations of TSH lower than 1 μIU/ml when there is no color change at the T-line on the strip. (c) The operating principle of the SERS-based lateral flow immunoassay (LFIA) for TSH detection. Source: Reproduced with permission from Choi et al. [137]. Copyright 2017, Elsevier.

References

In a nutshell, small molecules play an important role in human health. To monitor and detect the level of small molecules, microfluidics applied to POC diagnostics is a rapidly emerging field. Detecting common metabolites from sweat is an interesting direction to measure physiological health status and report diagnosis results of the disease. Design of a quantitative readout is necessary before fabricating an effective device. New devices with sensitive, low-cost, portable, and time-saving performance are optimized to detect small molecules by integrating chemical/electrochemical methods, which are of vital importance in environmental monitoring and health care diagnostics. Furthermore, using nanoparticles as part of the detection chemistry show great potential in POC diagnosis systems.

8.6 Conclusion and Outlook In this chapter, we have briefly reviewed several representative microfluidic devices according to the type of the biomarker and their applications in POCT. Although microfluidics is a discipline with a history over 30 years, its applications in POCT have emerged only in the past decade. As a matter of that, the microfluidic-based POCT device is still not wildly used in disease diagnosis. With the continuous growth of demand for fast and accurate diagnosis equipment in contrast to the scarcity of the medical resource, one can foresee that more and more resources will be invested into quick diagnosis field. As a powerful technique to manipulate fluidic at the microscale, it comes naturally that microfluidic systems will continue to serve purposes related to biomedical science. It is definite that the microfluidic-based POCT device will prove their great value in disease diagnosis and demonstrate their clinical potential.

Acknowledgments We gratefully appreciate the financial support by the General Research Fund (GRF) from the Research Grant Council (RGC) of Hong Kong (Project Numbers GRF16325116, GRF16306115, and GRF16308818).

References 1 Whitesides, G.M. (2006). The origins and the future of microfluidics. Nature

442 (7101): 368–373. 2 Pinto, I.F., Santos, D.R., Soares, R.R.G. et al. (2018). A regenerable microflu-

idic device with integrated valves and thin-film photodiodes for rapid optimization of chromatography conditions. Sens. Actuators, B 255: 3636–3646. 3 Regnier, F.E., He, B., Lin, S., and Busse, J. (1999). Chromatography and electrophoresis on chips: critical elements of future integrated, microfluidic analytical systems for life science. Trends Biotechnol. 17 (3): 101–106.

271

272

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

4 Pandya, H.J., Dhingra, K., Prabhakar, D., and Chandrasekar, V. (2017). A

5 6

7

8 9

10

11

12

13 14

15

16

17

18 19

micro fluidic platform for drug screening in a 3D cancer microenvironment. Biosens. Bioelectron. 94: 632–642. Gubala, V., Harris, L.F., Ricco, A.J. et al. (2012). Point of care diagnostics: status and future. Anal. Chem. 84 (2): 487–515. Huang, L., Chen, Y., Chen, Y., and Wu, H. (2015). Centrifugation-assisted single-cell trapping in a truncated cone-shaped microwell array chip for the real-time observation of cellular apoptosis. Anal. Chem. 87 (24): 12169–12176. Kamalakshakurup, G. and Lee, A.P. (2017). High-efficiency single cell encapsulation and size selective capture of cells in picoliter droplets based on hydrodynamic micro-vortices. Lab Chip 17 (24): 4324–4333. Nie, M. and Takeuchi, S. (2018). Bottom-up biofabrication using microfluidic techniques. Biofabrication 10 (4): 044103. Chen, Y., Huang, L., Dai, X. et al. (2017). Facile formation of a microporous chitosan hydrogel based on self-crosslinking. J. Mater. Chem. B 5: 9291–9299. Chaipan, C., Pryszlak, A., Dean, H. et al. (2017). Single-virus droplet microfluidics for high-throughput screening of neutralizing epitopes on HIV particles. Cell Chem. Biol. 24 (6): 751–757.e3. Koo, K.M., Wee, E.J.H., and Trau, M. (2017). High-speed biosensing strategy for non-invasive profiling of multiple cancer fusion genes in urine. Biosens. Bioelectron. 89: 715–720. Majors, C.E., Smith, C.A., Natoli, M.E. et al. (2017). Point-of-care diagnostics to improve maternal and neonatal health in low-resource settings. Lab Chip 17 (20): 3351–3387. Zhang, L., Ding, B., Chen, Q. et al. (2017). Point-of-care-testing of nucleic acids by microfluidics. TrAC, Trends Anal. Chem. 94: 106–116. Hu, J., Cui, X., Gong, Y. et al. (2016). Portable microfluidic and smartphone-based devices for monitoring of cardiovascular diseases at the point of care. Biotechnol. Adv. 34 (3): 305–320. Jung, W., Han, J., Choi, J.W., and Ahn, C.H. (2014). Point-of-care testing (POCT) diagnostic systems using microfluidic lab-on-a-chip technologies. Microelectron. Eng. 132: 46–57. Hong, Y., Wu, M., Chan, G. et al. (2016). 3D printed microfluidic device with microporous Mn2 O3 -modified screen printed electrode for real-time determination of heavy metal ions. ACS Appl. Mater. Interfaces 8 (48): 32940–32947. Gao, W., Zhang, W., and Meldrum, D.R. (2011). RT-qPCR based quantitative analysis of gene expression in single bacterial cells. J. Microbiol. Methods 85 (3): 221–227. Jiang, X., Jing, W., Sun, X. et al. (2016). High-throughput microfluidic device for LAMP analysis of airborne bacteria. ACS Sensors 1 (7): 958–962. Kalsi, S., Sellars, S.L., Turner, C. et al. (2017). A programmable digital microfluidic assay for the simultaneous detection of multiple anti-microbial resistance genes. Micromachines 8 (4): 111.

References

20 Patel, P., Abd El Wahed, A., Faye, O. et al. (2016). A field-deployable reverse

21 22

23

24

25

26 27

28

29

30

31

32 33

34

35

transcription recombinase polymerase amplification assay for rapid detection of the chikungunya virus. PLoS Negl. Trop. Dis. 10 (9): 1–15. Heikali, D. and Di Carlo, D. (2010). A niche for microfluidics in portable hematology analyzers. JALA - J. Assoc. Lab. Autom. 15 (4): 319–328. Holmes, D. and Morgan, H. (2010). Single cell impedance cytometry for identification and counting of CD4 T-cells in human blood using impedance labels. Anal. Chem. 82 (4): 1455–1461. Holmes, D., Sun, T., Morgan, H. et al. (2007). Label-free differential leukocyte counts using a microfabricated, single-cell impedance spectrometer. Proceedings 2007 IEEE Sensors, 1452–1455. Holmes, D., Pettigrew, D., Reccius, C.H. et al. (2009). Leukocyte analysis and differentiation using high speed microfluidic single cell impedance cytometry. Lab Chip 9 (20): 2881–2889. Han, X., Van Berkel, C., Gwyer, J. et al. (2012). Microfluidic lysis of human blood for leukocyte analysis using single cell impedance cytometry. Anal. Chem. 84 (2): 1070–1075. Evander, M., Ricco, A.J., Morser, J. et al. (2013). Microfluidic impedance cytometer for platelet analysis. Lab Chip 13 (4): 722–729. Smith, S., Madzivhandila, P., Sewart, R. et al. (2017). Microfluidic cartridges for automated, point-of-care blood cell counting. SLAS Technol. 22 (2): 176–185. Roy, M., Jin, G., Seo, D. et al. (2014). A simple and low-cost device performing blood cell counting based on lens-free shadow imaging technique. Sens. Actuators, B 201: 321–328. Li, J., Liu, Q., Xiao, L. et al. (2013). Label-free method for cell counting in crude biological samples via paramagnetic bead aggregation. Anal. Chem. 85 (23): 11233–11239. Li, S.H., Jain, A., Tscharntke, T. et al. (2018). Hand-held photometer for instant on-spot quantification of nucleic acids, proteins, and cells. Anal. Chem. 90 (4): 2564–2569. Bhandari, V., Wang, C., Rinder, C., and Rinder, H. (2008). Hematologic profile of sepsis in neonates: neutrophil CD64 as a diagnostic marker. Pediatrics 121 (1): 129–134. Hoffmann, J.J.M.L. (2009). Neutrophil CD64: a diagnostic marker for infection and sepsis. Clin. Chem. Lab. Med. 47 (8): 903–916. Hassan, U., Ghonge, T., Reddy, B. et al. (2017). A point-of-care microfluidic biochip for quantification of CD64 expression from whole blood for sepsis stratification. Nat. Commun. 8 (15949): 1–12. Zhang, Y., Zhou, Y., Li, W. et al. (2018). Multiparameter affinity microchip for early sepsis diagnosis based on CD64 and CD69 expression and cell capture. Anal. Chem. 90 (12): 7204–7211. Qiu, X., Yang, S., Wu, D. et al. (2018). Rapid enumeration of CD4+ T lymphocytes using an integrated microfluidic system based on chemiluminescence image detection at point-of-care testing. Biomed. Microdevices 20 (15): 1–10.

273

274

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

36 Liu, Q., Chernish, A., Duvall, J.A. et al. (2016). The ARTμS: a novel

37

38

39 40

41

42

43

44

45

46

47 48

49 50

51

microfluidic CD4+ T-cell enumeration system for monitoring antiretroviral therapy in HIV patients. Lab Chip 16 (3): 506–514. Glynn, M.T., Kinahan, D.J., and Ducrée, J. (2014). Rapid, low-cost and instrument-free CD4+ cell counting for HIV diagnostics in resource-poor settings. Lab Chip 14 (15): 2844–2851. Wasserberg, D., Zhang, X., Breukers, C. et al. (2018). All-printed cell counting chambers with on-chip sample preparation for point-of-care CD4 counting. Biosens. Bioelectron. 117: 659–668. Li, P., Mao, Z., Peng, Z. et al. (2015). Acoustic separation of circulating tumor cells. Proc. Natl. Acad. Sci. U.S.A. 112 (16): 4970–4975. Kim, T.-H., Lim, M., Park, J. et al. (2017). FAST: size-selective, clog-free isolation of rare cancer cells from whole blood at a liquid–liquid interface. Anal. Chem. 89 (2): 1155–1162. Park, M.-H., Reátegui, E., Li, W. et al. (2017). Enhanced isolation and release of circulating tumor cells using nanoparticle binding and ligand exchange in a microfluidic chip. J. Am. Chem. Soc. 139 (7): 2741–2749. Safaei, T.S., Mohamadi, R.M., Sargent, E.H., and Kelley, S.O. (2015). In situ electrochemical ELISA for specific identification of captured cancer cells. ACS Appl. Mater. Interfaces 7 (26): 14165–14169. Strachan, B.C., Sloane, H.S., Houpt, E. et al. (2016). A simple integrated microfluidic device for the multiplexed fluorescence-free detection of Salmonella enterica. Analyst 141 (3): 947–955. Sposito, A., Hoang, V., and DeVoe, D.L. (2016). Rapid real-time PCR and high resolution melt analysis in a self-filling thermoplastic chip. Lab Chip 16 (18): 3524–3531. Houssin, T., Cramer, J., Grojsman, R. et al. (2016). Ultrafast, sensitive and large-volume on-chip real-time PCR for the molecular diagnosis of bacterial and viral infections. Lab Chip 16 (8): 1401–1411. Roy, E., Stewart, G., Mounier, M. et al. (2015). From cellular lysis to microarray detection, an integrated thermoplastic elastomer (TPE) point of care Lab on a Disc. Lab Chip 15 (2): 406–416. Kim, T.H., Sunkara, V., Park, J. et al. (2016). A lab-on-a-disc with reversible and thermally stable diaphragm valves. Lab Chip 16 (19): 3741–3749. Shin, D.J., Trick, A.Y., Hsieh, Y.H. et al. (2018). Sample-to-answer droplet magnetofluidic platform for point-of-care hepatitis C viral load quantitation. Sci. Rep. 8 (1): 1–12. Notomi, T., Okayama, H., Masubuchi, H. et al. (2000). Loop-mediated isothermal amplification of DNA. Nucleic Acids Res. 28 (12): E63. Giuffrida, M.C. and Spoto, G. (2017). Integration of isothermal amplification methods in microfluidic devices: recent advances. Biosens. Bioelectron. 90: 174–186. Dou, M., Dominguez, D.C., Li, X. et al. (2014). A versatile PDMS/paper hybrid microfluidic platform for sensitive infectious disease diagnosis. Anal. Chem. 86 (15): 7978–7986.

References

52 Chen, C., Liu, P., Zhao, X. et al. (2017). A self-contained microfluidic in-gel

53

54

55

56

57

58

59 60 61 62 63

64 65

66

67

loop-mediated isothermal amplification for multiplexed pathogen detection. Sens. Actuators, B 239: 1–8. Kadimisetty, K., Song, J., Doto, A.M. et al. (2018). Fully 3D printed integrated reactor array for point-of-care molecular diagnostics. Biosens. Bioelectron. 109: 156–163. Dou, M., Sanjay, S.T., Dominguez, D.C. et al. (2017). Multiplexed instrument-free meningitis diagnosis on a polymer/paper hybrid microfluidic biochip. Biosens. Bioelectron. 87: 865–873. Dou, M., Sanjay, S.T., Dominguez, D.C. et al. (2017). A paper/polymer hybrid CD-like microfluidic SpinChip integrated with DNA-functionalized graphene oxide nanosensors for multiplex qLAMP detection. Chem. Commun. 53 (79): 10886–10889. Shin, Y., Perera, A.P., Tang, W.Y. et al. (2015). A rapid amplification/detection assay for analysis of Mycobacterium tuberculosis using an isothermal and silicon bio-photonic sensor complex. Biosens. Bioelectron. 68: 390–396. Heo, H.Y., Chung, S., Kim, Y.T. et al. (2016). A valveless rotary microfluidic device for multiplex point mutation identification based on ligation-rolling circle amplification. Biosens. Bioelectron. 78: 140–146. Wei, X., Zhou, W., Sanjay, S.T. et al. (2018). Multiplexed instrument-free bar-chart SpinChip integrated with nanoparticle-mediated magnetic aptasensors for visual quantitative detection of multiple pathogens. Anal. Chem. 90 (16): 9888–9896. Alp, A. (2018). Advancement in POCT molecular testing: the multiplex PCR POCT devices for infectious diseases. EJIFCC 29: 205–209. Lindenmann, J. (1984). Origin of the terms “antibody” and “antigen”. Scand. J. Immunol. 19 (4): 281–285. Wild, D. (2013). The Immunoassay Handbook: Theory and Applications of Ligand Binding, ELISA and Related Techniques. Newnes. Chin, C.D., Linder, V., and Sia, S.K. (2012). Commercialization of microfluidic point-of-care diagnostic devices. Lab Chip 12 (12): 2118. Gervais, L. and Delamarche, E. (2009). Toward one-step point-of-care immunodiagnostics using capillary-driven microfluidics and PDMS substrates. Lab Chip 9 (23): 3330–3337. Sista, R., Hua, Z., Thwar, P. et al. (2008). Development of a digital microfluidic platform for point of care testing. Lab Chip 8 (12): 2091–2104. Chin, C.D., Laksanasopin, T., Cheung, Y.K. et al. (2011). Microfluidics-based diagnostics of infectious diseases in the developing world. Nat. Med. 17 (8): 1015–1019. Shen, J., Zhou, Y., Fu, F. et al. (2015). Immunochromatographic assay for quantitative and sensitive detection of hepatitis B virus surface antigen using highly luminescent quantum dot-beads. Talanta 142: 145–149. Ran, B., Xianyu, Y., Dong, M. et al. (2017). Bioorthogonal reaction-mediated ELISA using peroxide test strip as signal readout for point-of-care testing. Anal. Chem. 89 (11): 6113–6119.

275

276

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

68 Hu, M., Yan, J., He, Y. et al. (2010). Ultrasensitive, multiplexed detection of

69

70

71

72

73

74

75

76

77

78

79

80 81 82 83

cancer biomarkers directly in serum by using a quantum dot-based microfluidic protein chip. ACS Nano 4 (1): 488–494. Li, J., Baird, M.A., Davis, M.A. et al. (2017). Dramatic enhancement of the detection limits of bioassays via ultrafast deposition of polydopamine. Nat. Biomed. Eng. 4 (0082): 1–28. Wu, J., Chen, Y., Yang, M. et al. (2017). Streptavidin-biotin-peroxidase nanocomplex-amplified microfluidics immunoassays for simultaneous detection of inflammatory biomarkers. Anal. Chim. Acta 982: 138–147. Shohatee, D., Keifer, J., Schimmel, N. et al. (2018). Hydrogel-based suspension array for biomarker detection using horseradish peroxidase-mediated silver precipitation. Anal. Chim. Acta 999: 132–138. Guo, T., Patnaik, R., Kuhlmann, K. et al. (2015). Smartphone dongle for simultaneous measurement of hemoglobin concentration and detection of HIV antibodies. Lab Chip 15 (17): 3514–3520. Li, X. and Liu, X. (2016). A microfluidic paper-based origami nanobiosensor for label-free, ultrasensitive immunoassays. Adv. Healthc. Mater. 5 (11): 1326–1335. Yu, X., Xia, Y., Tang, Y. et al. (2017). A nanostructured microfluidic immunoassay platform for highly sensitive infectious pathogen detection. Small 13 (1700425): 1–12. Xia, Y., Tang, Y., Yu, X. et al. (2017). Label-free virus capture and release by a microfluidic device integrated with porous silicon nanowire forest. Small 13 (1603135): 1–11. Yeh, Y.T., Tang, Y., Sebastian, A. et al. (2016). Tunable and label-free virus enrichment for ultrasensitive virus detection using carbon nanotube arrays. Sci. Adv. 2 (10): e1601026. Shafiee, H., Jahangir, M., Inci, F. et al. (2013). Acute on-chip HIV detection through label-free electrical sensing of viral nano-lysate. Small 9 (15): 2553–2563. Chen, Y., Xianyu, Y., Wu, J. et al. (2017). Double-enzymes-mediated bioluminescent sensor for quantitative and ultrasensitive point-of-care testing. Anal. Chem. 89 (10): 5422–5427. Kim, S.J., Song, Y.A., and Han, J. (2010). Nanofluidic concentration devices for biomolecules utilizing ion concentration polarization: theory, fabrication, and applications. Chem. Soc. Rev. 39 (3): 912–922. Au, A.K., Huynh, W., Horowitz, L.F., and Folch, A. (2016). 3D-printed microfluidics. Angew. Chem. Int. Ed. 55 (12): 3862–3881. Sochol, R.D., Sweet, E., Glick, C.C. et al. (2018). 3D printed microfluidics and microelectronics. Microelectron. Eng. 189: 52–68. Chan, H.N., Tan, M.J.A., and Wu, H. (2017). Point-of-care testing: applications of 3D printing. Lab Chip 17 (16): 2713–2739. Begolo, S., Zhukov, D.V., Selck, D.A. et al. (2014). The pumping lid: Investigating multi-material 3D printing for equipment-free, programmable generation of positive and negative pressures for microfluidic applications. Lab Chip 14 (24): 4616–4628.

References

84 Sweet, E.C., Mehta, R.R., Lin, R., and Lin, L. (2017). Finger-powered, 3D

85

86

87

88

89 90

91

92

93

94

95

96

97

98

printed microfluidic pumps. TRANSDUCERS 2017 - 19th Internaitonal Conference Solid-State Sensors, Actuators Microsystems, pp. 1766–1769. Anderson, K.B., Lockwood, S.Y., Martin, R.S., and Spence, D.M. (2013). A 3D printed fluidic device that enables integrated features. Anal. Chem. 85 (12): 5622–5626. Jue, E., Schoepp, N.G., Witters, D., and Ismagilov, R.F. (2016). Evaluating 3D printing to solve the sample-to-device interface for LRS and POC diagnostics: example of an interlock meter-mix device for metering and lysing clinical urine samples. Lab Chip 16 (10): 1852–1860. Plevniak, K., Campbell, M., Myers, T. et al. (2016). 3D printed auto-mixing chip enables rapid smartphone diagnosis of anemia. Biomicrofluidics 10 (5): 054113. Li, F., Smejkal, P., Macdonald, N.P. et al. (2017). One-step fabrication of a microfluidic device with an integrated membrane and embedded reagents by multimaterial 3D printing. Anal. Chem. 89 (8): 4701–4707. Yuen, P.K. (2016). Embedding objects during 3D printing to add new functionalities. Biomicrofluidics 10 (4): 9–12. Erkal, J.L., Selimovic, A., Gross, B.C. et al. (2014). 3D printed microfluidic devices with integrated versatile and reusable electrodes. Lab Chip 14 (12): 2023–2032. Comina, G., Suska, A., and Filippini, D. (2015). Autonomous chemical sensing interface for universal cell phone readout. Angew. Chem. Int. Ed. 54 (30): 8708–8712. Comina, G., Suska, A., and Filippini, D. (2016). A 3D printed device for quantitative enzymatic detection using cell phones. Anal. Methods 8 (32): 6135–6142. Chan, H.N., Shu, Y., Xiong, B. et al. (2016). Simple, cost-effective 3D printed microfluidic components for disposable, point-of-care colorimetric analysis. ACS Sensors 1 (3): 227–234. Tang, C.K., Vaze, A., and Rusling, J.F. (2017). Automated 3D-printed unibody immunoarray for chemiluminescence detection of cancer biomarker proteins. Lab Chip 17 (3): 484–489. Kadimisetty, K., Malla, S., Bhalerao, K. et al. (2018). Automated 3D-printed microfluidic array for rapid nanomaterial-enhanced detection of multiple proteins. Anal. Chem. 90 (12): 7569–7577. Kadimisetty, K., Spak, A.P., Bhalerao, K.S. et al. (2018). Automated 4-sample protein immunoassays using 3D-printed microfluidics. Anal. Methods 10 (32): 4000–4006. Dirkzwager, R.M., Liang, S., and Tanner, J.A. (2016). Development of aptamer-based point-of-care diagnostic devices for malaria using three-dimensional printing rapid prototyping. ACS Sensors 1 (4): 420–426. Fraser, L.A., Kinghorn, A.B., Dirkzwager, R.M. et al. (2018). A portable microfluidic Aptamer-Tethered Enzyme Capture (APTEC) biosensor for malaria diagnosis. Biosens. Bioelectron. 100: 591–596.

277

278

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

99 Gong, H., Bickham, B.P., Woolley, A.T., and Nordin, G.P. (2017). Custom 3D

100

101 102

103

104

105

106

107

108

109

110

111 112

113

114

printer and resin for 18 μm × 20 μm microfluidic flow channels. Lab Chip 17 (17): 2899–2909. Bauer, M. and Kulinsky, L. (2018). Fabrication of a lab-on-chip device using material extrusion (3D printing) and demonstration via malaria-Ab ELISA. Micromachines 9 (1): 27. Choi, K., Ng, A.H.C., Fobel, R., and Wheeler, A.R. (2012). Digital microfluidics. Annu. Rev. Anal. Chem. 5 (1): 413–440. Xu, X., Sun, L., Chen, L. et al. (2014). Electrowetting on dielectric device with crescent electrodes for reliable and low-voltage droplet manipulation. Biomicrofluidics 8 (6): 064107. Hase, M., Watanabe, S.N., and Yoshikawa, K. (2006). Rhythmic motion of a droplet under a dc electric field. Phys. Rev. E: Stat. Nonlinear Soft Matter Phys. 74 (4): 3–6. Miralles, V., Huerre, A., Williams, H. et al. (2015). A versatile technology for droplet-based microfluidics: thermomechanical actuation. Lab Chip 15 (9): 2133–2139. Zhang, Y. and Wang, T.H. (2013). Full-range magnetic manipulation of droplets via surface energy traps enables complex bioassays. Adv. Mater. 25 (21): 2903–2908. Kong, T., Brien, R., Njus, Z. et al. (2016). Motorized actuation system to perform droplet operations on printed plastic sheets. Lab Chip 16 (10): 1861–1872. Hernandez-Perez, R., Fan, Z.H., and Garcia-Cordero, J.L. (2016). Evaporation-driven bioassays in suspended droplets. Anal. Chem. 88 (14): 7312–7317. Zhang, D., Gao, B., Chen, Y., and Liu, H. (2018). Converting colour to length based on the coffee-ring effect for quantitative immunoassays using a ruler as readout. Lab Chip 18 (2): 271–275. Mok, J., Mindrinos, M.N., Davis, R.W., and Javanmard, M. (2014). Digital microfluidic assay for protein detection. Proc. Natl. Acad. Sci. U.S.A. 111 (6): 2110–2115. Fobel, R., Fobel, C., and Wheeler, A.R. (2013). DropBot: an open-source digital microfluidic control system with precise control of electrostatic driving force and instantaneous drop velocity measurement. Appl. Phys. Lett. 102 (19): 193513. Alistar, M. and Gaudenz, U. (2017). OpenDrop: an integrated do-it-yourself platform for personal use of biochips. Bioengineering 4 (2): 45. Ng, A.H.C., Fobel, R., Fobel, C. et al. (2018). A digital microfluidic system for serological immunoassays in remote settings. Sci. Transl. Med. 10 (eaar6076): 1–13. Shu, B., Li, Z., Yang, X. et al. (2018). Active droplet-array (ADA) microfluidics enables multiplexed complex bioassays for point of care testing. Chem. Commun. 54 (18): 2232–2235. Koh, A., Kang, D., Xue, Y. et al. (2017). A soft, wearable microfluidic device for the capture, storage, and colorimetric sensing of sweat. Sci. Transl. Med. 8: 366ra165.

References

115 Salvo, P., Di Francesco, F., Costanzo, D. et al. (2010). A wearable sensor for

measuring sweat rate. IEEE Sens. J. 10 (10): 1557–1558. 116 Jadoon, S., Karim, S., Akram, M.R. et al. (2015). Recent developments in

sweat analysis and its applications. Int. J. Anal. Chem. 2015: 164974. 117 LeGrys, V.A., Applequist, R., Briscoe, D.R. et al. (2009). Sweat testing: sam-

118

119

120

121

122 123

124

125

126

127

128

129

130

ple collection and quantitative chloride analysis; approved guideline—third edition. Clin. Lab. Stand. Inst. 29: C34–A2. Jia, W., Bandodkar, A.J., Valde, G. et al. (2013). Electrochemical tattoo biosensors for real-time noninvasive lactate monitoring in human perspiration. Anal. Chem. 85: 6553–6560. Bandodkar, A.J., Molinnus, D., Mirza, O. et al. (2014). Epidermal tattoo potentiometric sodium sensors with wireless signal transduction for continuous non-invasive sweat monitoring. Biosens. Bioelectron. 54: 603–609. Curto, V.F., Fay, C., Coyle, S. et al. (2012). Real-time sweat pH monitoring based on a wearable chemical barcode micro-fluidic platform incorporating ionic liquids. Sens. Actuators, B 171–172: 1327–1334. Oncescu, V., O’Dell, D., and Erickson, D. (2013). Smartphone based health accessory for colorimetric detection of biomarkers in sweat and saliva. Lab Chip 13: 3232–3238. Shen, L., Hagen, A., and Papautsky, I. (2012). Point-of-care colorimetric detection with a smartphone. Lab Chip 12: 4240–4243. Rose, D.P., Member, S., Ratterman, M.E. et al. (2015). Adhesive RFID sensor patch for monitoring of sweat electrolytes. IEEE Trans. Biomed. Eng. 62 (6): 1457–1465. Choi, J., Kang, D., Han, S. et al. (2017). Thin, soft, skin-mounted microfluidic networks with capillary bursting valves for chrono-sampling of sweat. Adv. Healthc. Mater. 6: 1601355. Kim, S.B., Zhang, Y., Won, S.M. et al. (2018). Super-absorbent polymer valves and colorimetric chemistries for time-sequenced discrete sampling and chloride analysis of sweat via skin-mounted soft microfluidics. Small 14 (12): 1703334. Lewis, G.G., Robbins, J.S., and Phillips, S.T. (2013). Point-of-care assay platform for quantifying active enzymes to femtomolar levels using measurements of time as the readout. Anal. Chem. 85: 10432–10439. Lewis, G.G., Robbins, J.S., and Phillips, S.T. (2014). A prototype point-of-use assay for measuring as a quantitative readout †. Chem. Commun. 50: 5352–5354. Cate, D.M., Noblitt, S.D., and Henry, C.S. (2015). Multiplexed paper analytical device for quantification of metals using distance-based detection. Lab Chip 15: 2808–2818. Gao, B., Liu, H., and Gu, Z. (2016). An exothermic chip for point-of-care testing using a forehead thermometer as a readout. Lab Chip 16 (3): 525–531. Li, Y., Uddayasankar, U., He, B. et al. (2017). Fast, sensitive, and quantitative point-of-care platform for the assessment of drugs of abuse in urine, serum, and whole blood. Anal. Chem. 89: 8273–8281.

279

280

8 Recent Developments in Microfluidic-Based Point-of-care Testing (POCT) Diagnoses

131 Zhang, D., Ma, B., Tang, L., and Liu, H. (2018). Toward quantitative chemi-

132

133

134

135

136

137

cal analysis using a ruler on paper: an approach to transduce color to length based on coffee-ring effect. Anal. Chem. 90 (3): 1482–1486. Ariza-Avidad, M., Salinas-Castillo, A., and Capitán-Vallvey, L.F. (2016). A 3D μPAD based on a multi-enzyme organic-inorganic hybrid nanoflower reactor. Biosens. Bioelectron. 77: 51–55. Liu, Y., Matharu, Z., Howland, M.C. et al. (2012). Affinity and enzyme-based biosensors: recent advances and emerging applications in cell analysis and point-of-care testing. Anal. Bioanal. Chem. 404 (4): 1181–1196. Burrs, S.L., Bhargava, M., Sidhu, R. et al. (2016). A paper based graphene-nanocauliflower hybrid composite for point of care biosensing. Biosens. Bioelectron. 85: 479–487. Pappa, A.M., Curto, V.F., Braendlein, M. et al. (2016). Organic transistor arrays integrated with finger-powered microfluidics for multianalyte saliva testing. Adv. Healthc. Mater. 5 (17): 2295–2302. Agnes, R.S., Jernigan, F., Shell, J.R. et al. (2010). Suborganelle sensing of mitochondrial cAMP-dependent protein kinase activity. J. Am. Chem. Soc. 132 (17): 6075–6080. Choi, S., Hwang, J., Lee, S. et al. (2017). Quantitative analysis of thyroid-stimulating hormone (TSH) using SERS-based lateral flow immunoassay. Sens. Actuators, B 240: 358–364.

281

9 Microfluidics in Microbiome and Cancer Research Barath Udayasuryan 1 , Daniel J. Slade 2 , and Scott S. Verbridge 1 1 Virginia Polytechnic Institute and State University – Wake Forest University, School of Biomedical Engineering and Sciences, 325 Stanger Street, Blacksburg, VA 24061, USA 2 Virginia Polytechnic Institute and State University, Department of Biochemistry, 340 West Campus Drive, Blacksburg, VA 24061, USA

9.1 Introduction Research into the role of the microbiome in human health is a burgeoning field of study that is poised to create a major paradigm shift in the way we approach the treatment and prevention of a vast number of diseases. With the discovery that the microbiome can profoundly influence digestion, metabolite production, drug absorption and metabolism, pharmacokinetics, and immunomodulation, efforts have been directed toward understanding the molecular mechanisms microbes use to regulate these processes. A stable host-microbiome interaction is requisite for normal gastrointestinal physiology and its alteration leads to a plethora of undesirable consequences including opportunistic infections, inflammatory diseases, cardiovascular disease, and even disruptive effects on the brain and endocrine systems. However, a significant bottleneck in this field is a lack of mechanistic understanding of individual microbiota and how they function holistically in a complex and diverse community, as seen within the gut epithelium. Additional evidence implicates the role of specific bacteria in cancer initiation and progression through the effects of immunomodulation and chronic inflammation. Studies investigating these key links between the microbiome and cancer are being brought together through interdisciplinary research bridging advances in biomaterials, tissue engineering, genetic engineering, systems biology, microfluidics, and microbiology. Microfluidic technologies combining these disciplines hold the promise of delineating host-microbe crosstalk and community dynamics, as well as unraveling the cellular and molecular mechanisms and interactions initiated by these bacteria. Coupled with advances in tissue engineering, organoid, and organ-on-chip technologies, there is much promise to drive this field forward. The microbiome can be leveraged to develop novel diagnostic and prognostic tools to help prevent and manage disease. Probiotics are already available as dietary elements or prescriptions and fecal microbial transplantations (bacteriotherapy) appear to be promising therapies for certain Nanotechnology and Microfluidics, First Edition. Edited by Xingyu Jiang. © 2020 Wiley-VCH Verlag GmbH & Co. KGaA. Published 2020 by Wiley-VCH Verlag GmbH & Co. KGaA.

282

9 Microfluidics in Microbiome and Cancer Research

dysbiosis-associated diseases (e.g. Clostridium difficile infections) [1] and recently have also been shown to help patients on cancer immunotherapy drugs (PD-1 blockers) [2]. Alterations in the gut microbiome have also been speculated to play a pathophysiological role in brain diseases and psychological disorders [3, 4]. Furthermore, influences of the mycobiome and virome add further complexity to this intricate ecosystem that exists within our body [5, 6]. Thus, there is tremendous opportunity to exploit microbial functionalities to benefit human health, and the first step toward that is to identify and understand their basic cellular and molecular interactions amongst the microbes and with the host. This chapter will provide an introduction to the role of microbiome in human health and disease including a brief review of the current techniques used to study the complex interactions between the host and microbes. This will be followed by a discussion of the principal features of emerging technologies in microfluidic device development and 3D tissue microenvironment engineering that enable the dissection of key microscale host-microbiome interactions. We will later examine select organ-on-chip devices that recapitulate salient features of the gut epithelium and how such technologies can be leveraged to study the role of bacteria in cancer initiation and progression.

9.2 What is the Microbiome? 9.2.1

Composition and Biogeography

The microbiota consists of the entire collection of microbes present in the myriad tissues of the body including the skin, mammary glands, placenta, seminal fluid, uterus, ovarian follicles, lung, saliva, oral mucosa, conjunctiva, biliary, and gastrointestinal tracts. The microbiome refers to the complete genomic constitution of these microbiota, although these terms are sometimes used interchangeably. Current estimates show that at least 3.8 × 1013 microorganisms inhabit the human intestines and form their own micro-ecosystem [7]. Incredibly, the number of human cells is estimated to be around 3.0 × 1013 cells which means that the microbes outnumber the human cells 1.3 : 1. These microbes have a strong influence on human health and disease [8]. Most of these organisms are commensal and maintain a symbiotic relationship with the host, however opportunistic pathogens can also be present under certain microenvironmental conditions [9]. Additionally, this complex ecosystem includes species that are not present anywhere else in nature possibly indicating co-evolution with the host over several millennia [10]. Normal gut homeostasis is maintained by the trillions of microorganisms present in the gut. This community is involved in aiding digestion, metabolizing polysaccharides, secreting metabolites, regulating drug metabolism, maintaining epithelial barrier function, and affecting immune response [11]. The dominant phyla that are present are the Firmicutes, Bacteroidetes, Actinobacteria, Verrucomicrobia, and Proteobacteria [10]. As shown in Figure 9.1, the composition and diversity vary within different parts of the digestive tract. The stomach and the small intestine have fewer species than the colon, and diversity

9.2 What is the Microbiome?

Dominant gut phyla: Bacteroidetes, Firmicutes, Actinobacteria, Proteobacteria, Verrucomicrobia

Predominant families in the: Small intestine

Colon

Lactobacillaceae, Enterobacteriaceae

Bacteroidaceae, Prevotellaceae, Rikenellaceae, Lachnospiraceae, Ruminococcaceae

Bile duct

Descending

Ascending

Transverse

Appendix Caecum Rectum 102 cfu/g

Inter-fold regions Lachnospiraceae, Ruminococcaceae

Digesta Bacteroidaceae, Prevotellaceae, Rikenellaceae

1011 cfu/g

Bacterial load pH Antimicrobials Oxygen Proximal

Distal

Figure 9.1 The dominant phyla and families present in the gut which reflect the physiological differences along the length of the gut. There are varied gradients of pH, antimicrobials, and oxygen which in turn affects the bacterial load along the gut. Source: Donaldson et al. 2016 [10]. Reproduced with permission of Springer Nature.

increases from proximal to distal within the intestines [12]. There exists various chemical and nutrient gradients, as well as compartmentalized host-immune activity along the lengths of the small intestine and colon that greatly influence microbial diversity [10]. The small intestine is acidic due to the presence of bile acids and has higher levels of oxygen than within the colon. There is a shorter transit time through the small intestine compared to the colon and hence bacterial adherence to the tissue or mucus also plays a role in persistent colonization [10, 12]. The colon and the caecum have the most dense and diverse microbial communities where their biomass exceeds 1011 cells per gram content [13]. The slower transit time and lack of simple carbon sources allow fermentative polysaccharide-degrading anaerobes to flourish. Additionally, the inter-fold regions of the lumen accumulate mucus which can also act as a source of nutrients for specific bacteria [10]. For microorganisms to occupy the crypts

283

284

9 Microfluidics in Microbiome and Cancer Research

of the colon, they need to be highly specialized to evade immune response and utilize specific host-derived nutrients to facilitate their close association with the host [14]. Furthermore, it has been found that the mucosal microbial populations occur in patches indicating spatial aggregation of interacting microorganisms and the presence of heterogenous biofilm microenvironments [15, 16]. The cellular architecture of the intestine is quite complex and consists of multiple cell types. These include absorptive enterocytes, hormone-secreting enteroendocrine cells, Paneth cells that participate in antimicrobial production, goblet cells responsible for mucin production, and M cells and Tuft cells which participate in immunity [17, 18]. In addition, macrophages, dendritic cells, T cells, and B cells are present and organized as lymphoid structures termed Peyer’s patches which are responsible for mucosal immunity. One of the most important functions of the intestinal epithelial cells is the establishment and maintenance of the intestinal barrier [19]. Any compromise to this barrier can lead to inflammation or unrestrained growth of microbiota leading to disease including inflammatory bowel disorders, autoimmune disease, and other metabolic disorders. The barrier is a one-cell thick lining of the gut and its function is established by epithelial cell polarity, where the cell’s apical surface faces the lumen of the villi and its basal surface faces the basement membrane and lamina propria [19]. These surfaces have distinct biochemical compositions that facilitate the permeability of essential ions, nutrients, and water while at the same time restricting the entry of bacterial pathogens and toxins. This layer is dynamic and has a high turnover rate (the intestinal epithelial cells are replaced every couple of days). Microbes have evolved to recognize these surface-specific molecules and, upon binding, can compromise the intestinal barrier, which would enable the uptake and dissemination of pathogens. Maintenance of barrier integrity requires the proper expression and localization of tight and adherens junctions composed of transmembrane and cytosolic proteins such as claudins, Zona occludin-1, occludin, tricellulin, cingulin, and junctional adhesion molecules (JAMs) [19]. It has been found that microbial metabolites also may function to maintain the gut barrier integrity [20]. The epithelial lining is further covered by a thin layer of mucus secreted by goblet cells that separates the commensal bacteria from directly interacting with the host epithelium and immune system [21]. During the first three years of life, the composition of the human gut microbiota is highly variable and unstable. Initially, the sterile intestinal tract is colonized by facultative anaerobes such as Lactobacilli, Enterobacteria, and Enterococci, species which are present in the environment and the mother [22, 23]. These early colonizers induce changes in the gut that transform the environment and make it conducive to strict anaerobes such as Bacteroides, Bifidobacterium, and Clostridium, which are more prevalent in adults. The mode of child birth (either vaginal or Cesarean), breast feeding, diet, and sanitation greatly shape the composition of the microbial community in the gut [22]. In fact, breast milk has been found to contain up to 107 bacterial cells/800 ml which act as an inoculum for newborns [23]. Usually by the age of three, the composition and diversity of the microbiota resembles that of adults. This bacterial composition is found to be a stable community with approximately 40 species accounting

9.2 What is the Microbiome?

for 75% of the overall microbial mass [10]. However, there may still be some variability based on diet, lifestyle, host genetics, as well as other life events including developmental stage (e.g. adolescence), ovarian cycles, pregnancy, and menopause [24]. The consumption of specific protein, fat, digestible and non-digestible carbohydrates, probiotics, and polyphenols directly influence the relative abundance of certain microbial genera and have secondary effects on host immunological and metabolic markers [13, 25–27]. Elderly people show a loss of microbial diversity and decreased temporal compositional stability [28]. The microbiome is known to directly affect the development of the host-immune system. For example, Bifidobacterium longum strongly stimulates production of interleukin 10 and proinflammatory cytokines such as tumor necrosis factor (TNFα) that protects against tumor proliferation [29]. Bacteroides fragilis affects mucosal T cell homeostasis and regulatory T cell function [30]. Clostridia of the phylum Firmicutes are highly heterogenous and their presence has been shown to promote accumulation and differentiation of CD4+ T cells [31]. Thus, it is not surprising that correlations have been reported between gut bacterial composition and immunogenicity of orally administered vaccines [22]. In the case of chronic disorders, such as Crohn’s disease, a significant decrease was observed in the population of beneficial Bifidobacteriaceae and an increase in phyla including Enterobacteriaceae, Pasteurellaceae, Fusobacteriaceae, and Neisseriaceae [29]. The resident microbiota also serves a protective role. Commensal bacteria, which are well adapted to the gut environment, can protect the host from opportunistic pathogens by producing inhibitory molecules which have bacteriostatic or bactericidal activity or they can even out-compete the utilization of nutrients required by pathogens [14]. 9.2.2

The Microbiome and Cancer

The gut microbiota are often collectively thought of as the “forgotten organ” and live in mutually advantageous harmony with their human host, a state termed as eubiosis. However, a number of factors can lead to dysbiosis, perturbing the equilibrium, which in turn has been found to potentiate the development and progression of neoplasms by a number of mechanisms including damaging DNA, activating oncogenic signaling pathways, producing tumor-promoting metabolites, suppressing anti-tumor immunity, and modulating responses to cancer therapy [32–37]. The compositional perturbation due to dysbiosis has been associated with disease states such as obesity, diabetes, colorectal cancer (CRC) and allergies [29, 36]. Realization of the association between bacteria and tumors is not new and has been reported in patients for decades [38]. Several species of bacteria have been implicated in tumorigenesis including Clostridium sp., Salmonella sp., Bifidobacterium sp., Campylobacter jejuni, Citrobacter rodentium, Escherichia coli, Vibrio cholerae, and Listeria monocytogenes [38, 39]. However, there is still a lack of substantial direct evidence that microbiota modulate carcinogenesis in humans [40] and their role is quite complex as bacteria have been shown to exhibit both pro- and anti-carcinogenic effects [41]. The bacterium Helicobacter pylori plays

285

286

9 Microfluidics in Microbiome and Cancer Research

a role in cancers such as gastric cancer and mucosa-associated lymphoid tissue (MALT) lymphoma through stimulating chronic inflammatory response [42]. Another intriguing example is that of the bacterium Fusobacterium nucleatum, whose invasion of the host tumor tissue in CRC accelerates, by largely unknown mechanisms, cancer progression and reduces patient survival [43]. Current knowledge on these two bacteria and their association with tumorigenesis are summarized below.

9.2.3

Helicobacter pylori and Gastric Cancer

Helicobacter pylori was first discovered in 1982 by Robin Warren and Barry Marshall. They were, in fact, awarded the Nobel Prize in Medicine in 2005 for the seminal discovery of this bacterium and its role in peptic ulcer disease [42]. The presence of the bacterium H. pylori represents the strongest known risk factor for gastric cancer, so much so that it was the first bacterium to be classified as a Type I carcinogen [41]. Strain biodiversity, geographical distribution, and environmental factors all play a role in determining which demographic is most susceptible to infection. This bacterium has also been associated with an increased risk for pancreatic cancer and CRC [42]. Helicobacter pylori is a Gram-negative bacterial pathogen that colonizes the gastric epithelium. It is spiral shaped and produces 3–5 polar flagella for motility. One of its most remarkable features is that their colonies can persist for decades in the harsh gastric environment (pH < 2). It is able to accomplish this feat by metabolizing urea to ammonia via urease forming a neutral environment that envelopes the bacterium [44]. There are ∼1 million cases of gastric cancer diagnosed each year making it the fourth most common cancer worldwide. Among the bacterium’s virulence factors is the protein CagA (cytotoxin associated gene A), which is a 120–140 kDa protein that is translocated into host gastric epithelial cells via the bacterial type IV secretion system. Here CagA undergoes tyrosine phosphorylation by Src family of kinases which leads to the activation of extracellular regulated kinase and mitogen activated protein kinase cascades such as the PI3K-AKT and WNT pathways that promote pro-oncogenic cell proliferation [44]. In addition, H. pylori enhances DNA damage through host mediated over-production of reactive oxygen species (ROS). It can also activate stemness properties and induces epithelial-mesenchymal transition in gastric epithelial cells [44]. Another virulence factor, VacA (vacuolating cytotoxin A), alters host cell membrane permeability and has been shown to suppress host immunity by inhibiting T cell activation and inducing the production of regulatory T cells. It has been shown that H. pylori directly induces gene mutations in epithelial cells by enhancing AID expression through NF-𝜅B activation. This acts as an editor of DNA and RNA and is essential for somatic hypermutation and class switch recombination of immunoglobulin genes in B lymphocytes. Recent studies have also suggested the role of H. pylori in DNA methylation that act to silence tumor suppressor genes and reduce epithelial cell apoptosis by the inhibition of TP53 [41].

9.2 What is the Microbiome?

9.2.4

Fusobacterium nucleatum and CRC

CRC is the third leading cause of cancer-related deaths in the US. The bacterium F. nucleatum plays a significant role in the progression of this cancer. F. nucleatum is an adhesive bacterium and normally co-aggregates with various microbial species in the oral cavity and is usually associated with gingivitis and periodontitis [45]. Its initial link to cancer was discovered when highly abundant F. nucleatum gene signatures were found in biobanked colorectal carcinoma using metagenomics methods [35, 45]. They were highly enriched in human CRCs compared to adjacent normal tissue. Further research clearly showed that F. nucleatum accelerates CRC in pre-clinical models using both in vitro and in vivo systems [45]. It has been shown that in ApcMin/+ mice (adenomatous polyposis coli [APC] multiple intestinal neoplasia [Min] murine model) there was increased carcinogenesis with the introduction of F. nucleatum or B. fragilis [46, 47]. These studies, however, have not been confirmed by observational studies in humans. The mechanism of enrichment of F. nucleatum is not clear. However, a few proteins have recently been discovered that appear to facilitate localization and internalization of this bacterium into the host. This bacterium encodes several adhesins for interspecies interactions, including Fap2, RadD, FadA, and Aid1. It binds to and invades a variety of mammalian cells including epithelial and endothelial cells, monocytes, erythrocytes, keratinocytes, fibroblasts, HeLa cells, and natural killer (NK) cells, and interacts with host molecules such as salivary macromolecules, extracellular matrix (ECM) proteins, human IgG, and cadherins [45]. Most recently, it was shown that F. nucleatum is present in distant hepatic and lymph node metastases of CRC suggesting that this pathogenic bacterium could potentially migrate along with CRC cells to metastatic sites in the human body [43]. Adherence and invasion are believed to be essential mechanisms for colonization, dissemination, evasion of host defense, and induction of host response. It was found that binding interactions between Fap2 and the host sugar Gal-GalNAc (d-galactose-β(1-3)-N-acetyl-d-galactosamine), which is overexpressed in human colorectal adenocarcinoma, may provide an explanation for localization. Intravenously injected F. nucleatum localizes to mouse tumor tissues in a Fap2-dependent manner [48]. Fap2 also mediates binding to erythrocytes possibly confirming a hematogenous route to the tumor. Another protein, the fusobacterial FadA protein, binds to host E-cadherin, activates β-catenin signaling, and differentially regulates inflammatory and oncogenic responses. This leads to suppression of anti-tumor immunity and inhibits tumor killing by NK cells. The recently discovered FplA (Fusobacterium phospholipase autotransporter) from F. nucleatum is a Type Vd autotransporter phospholipase and has a proposed role in altered host signaling and evasion of autophagy [49]. F. nucleatum strains are enriched in Type V secreted autotransporters, which are large Gram-negative specific bacterial virulence factors critical for binding and entry into host cells. It is hypothesized that upon intracellular invasion of the host, FplA could play a role in phagosomal escape, subversion of autophagy, or eicosanoid mediated inflammatory signaling, and it was shown that FplA

287

288

9 Microfluidics in Microbiome and Cancer Research

binds with high affinity to host phosphoinositide signaling lipids critical to these processes [49]. Characterization of these secreted autotransporters is currently an area of active research. Apart from microbial associations in gastric cancer and CRCs, there have been studies linking gastrointestinal microbiota in cancers present in other organs including the pancreas [50, 51], liver [52], prostate [53], and breast tissue [54]. For example, Helicobacter hepaticus along with H. pylori play a role in toxin and virus induced hepatocellular carcinoma [55] and Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans, members of the oral microbiome, have been associated with an increased risk of pancreatic cancer [37, 56]. F. nucleatum and Granulicatella adiacens were found to be enriched in the intraductal papillary mucinous neoplasms (IPMNs) of patients [57]. IPMNs are pancreatic cysts which can progress to invasive pancreatic cancer and further investigation is needed to understand the role of microbes in these cancers. Interestingly, P. gingivalis was recently identified in the brain of Alzheimer’s disease patients giving rise to new questions about bacterial invasion and activity in disease progression generally thought to be independent of microbe involvement [58]. 9.2.5

Bacterial Invasion

Research has shown that some bacteria, when systematically administered, are naturally capable of homing to tumors resulting in high levels of local replication [38]. This indicates that tumor specific factors attract the bacteria and promote their adherence and invasion. Aberrant neovasculature and local immune suppression are likely to be major elements in this process. The leaky blood vasculature and slow blood flow in tumor tissue allows circulating bacteria to enter tumors and embed locally [38]. Their ability to survive and proliferate in this environment is influenced by the presence of nutrients in the hypoxic region of the tumor. Hypoxia provides a unique growth environment for strictly and facultative anaerobic bacteria [38]. Bacterial nutrients such as purines are abundant in the necrotic regions of the tumor in addition to chemoattractants such as aspartate, serine, citrate, ribose, and galactose that are produced by quiescent cancer cells [38]. In the case of F. nucleatum, it is documented that this bacterium does not utilize host carbohydrates as a carbon source, and therefore the availability of proteins and amino acids in this environment is critical for survival and colonization. Bacteria use highly complex mechanisms for cell adhesion and invasion, which range from single monomeric proteins to intricate multimeric macromolecules that function as sophisticated nanomachines to penetrate the host [59]. They exhibit cell-surface adhesins that can recognize many elements of the host-cell surfaces, including components of the ECM such as collagens, laminins, elastin, proteoglycans, fibronectin, and hyaluronan [59]. In Yersinia sp., invasin, a surface protein first discovered in 1987, binds with high affinity to members of the β1 chain integrin family which normally binds fibronectin [59]. This engagement leads to activation of GTPases such as Rac1, an actin regulator, that modulates phosphatidylinositol metabolism to induce actin rearrangements at the site of bacterial entry to promote invasion of the bacteria [59]. In the gastrointestinal tract, bacteria can translocate and invade tissue through transcellular (through

9.3 Studying the Microbiome

enterocytes) or paracellular (using tight junctions) mechanisms. They can also enter through the enteric venous system to the portal vein and following lymphatic drainage [59]. These processes are major focuses of research as the precise mechanisms by which several microorganisms translocate, survive, and proliferate in extra-intestinal tissues is currently unknown. Particularly, the ubiquity of these interactions in cancer is relatively unknown and new technologies are required to obtain insight into these mechanisms. Despite the examples of tumor-promoting bacteria that are described, it should be remembered that most bacteria in our microbiome are not harmful. In fact, it has been shown that in some cases there exists a correlation between bacterial infection and tumor regression. In 1868, William Coley (the “father of immunotherapy”) [60] identified the presence of Streptococcus pyogenes in soft tissue sarcoma, and was one of the first to characterize concomitant infection of tumors and how this could lead to remission of incurable neoplastic malignancy [38]. Bacterial products can also have therapeutic applications and can affect chemotherapeutic response [61]. It was recently discovered that N-(3-oxododecanoyl)-l-homoserine (OdDHL), a quorum-sensing molecule from Pseudomonas aeruginosa that regulates bacterial stress responses, has a selective effect in inhibiting proliferation and inducing apoptosis in breast cancer cells [62]. In another study, quorum sensing peptides like PhrG from Bacillus subtilis, competence stimulating peptide (CSP) from Streptococcus mitis and extracellular death factor (EDF) from E. coli and its tripeptide analogue were able to promote tumor cell invasion and angiogenesis influencing metastasis [63]. Bacteria also greatly influence drug toxicity and efficacy and can be harnessed in bacteriotherapy [41]. One interesting proof-of-concept study showed that the tumor targeting bacterium, Salmonella enterica serovar Typhimurium VNP20009 can be exploited for the targeted delivery of nanotherapeutics. The bacterium’s surface is functionalized with poly(lactic-co-glycolic acid) (PLGA) nanoparticles and when injected into solid tumors these nanoparticles showed enhanced retention and distribution in solid tumors [64].

9.3 Studying the Microbiome Various approaches can be used to study the microbiome and host-microbiome interactions. The majority of early findings in this field have been correlative and associative [28, 65–67]. To obtain actionable insights to develop therapeutic interventions and precision diagnostics leveraging aspects of the microbiome, it is important to tailor the experiments to the level of complexity and resolution that is required. One can study the macroscopic community diversity and dynamics or zoom into the level of individual bacterial species or still further probe the molecular interactions of individual proteins. Usually, host-microbiome interaction studies begin with selecting a phenotype linked to a disease and seeking the microbial taxa or genes responsible for it. Fischbach [65] compares this approach to that of “forward genetics” but with two distinct genomes (host and microbe) and utilizing differential sequence

289

290

9 Microfluidics in Microbiome and Cancer Research

analysis to obtain the bacterial taxon responsible for the observed phenotype. These studies have shown that disease phenotypes can be transferred using a microbiome transplant in germ-free mice [68, 69] and even identifying individual microbial species that modulate host phenotypes [70, 71]. However, the main challenge of this approach is discriminating strains that are causally related to disease phenotypes and establishing the mechanistic link to the observed phenotype [65]. Contrarily, the “reverse genetics” approach involves determining the causative microbial strain by colonizing mice with microbial communities that differ only by a single strain or alternatively colonizing mice with wild-type versus mutant version of the strain [65]. This approach can more directly determine the causative strain; however, this approach suffers from the disadvantage of cherry-picking the likely pathogenic strains. These studies can be supplemented with the rational design and development of complex model microbiomes, improvements in technology to directly observe host-microbiome interactions, improved animal models, and scaled-up in vitro models that adequately recapitulate human gut physiology. One of the most significant challenges to overcome in designing experiments to study the microbiome is the ability to control and manipulate multi-species microbial communities while avoiding over-proliferation of individual species [72]. It has been proven to be extremely difficult to co-culture microbes with living epithelium for more than one day using conventional culture models or with more sophisticated intestinal organoid cultures [73]. The microbial cells overgrow and compromise intestinal barrier function. As we will see below, conventional 2D and 3D cell culture systems struggle to recapitulate the physiological conditions including the microenvironment of the tissue and thus fail to mimic the salient features of the intestinal walls including the mucosa and muscular layers. Using static cultures further inhibits the differentiation of intestinal villi and fails to confine microbial growth. The peristaltic-like motions within the intestine are especially crucial for this function [74]. Due to these limitations, the analysis of the gut microbiome has been limited to genetic or metagenomic analysis using techniques like 16S ribosomal RNA sequencing. Efforts have been directed to develop experimental in vitro and ex vivo models of the gut and other tissues using emerging technologies in order to study the microbiome. A third approach is to use surrogate animal models. However, studying host-gut microbiome crosstalk has been limited because the colonization and stable maintenance of a defined gut microbiome in these animal models is challenging. The real-time monitoring of the function of the microbiota in situ has also proven to be difficult. Additionally, the inability to create and maintain cultures of well-defined microbial communities to operate in natural environments has hampered progress in this field [75, 76]. May et al. [74], Park et al. [72], and Trujillo-de Santiago et al. [77] have extensively reviewed some of the conventional and current in vitro modeling technologies to study host-microbiome cross-talk. The following sections of this chapter describe the advantages and disadvantages of existing approaches, and propose a scope for future improvements in their design and construction using microfluidic technologies.

9.3 Studying the Microbiome

9.3.1

2D Models

2D models are the simplest in vitro models, and essentially contain human cell lines grown as a monolayer culture with specific microbial species inoculated into the culture media to subsequently interact with host cells. The basic concept has been re-engineered to suit specific requirements as shown in Figure 9.2. Some examples such as the Hanging Basket method use well plates which are modified so as to have a basket hanging over a monolayer of epithelial cells. An oral biofilm consisting of four pathogenic microbes was suspended on this basket over the cells and this set-up was used to study gingival inflammation. The biofilm was pre-grown on a coverslip before being attached to the basket [72]. Another method is to use a simple centrifuge tube consisting of host cells attached to a coverslip and positioned over a medium of bacterial cells to simulate the aerobic-anaerobic interface that occurs in the gut. Transwell approaches have also been used to quantify enhanced or compromised intestinal barrier function through the measurement of TEER or transepithelial electrical resistance. The cells are coated on an ECM or porous membranes within Transwell culture systems. The microbes are then added to the apical surface and any compromise to the barrier integrity can be measured using electrodes. This setup can be used to monitor immune interactions with immune cells added to the basolateral side. Although simple and robust, in vitro 2D cultures have several disadvantages. They fail to induce complete cell differentiation due to the lack of microenvironmental cues and hence are unable to replicate tissue-specific functions. Due to static culture conditions, overgrowth of microbes cannot be controlled and hence long-term cultures are difficult or impossible to maintain [72]. With respect to the gut epithelial models in particular, 2D cultures fail to recapitulate physiological 3D intestinal cell and tissue morphology and re-establish other key intestinal differentiated functions such as mucus production, villi formation, and cytochrome P450 based drug metabolism. Most importantly, drug efficacy and dosage studies on 2D models fail to translate effectively to the clinic. 9.3.2

3D Models

Due to the limitations encountered with 2D cultures, there has been a transition to 3D culture models. 3D cultures make it possible to re-create (generally in a more simplistic form) the ECM that surrounds most mammalian cells, which in turn makes it possible to tune the microenvironment around the cells using physical and chemical cues to induce differentiation and affect cell functionality and response [78]. Some examples of in vitro 3D culture methods include the rotating wall vessel, hydrogel scaffolds, and organoids (Figure 9.3) [17, 72]. The rotating wall vessel bioreactor essentially contains the epithelial cells adhered to beads coated with ECM within a rotating drum. The continuous rotation induces surface shear stress on the cells mimicking the forces experienced physiologically. Honer zu Bentrup et al. showed that when the bacterium S. enterica serovar Typhimurium is added to the media, they adhere to the epithelial cells and exhibit physiological invasion levels [79]. Hydrogel scaffolds made from biocompatible hydrogels such as hyaluronic acid, collagen, PLGA, calcium-alginate, and polyethylene glycol (PEG)-peptide

291

9 Microfluidics in Microbiome and Cancer Research

Hanging basket Oral biofilm

Gingival epithelium (a) w/o Caco-2

w/ Caco-2

Cell culture medium

Coverslip

Density gradient

No density gradient

292

Bacterial medium Bacterial cells

(b)

Intestinal epithelium

Intestinal epithelium Bacterial cells AP

BL

Nanoporous membrane

(c)

Figure 9.2 2D in vitro static models. (a) Hanging Basket co-culture model consisting of a pre-grown biofilm attached at the bottom of a hanging basket to induce interactions with gingival epithelium cultured on the bottom. (b) Using a centrifuge tube to co-culture Caco-2 cells with Faecalibacterium prausnitzii grown on a solid agar medium with Caco-2 cells overlaid on a cover slip. (c) A Transwell set up with Caco-2 cells seeded on the Transwell with media in the basolateral (BL) side and bacterial cells seeded on the apical (AP) side. Source: Park et al. 2017 [72]. Reproduced with permission of Springer Nature.

have been used to model the lumen of the intestinal tract. These gels can be used in conjunction with 3D bioprinting technologies to build multilayered structures to recapitulate tissue morphology and functionality [80, 81]. One example shows the use of silk fibrin-based 3D scaffolding to replicate the villi like intestinal architecture using Caco-2 and HT29-MTX cell lines in co-culture with primary human intestinal myofibroblasts [82]. This model shows enhanced mucus production, increased cytodifferentiation, as well as the formation of an

Lumen

No sedimentation

Salmonella enterica

Yersinia pseudotuberculosis

3D patterned small intestine

Porous bead

(a)

2 mm

5 mm

Collagen

400 μm

m 8m

Intestinal epithelium

H-InMyoFib

Caco-2 & HT29-MTX

Po2

Silk

9.3 Studying the Microbiome

(b) Stomach organoid

Suction

H. pylori

Glass capillary

Microinjection needle

Control

Helicobacter pylori infection

Nuclei, CDH1, H. pylori

(c)

Figure 9.3 3D culture models to study host-microbiome cross-talk. (a) Rotating Wall Vessel that demonstrates adherence of Salmonella enterica to epithelial cells. (b) Silk-fibrin based hydrogel to recreate intestinal lumen and infected with Yersinia pseudotuberculosis. (c) Microinjection of Helicobacter pylori into stomach organoids. Source: Park et al. 2017 [72]. Reproduced with permission of Springer Nature.

oxygen gradient [82]. The authors also successfully tested infection with Yersinia pseudotuberculosis, and Lactobacillus rhamnosus GG (LGG) to demonstrate its utility for microbiome studies. Organoid cultures are growing in significance in the research community and are extensively used to study both basic and clinical biology [83–86]. They are defined as a 3D structure grown from stem cells consisting of organ-specific cell types that self-organize through cell sorting and spatially restricted lineage commitment [87]. The intestinal 3D organoid cultures can be derived from either the intestinal crypt containing endogenous intestine cells or from induced pluripotent stem cells, and can sustainably support various differentiated intestinal epithelial cell subtypes in vitro [88]. It has been shown that small intestinal organoids (enteroids) can spontaneously undergo villus-crypt morphological organization and intestinal histogenesis when cultured within a 3D ECM gel (Matrigel) in a medium containing Wnt, R-spondin, noggin, and additional growth factors [73]. For host-microbe interaction studies, microbes have been injected into the lumen of organoids or added directly to the media containing the organoids [89]. The intestinal organoids have been used in proof of concept studies to model infection. Forbester et al. generated intestinal organoids from human induced pluripotent stem cells and studied their interaction with S. enterica serovar Typhimurium [89, 90]. H. pylori, which is a known pathogen in stomach ulcers and a designated Type 1 carcinogen [91], has been shown

293

294

9 Microfluidics in Microbiome and Cancer Research

to recapitulate well-known hallmarks of infection upon microinjection into gastric organoids [92–95]. It was shown that bacterial colonization led to an increase in proliferation of Lgr5+ stem cells influenced by the CagA virulence factor produced by the bacteria [95]. Lgr5+ stem cells are the cells found at the bottom of the crypt that enable the formation of crypt-villus structures. In an organoid model of the infection of the obligate anaerobe, C. difficile, its toxins Toxin A (TcdA) and Toxin B (TcdB) were microinjected into stem cell derived intestinal organoids which resulted in disruption of the epithelial barrier [85, 96]. It was further observed that there were E-cadherin and actin cytoskeletal rearrangements only in the organoids injected with TcdA and not those injected with TcdB. These organoids represent highly modular and adaptable systems to study the host-microbe interface. However, some of the disadvantages of organoids include a lack other supporting cell and tissue types found within the living intestine, such as endothelium lined blood vessels and immune cells, which are important for drug transport, pharmacokinetic analysis, and disease modeling [86]. Additionally, it is not possible to replicate the fluid flows and cyclic mechanical deformation experienced during the peristaltic motion within the intestine that contributes significantly to intestinal health and function. Experimentally, it is challenging to manipulate or extract the luminal contents of the organoids and microbial overgrowth still remains a problem in these culture systems. There is much variability amongst experimental preparations due to heterogenous organoid shape and structure and batch to batch variation in Matrigel and growth factor sources [17]. Finally, there is the high cost for the maintenance of organoids and the need for well-defined ECM cell culture reagents that support organoid growth [90]. Spheroids provide a simpler approach compared with organoids. 3D cultures that incorporate the use of hydrogels are quite common to develop cellular spheroids. Materials such as PEG, PLGA, and poly(N-isopropylacrylamide) are used. The advantages include better cell differentiation, development of the polarized GI surface, and extended co-culture [72]. Other 3D constructs include models developed by SynVivo and PCTS (precision cut tissue slices). SynVivo uses microfluidic models which recapitulate in vivo relevant flow and pressures to study human tissue microvasculature and cell drug interactions in a 3D model [74]. These technologies have been used to study the blood–brain barrier, cancer, inflammation, and toxicology. Some of their synthetic network assays can model leaky tumor vasculature which can be used to study microbial movement within the tumor microenvironment. PCTS are used to model solid organs (liver, kidney) and agarose embedded non-solid organs intestine. The prevailing assumption is that the cells within the tissue will function optimally when they are cultured in conditions similar to their in vivo microenvironment [97]. They have certain advantages over organoids and organ-on-a-chip since they represent all regions of the tissue [74]. However, currently they are only amenable to aerobic bacterial cultures and further work needs to be done to adopt the tissue slices to anaerobic microbes [98].

9.3 Studying the Microbiome

9.3.3

Organ-on-a-Chip and the Application of Microfluidics

Some of the challenges associated with basic 2D and 3D cultures have been overcome by the development of microfluidic Organ-on-a-Chip models [99]. These microfluidic models have been developed to mimic the functionality of several organs and tissues including the lung [100–102], liver [103], pancreas [104], small intestine [105], neuronal tissue [106, 107], kidney [108, 109], heart [110], as well as tumor models [111–113]. Even multi-organ chips have been developed [114, 115]. One intriguing example is that of the design of a microfluidic chip coupling pancreatic islets and liver spheroids to model Type 2 diabetes ex vivo by Bauer et al. which artificially maintains glucose homeostasis within the device [104]. A combination of the advantages offered by 3D culture models with that of coupling flow and mechanical movement using microfluidics is highly valuable in recapitulating tissue differentiation and function and most importantly, the long-term sustainability of the culture models. Barkal et al. [116] reviewed the ways microscale systems can be used to study complex host-pathogen interactions (Figure 9.4). Microscale approaches allow for the precise tuning of shear forces and the development of exact chemical gradients that are quite difficult to obtain at the macroscale. In fact, they enable spatial patterning of cell cultures, extraction and analysis of low concentration soluble factors, and even help in developing organotypic models that closely approximate the cellular microenvironment. Microfluidic devices can be fabricated using soft lithography techniques and consist of channels and chambers which can be perfused with media and can house cells and spheroids [99]. Polydimethylsiloxane (PDMS) is the most common biomaterial used to build microfluidic platforms due to its excellent optical properties, gas permeability, and biocompatibility. The mold can be designed to incorporate a number of microstructures that support different functionalities including fluid storage, transport, mixing, and splitting [117]. The culture medium can be perfused through the devices using a syringe or a peristaltic pump. The resulting laminar flow profiles can be dynamically controlled to recapitulate physiological fluid flow and associated shear stresses on the cell surfaces as observed in the human intestinal lumen and blood capillaries, forces which are key to differentiation and morphogenesis. The microfluidic control also enables the precise delivery of nutrients, growth factors, drugs, and toxins in a highly regulated spatiotemporal manner [73]. These platforms have been engineered extensively to incorporate multiple cell types, tissues, microvascular channels, as well as immune cells and bacteria and provide high spatial and temporal resolution to enable single cell tracking. In microfluidics, flows are incompressible and result from a balance between viscous and pressure forces. Mixing is slow and aids in the creation of chemical gradients which can be increased in complexity by modifying the design of the microfluidic system. Soluble factor gradients are important in immune response to infection. Diffusion forces can be made to dominate over convective mixing and can be used to generate and study reproducible gradients. With respect to studying the microbiome, fluids in the body such as mucus, blood, and urine play a significant role in clearing pathogens through shear and

295

296

9 Microfluidics in Microbiome and Cancer Research

(a) Infection at the epithelial barrier

(b) Microscale models

Virulence strategies Mucosal adherence Antigen camouflage Phagocyte-hijacking Host protein interference Biofilm formation Multispecies coordination Secondary metabolites

Viruses

1. Microbial interactions under shear forces

Fungi 3

4

Soluble factors

Epithelium

2. Chemical gradients and immune recruitment

Bacteria Chemokines Mucus

1

3. Soluble factor communication

2 Recruited immune cells

Defense strategies Epithelial barrier Mucociliary flow Fluid flow Phagocytosis Cytokine response Immune cell recruitment NET formation

4. 3D architecture of the infection environment Vasculature Resident immune cells Stromal cells

Figure 9.4 The complex host pathogen microenvironment can be modelled using microfluidics. (a) The virulence strategies of the microbes and the defense strategies of the host at the epithelial barrier forming the complex infection microenvironment. (b) Four aspects of the infection microenvironment that can be readily modeled including fluid flow, gradients for chemotaxis, co-culture and soluble factor communication, and 3D organotypic modeling. Source: Barkal et al. 2017 [116]. https://journals.plos.org/plospathogens/article? id=10.1371/journal.ppat.1006424. Licensed under CC BY 4.0.

binding forces to wash away pathogens. Additionally, the shear forces can be exploited by the pathogens and they tend to form clusters in regions of high shear. These effects can be replicated in microfluidic platforms. For example, Andersen et al. developed a microfluidic device to study the invasion of bladder epithelial cells by uropathogenic E. coli. They exposed the invaded bladder epithelial cells (BEC) cells to urine flow and identified the outgrowth of filamentous bacteria replicating observations in vivo in mouse models [118]. In most infections, the microenvironment tends to contain multiple microbial populations which interact chemically and physically. In fact, some metabolites and small molecules are produced only when the organisms are in mixed/co-culture populations. For example, Cryptococcus neoformans increases melanin synthesis when in coculture with Klebsiella aerogenes. Thus co-culture stability is greatly facilitated by the microscale control of the spatiotemporal distribution of cells and soluble factors [116]. The subtle balances of chemicals and metabolites as well as the relative placement of cells and microbes in organs is required to accurately model infections. Mechanobiology of the cells may also play a significant role in the probability of infection [119]. Responses to environmental stiffness and shear stress can activate certain signaling pathways that may matter to microbial cells. Microfluidics

9.4 Microfluidic Intestine Chip Models

and bioprinting technologies may be able to incorporate some of these physical cues which can mimic salient features of organs such as the lung and the digestive tract [116].

9.4 Microfluidic Intestine Chip Models There have been several significant breakthroughs in the development of microfluidic chips to study host-microbiome crosstalk. The Ingber Lab has developed Gut-on-a-Chip and Intestine Chip models which support co-culture of host intestinal epithelium with specific microbial populations. Recent improvements to their device additionally support the culture of anaerobic microbial communities. Their chips consist of two hollow channels separated by a porous, ECM-coated membrane with immortalized intestinal epithelial cells coated on one side. This layer can be probed from both the apical and basolateral sides and is amenable to studies on tight junction barrier function [73]. Some of these devices integrate multiple chambers to study pharmacokinetic and pharmacodynamic properties of drugs. These models were improved with the addition of microfluidic flow and micro-molding of polymeric scaffolds into crenulated surfaces to mimic a villus-like structure. Culturing Caco-2 cells on these surfaces promoted the formation of a similar crenulated epithelium, and showed increased absorption of drug compounds and enhanced cytochrome P-450 activity in response to fluid shear [120]. Another model, the modular HuMiX (human-microbial-crosstalk) platform developed by Shah et al. [121] was among the first to support anaerobic microbial co-culture. The construction and salient features of these models are described below. 9.4.1

Gut-on-a-Chip Model

The “Gut-on-a-Chip” model was first developed in the Ingber Lab and enables the growth and co-culture of human intestinal epithelium, capillary endothelium, immune cells, and even commensal microbes [122, 123]. Its most important characteristic is that it mimics physiologically relevant fluid flow and peristalsis-like mechanical deformations in vitro. The Gut-on-a-Chip model is made of PDMS so that it allows high-resolution imaging by phase contrast, differential interference contrast, or immunofluorescence confocal microscopy. It has two parallel microchannels (