Bioelectrosynthesis 9783030032982, 3030032981

Table of contents :
Front Matter ....Pages i-ix
Electrification of Biotechnology: Status quo (Falk Harnisch, Dirk Holtmann)....Pages 1-14
Extracellular Electron Transfer and Biosensors (Francesca Simonte, Gunnar Sturm, Johannes Gescher, Katrin Sturm-Richter)....Pages 15-38
Electron Transfer Between Enzymes and Electrodes (Tanja Vidakovic-Koch)....Pages 39-85
Enzyme-Based Electrobiotechnological Synthesis (Lisa Marie Schmitz, Katrin Rosenthal, Stephan Lütz)....Pages 87-134
Engineering of Microbial Electrodes (Sven Kerzenmacher)....Pages 135-180
Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering (Miriam A. Rosenbaum, Carola Berger, Simone Schmitz, Ronny Uhlig)....Pages 181-202
Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates (Annemiek ter Heijne, Florian Geppert, Tom H. J. A. Sleutels, Pau Batlle-Vilanova, Dandan Liu, Sebastià Puig)....Pages 203-229
Reactors for Microbial Electrobiotechnology (Thomas Krieg, Joana Madjarov, Luis F. M. Rosa, Franziska Enzmann, Falk Harnisch, Dirk Holtmann et al.)....Pages 231-271
Modeling Microbial Electrosynthesis (Benjamin Korth, Falk Harnisch)....Pages 273-325
Electrochemical Applications in Metal Bioleaching (Christoph Kurt Tanne, Axel Schippers)....Pages 327-359
Generating Electric Current by Bioartificial Photosynthesis (Babu Halan, Jenny Tschörtner, Andreas Schmid)....Pages 361-393
Electrification of Biotechnology: Quo Vadis? (Dirk Holtmann, Falk Harnisch)....Pages 395-411
Erratum to: Engineering of Microbial Electrodes (Sven Kerzenmacher)....Pages 413-413
Back Matter ....Pages 415-420

Citation preview

Advances in Biochemical Engineering/Biotechnology  167 Series Editor: T. Scheper

Falk Harnisch Dirk Holtmann Editors

Bioelectrosynthesis

167 Advances in Biochemical Engineering/Biotechnology Series editor T. Scheper, Hannover, Germany Editorial Board S. Belkin, Jerusalem, Israel T. Bley, Dresden, Germany J. Bohlmann, Vancouver, Canada M.B. Gu, Seoul, Korea (Republic of ) W.-S. Hu, Minneapolis, Minnesota, USA B. Mattiasson, Lund, Sweden J. Nielsen, Gothenburg, Sweden H. Seitz, Potsdam, Germany R. Ulber, Kaiserslautern, Germany A.-P. Zeng, Hamburg, Germany J.-J. Zhong, Shanghai, Minhang, China W. Zhou, Shanghai, China

Aims and Scope This book series reviews current trends in modern biotechnology and biochemical engineering. Its aim is to cover all aspects of these interdisciplinary disciplines, where knowledge, methods and expertise are required from chemistry, biochemistry, microbiology, molecular biology, chemical engineering and computer science. Volumes are organized topically and provide a comprehensive discussion of developments in the field over the past 3–5 years. The series also discusses new discoveries and applications. Special volumes are dedicated to selected topics which focus on new biotechnological products and new processes for their synthesis and purification. In general, volumes are edited by well-known guest editors. The series editor and publisher will, however, always be pleased to receive suggestions and supplementary information. Manuscripts are accepted in English. In references, Advances in Biochemical Engineering/Biotechnology is abbreviated as Adv. Biochem. Engin./Biotechnol. and cited as a journal. More information about this series at http://www.springer.com/series/10

Falk Harnisch • Dirk Holtmann Editors

Bioelectrosynthesis With contributions by P. Batlle-Vilanova
C. Berger
F. Enzmann
F. Geppert
J. Gescher
B. Halan
F. Harnisch
D. Holtmann
S. Kerzenmacher
B. Korth
T. Krieg
D. Liu
S. L€ utz
J. Madjarov
S. Puig
K. Rabaey
L. F. M. Rosa
M. A. Rosenbaum
K. Rosenthal
A. Schippers
A. Schmid
L. M. Schmitz
S. Schmitz
F. Simonte
T. H. J. A. Sleutels
G. Sturm
K. Sturm-Richter
C. K. Tanne
A. ter Heijne
J. Tsch€ortner
R. Uhlig
T. Vidakovic-Koch

Editors Falk Harnisch Department of Environmental Microbiology Helmholtz-Centre for Environmental Research GmbH - UFZ Leipzig, Germany

Dirk Holtmann Industrial Biotechnology DECHEMA-Forschungsinstitut Frankfurt am Main, Germany

ISSN 0724-6145 ISSN 1616-8542 (electronic) Advances in Biochemical Engineering/Biotechnology ISBN 978-3-030-03298-2 ISBN 978-3-030-03299-9 (eBook) https://doi.org/10.1007/978-3-030-03299-9 Library of Congress Control Number: 2018964403 © Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

The transformation of our current fossil fuel-based economy—and hence society as a whole—depends on different factors. Each of these factors, such as the biobased economy, bioeconomy, circular economy, and sustainable economy, slightly emphasizes one aspect over the others. However, they stand united when it comes to one point: the material resources and energy carriers that are exploited each year need to be recovered by nature within a similar time frame. In the words of Wilhelm Ostwald, who won the Nobel Prize in 1909: “Die dauerhafte Wirtschaft muß ausschließlich auf die regelm€ aßige Benutzung der j€ ahrlichen Strahlungsenergie begr€ undet werden [The permanent economy has to be based solely by the utilization of the annual solar energy]” [1]. The generation of electric energy from renewable resources, such as photovoltaics or wind turbines, is already an advanced technology. However, its storage and transformation to chemicals in electrochemical reactions is a key challenge. Furthermore, because of its versatility and inherent advantages for a green economy, biotechnology is undisputed as a key technology of the twenty-first century. This nexus of electric energy and biobased chemicals is the essence of electrobiotechnology or bioelectrotechnology. Biotechnology on its own is a highly interdisciplinary scientific discipline that embraces areas such as microbiology, chemistry, molecular biology, genetics, bioinformatics, and engineering sciences. It becomes even more complex for the integration of electrochemical reactions catalysed by biological moieties. These enzymatic or microbial electrochemical reactions are based on the interfacing of enzymes and whole cells with electrodes. The archetype of the respective devices— bioelectrochemical systems—is the microbial fuel cell (MFC), which generates electric energy from resources such as waste water. MFCs were first described more than 100 years ago. In recent decades, a plethora of applications in electrobiotechnology were proposed using bioelectrosynthesis—that is, the production of fine and platform chemicals based on enzymatically or microbially catalyzed

v

vi

Preface

reactions driven by electric energy (as the most prominent example). However, for bringing microbial and enzymatic electrosynthesis from the laboratory bench to industrial applications, we must address the inherent complexity of the complete process chain. The aim of the book Bioelectrosynthesis is to address the experimental state of the art, the envisaged applications, and the routes to close the current knowledge gap between them. The high interdisciplinarity in the research and development of microbial and enzymatic electrosynthesis is reflected in the conception and diverse content of the chapters within. This book aims to complement the large number of publications on biofuel cells and the specific aspects of electrobiotechnology. We hope that it can stand as a single authoritative source because, for the first time, it brings together the relevant developments in the field of bioelectrosynthesis. We intent to address scientists who are already active in this field as well as those who want to familiarize themselves with this fascinating area. Furthermore, this book is addressed to scientists who focus only on one of the previously mentioned disciplines. We hope they can transfer their knowledge, methods, and findings to the wealth of electrobiotechnology. This book may also contain some helpful information for stakeholders, funding agencies, and policy advisors. Finally, the editors hope that this book will stimulate the growth and flourishing of the promising field that has gathered under the umbrella of the International Society of Microbial Electrochemistry and Technology. Leipzig, Germany Frankfurt am Main, Germany

Falk Harnisch Dirk Holtmann

Reference 1. Ostwald W (1909) Energetische Grundlagen der Kulturwissenschaft. Verlag von Dr. Werner Klinkhardt, Leipzig, p 44

Contents

Electrification of Biotechnology: Status quo . . . . . . . . . . . . . . . . . . . . . . Falk Harnisch and Dirk Holtmann

1

Extracellular Electron Transfer and Biosensors . . . . . . . . . . . . . . . . . . . Francesca Simonte, Gunnar Sturm, Johannes Gescher, and Katrin Sturm-Richter

15

Electron Transfer Between Enzymes and Electrodes . . . . . . . . . . . . . . . Tanja Vidakovic-Koch

39

Enzyme-Based Electrobiotechnological Synthesis . . . . . . . . . . . . . . . . . . Lisa Marie Schmitz, Katrin Rosenthal, and Stephan L€utz

87

Engineering of Microbial Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Sven Kerzenmacher Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Miriam A. Rosenbaum, Carola Berger, Simone Schmitz, and Ronny Uhlig Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Annemiek ter Heijne, Florian Geppert, Tom H.J.A. Sleutels, Pau Batlle-Vilanova, Dandan Liu, and Sebastia Puig Reactors for Microbial Electrobiotechnology . . . . . . . . . . . . . . . . . . . . . 231 Thomas Krieg, Joana Madjarov, Luis F.M. Rosa, Franziska Enzmann, Falk Harnisch, Dirk Holtmann, and Korneel Rabaey Modeling Microbial Electrosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Benjamin Korth and Falk Harnisch Electrochemical Applications in Metal Bioleaching . . . . . . . . . . . . . . . . 327 Christoph Kurt Tanne and Axel Schippers vii

viii

Contents

Generating Electric Current by Bioartificial Photosynthesis . . . . . . . . . 361 Babu Halan, Jenny Tsch€ortner, and Andreas Schmid Electrification of Biotechnology: Quo Vadis? . . . . . . . . . . . . . . . . . . . . . 395 Dirk Holtmann and Falk Harnisch Erratum to: Engineering of Microbial Electrodes . . . . . . . . . . . . . . . . . 413 Sven Kerzenmacher Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

About the Editors

Falk Harnisch is a group leader at the Helmholtz Centre for Environmental Research – UFZ in Leipzig (Germany), a post he has occupied since 2012. He earned his PhD in Environmental Chemistry (2009) and a Diploma in Biochemistry (2006), both from the University of Greifswald (Germany), and his Habilitation in Biophysical Chemistry (2016) from the University of Leipzig. He has received numerous awards and scholarships, most recently the Biotechnology 2020+ research award from the German Federal Ministry of Education and Research (2012), and a research award from the Helmholtz Centre for Environmental Research – UFZ (2015). Prior to his current post he was a post-doctoral fellow and senior researcher at the Braunschweig University of Technology (Germany) and a visiting academic at the University of Queensland (Australia). He has now been working at the interface of microbiology and electrochemistry for over a decade. His research interests cover the fields of microbial electrochemistry and microbial electrochemical technologies, as well as electrochemistry for energy conversion and storage, especially electroorganic chemistry. Dirk Holtmann was born in Bremen, Germany, and completed his diploma in chemical engineering/biotechnology in 1999 at University of Applied Science Emden. He obtained his PhD, on the electrochemical measurement of microbial activities in fermentation, at the Otto von Guericke University of Magdeburg. He is now group leader of the biochemical engineering group at the DECHEMA Research Institute in Frankfurt, Germany. The group’s research focuses on combinations of different disciplines – such as bio-, electro-, and chemocatalysis; molecular biology; and process engineering – to develop novel production routes. His current research activities concentrate on biocatalysis and biotransformation, with the development and evaluation of electroenzymatic processes, as well as microbial electrosynthesis.

ix

Adv Biochem Eng Biotechnol (2019) 167: 1–14 DOI: 10.1007/10_2017_41 © Springer International Publishing AG 2017 Published online: 6 December 2017

Electrification of Biotechnology: Status quo Falk Harnisch and Dirk Holtmann Abstract Interfacing microbial, enzymatic, and electrochemical transformations has led to the new field of electrobiotechnology. Among the plethora of applications (including electric energy generation via pollutant removal), the synthesis of chemicals and energy carriers (e.g. H2) has sparked great interest. The linked transformation of chemical and electric energy may allow the joint utilization of renewable feedstock and sustainable electricity to gain commodities and fuels. The overall field is now referred to as bioelectrosynthesis and is a focus of this book. Starting with the rationale for using bioelectrosynthesis in a bioeconomy, this chapter provides a brief introduction to the field of electrobiotechnology. Subsequently, the chapter discusses the framework for bioelectrosynthesis, which is based on enzymes as well as microorganisms, and provides a definition of bioelectrosynthesis. The chapter concludes with a short overview on the history of the field.

F. Harnisch (*) Department of Environmental Microbiology, Helmholtz-Centre for Environmental Research, Leipzig, Germany e-mail: [email protected] D. Holtmann (*) DECHEMA-Forschungsinstitut, Industrial Biotechnology, Frankfurt am Main, Germany e-mail: [email protected]

2

F. Harnisch and D. Holtmann

Graphical Abstract

partly adapted from Natalia Moskovkina under CC BY 4.0

Electrification of Biotechnology: Status quo

Keywords Bioelectrotechnology, Electroenzymatic Electrobiotechnology, Microbial electrosynthesis, Microbial fuel cells

3

synthesis,

Contents 1 Bioelectrosynthesis As Part of the Bioeconomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Fields of Application in Electrobiotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.1 Generation of Electric Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 Cleaning of Water and Soil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.3 Water Recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.4 Sensors and Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.5 Bioelectrosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 3 Definition of Bioelectrosynthesis and the Aim of This Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 4 Bioelectrosynthesis As Green Technology? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 5 A Short and Selective History of Electrobiotechnology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1 Bioelectrosynthesis As Part of the Bioeconomy Fossil resources such as oil, gas, and coal are finite. The already significant and foreseeable growing demand for chemical-based products and processes has changed Earth drastically. Due to the increasing world population and rising living standards, the chemical industry is steadily growing. The global market volume for chemicals will more than double from 2015 to 2035, with sales of 2.6 trillion Euros [1]. Industries that produce and use chemicals have a significant impact on employment, trade, economic growth, and thus societal welfare worldwide. At the same time, chemicals, especially those based or dependent on fossil resources, usually have adverse effects on the environment [2]. Currently, bio-based and ecoefficient processes for making chemical products are gaining increasing relevance. Most of the raw materials used in the chemical industry are petrochemicals, mainly crude oil. These finite fossil raw materials need to be replaced by renewable and sustainable resources. Furthermore, in the long run, the costs of oil and gas exploration are rising and climate change is forcing mankind to lower carbon and greenhouse gas emissions. Consequently, a new bio-based economy – a “bioeconomy” – has to be formed on national and global levels. In general, the transition from fossil-based raw materials to renewable resources is a grand challenge of the twenty-first century. The transition to a bioeconomy challenges established products, markets, and technologies but also opens unprecedented opportunities for new ones. A future bioeconomy should not only be based on renewable raw materials but also must assure circular utilization of resources and interlinking (bio)chemical transformations with the storage and utilization of electric energy [3]. Renewables in the electric energy sector, such as photovoltaics, face temporal fluctuations and spatial separation of source and sink, which creates a demand for storage and conversion technologies. However, linking the chemical and energy sectors cannot be achieved

4

F. Harnisch and D. Holtmann

(or only to very minor extent) with established technologies. This gap can be narrowed or even closed by electrobiotechnology. Electrobiotechnology is based on the combination of biotechnology and electrochemistry. It promises to create a new venue for the conversion of chemical and electric energy [4–6]. Electrobiotechnology covers a wide range of envisaged applications, from electric energy generation from wastewater via removal of decontaminants to the synthesis of complex chemicals, as briefly illustrated in this chapter. The latter application is termed bioelectrosynthesis and represents a technology platform that may overcome several of the fundamental challenges of a future bioeconomy.

2 Fields of Application in Electrobiotechnology Electrobiotechnology covers a broad variety of applications: energy, biotechnology, environmental, biosensing, and biocomputing, as illustrated in Fig. 1. This includes microbial electrochemical technologies (MET) and enzymatic electrochemical technologies. The most relevant and advanced fields of application are described in the following sections.

Autonomous devices

Energy applicaons

Biophotovoltaics

Environmental applicaons

Remediaon Desalinaon Resource recovery Waste water cleaning

Energy harvesng from waste water

Whole-cell biosensors

Biosensing & Biocompung

Biotechnological applicaons

Enzymac biosensors LOGIC gates Electrochemically steered fermentaon Enzymac electrosynthesis Microbial electrosynthesis

Fig. 1 Illustration of the different fields of electrobiotechnology and photograph of a biofilm electrode

Electrification of Biotechnology: Status quo

2.1

5

Generation of Electric Power

Microbial and enzymatic electrochemical technologies can convert (part of) the chemical energy of substances and mixtures thereof into electric energy. These include the archetypical technologies of microbial fuel cells (MFCs) [7, 8] and enzymatic fuel cells (EFCs) [9, 10]. The “fuels” range from pure substances such as acetate and glucose via wastewater to organic and inorganic matter contained in sediments. In particular, the development of microbial fuel cells is on the verge of transitioning to real-world applications [11, 12].

2.2

Cleaning of Water and Soil

Microbial electrochemical technologies are used for the cleaning of wastewater from its organic load (usually expressed as chemical oxygen demand) while harvesting energy, as previously mentioned. Furthermore, pollutants such as X-ray contrast agents [13] can be removed from wastewater, as well as nitrobenzene and nitrate from groundwater [14, 15]. MET have also been demonstrated for use in soil remediation, such as for hydrocarbons or metals [16, 17]. MET can also be used for recycling metals from solids and liquid streams, as discussed in [18].

2.3

Water Recycling

MET can be used for water desalination, as first shown by Cao et al. [19]. This application hold promise for making drinking water available in arid regions.

2.4

Sensors and Circuits

Because the electric current is immediately linked to the enzymatic or microbial activity, MET can be used for generating analytical signals [20] (see also [21]).

2.5

Bioelectrosynthesis

MET can also be used for bioelectrosynthesis, the purposeful transformation of a substance (educt) into a desired product, as defined in the next section.

6

F. Harnisch and D. Holtmann

3 Definition of Bioelectrosynthesis and the Aim of This Book In a technical sense, chemical synthesis is defined as the purposeful execution of chemical reactions to obtain a product (or several products). Thereby, most chemical reactions need a catalyst (i.e. a substance that is involved but not converted during a chemical reaction) to take place at a sufficient rate. Catalysts can be different inorganic and organic materials, but also biological moieties such as enzymes or microorganisms; the latter are termed biocatalysts. Biocatalysts are exploited in different biotechnological applications, ranging from the production of food and beverages to the industrial synthesis of fine chemicals and biological wastewater treatment. A special type of biological moieties are those that can exchange electrons directly or via an electron shuttle with solid electron conductors (i.e. electrodes) [4, 6]. These moieties are termed bioelectrocatalysts; enzymatic and microbial electrocatalysts can be distinguished. Enzymes and microorganisms are both biological moieties that can be exploited for bioelectrosynthesis. They have numerous properties in common, especially when looking from the viewpoint of “classical” electrochemistry. At the same time, enzyme and microbial electrodes possess significant differences in regard to lifetime, costs, self-reproducibility, complexity of reactions (complete metabolism vs. limited number of reaction steps), surface coverage/size of catalysts, selectivity (for a product), and specificity (for substrate/ educt). It is not possible to determine which catalyst is more advantageous in electrobiotechnology in general. As a rule of thumb, enzymes are more specific and selective and should be favored for one- to three-step syntheses. Whole cell catalysts would be preferred for processes with complex reaction sequences and long reaction times. Bioelectrocatalysts can be used as wild-type organisms or enzymes, or they can be engineered to optimize electron transfer or product synthesis. In general, bioelectrocatalysis allows one to wire the flow of electric energy with the transformation of chemicals (catalyzed by the biological moieties). Therefore, we define bioelectrosynthesis as the purposeful execution of a combination of biologically (i.e. by microorganisms, organelles, and enzymes) catalyzed reactions with electrochemical reactions to transform an educt into desired products. This book is devoted to bioelectrosynthesis. It comprises the latest developments and addresses the research needs of the wiring of enzymatic and microbial conversions with the flow of electricity for the production of chemicals. Thereby, microbial and enzymatic electrosynthesis are at different levels of technological readiness; thus, the understanding of their fundamentals as well as the arsenal of engineering tools differs significantly. However, we believe that the field of bioelectrosynthesis is not only a fascinating research playground: It also holds great potential to fulfill (at least parts of) its great promises to give birth to an economically relevant and sustainable process in the future as part of the bioeconomy. We hope that this book will serve as valuable resource in this journey

Electrification of Biotechnology: Status quo

7

of research and development – not only for experts already working in the field, but especially for researchers starting their endeavors in enzymatic and microbial electrochemistry and technology. After this brief introduction, the fundamentals of microbial and enzymatic electron transfer with electrodes are discussed in [21, 22]. The applications of enzymatic electrosynthesis are summarized in [23], with microbial electrosynthesis being detailed in several contributions. First, a conceptual summary on the engineering of electrodes is provided [24]. Thereafter the current state of the art on microbial electrosynthesis using pure and mixed cultures as well as strain engineering [25] and microbiomes [26] are provided. This is followed by a comprehensive analysis on the need of reactors used for bioelectrosynthesis [27] and considerations for modeling approaches [28]. Thereafter, the fields of microbial metal bioleaching [18] and microbial photovoltaics [29] are summarized. Finally, the technology readiness levels of the different technologies as well as the economic perspectives are discussed [30].

4 Bioelectrosynthesis As Green Technology? In 1998, Paul Anastas and John Warner defined the term green chemistry: “Green Chemistry is the utilization of a set of principles that reduces or eliminates the use or generation of hazardous substances in the design, manufacture and applications of chemical products” [31]. The 12 principles of green chemistry have been an inspiring guideline for researchers from different fields throughout the years. These are also guidelines that can be taken as measures for developing bioelectrosynthesis (Table 1). A major advantage of bioelectrosynthesis is the capability to use electricity produced from renewable sources (e.g. sun, wind, wastewater) as an energy source to stimulate and enhance the microbial production of chemicals and fuels. Another important asset is the capability to reuse and upgrade low-value and waste compounds into useful commodities with industrial relevance, higher economic value, or higher energetic value. As a result, electrobiotechnological processes are often regarded as green technology. Furthermore, bioelectrosynthesis aims to develop a CO2-based bio-production. Carbon dioxide (CO2) is widely available from a variety of fossil-based and bio-based sources. The carbon feedstock CO2 can be obtained from flue gas emissions from high-emission industries such as ceramic, glass, and steel production and power plants, or from anaerobic digestion and fermentation plants. When exploiting CO2, a cheap source of reducing power is necessary. This energy can in principle be provided by excess renewable-based electricity, such as that obtained during peak production periods. The utilization of surplus electric power for producing chemicals is a highly important part of a future bio-economy. This link between the chemical and power sector can be created using bioelectrosynthesis. These considerations give a first impression that the combination of electrochemistry and biotechnology provides a suitable concept for sustainable green

8

F. Harnisch and D. Holtmann

chemistry. Therefore, bioelectrosynthesis should be widely considered as one of the key technologies fulfilling the 12 principles and thereby being green chemistry per se. When assessing new technologies, a certain tendency can be noted to pick a few of the 12 principles to underline the “greenness” of different processes, which certainly is in contrast to the envisioned use of the 12 principles as a cohesive system [33]. Table 1 shows the 12 principles in comparison to the impact of bioelectrosynthesis on the different topics. Obviously, bioelectrosynthesis could theoretically fulfill all 12 principles at the same time; however, this is not achievable in practice. Therefore, bioelectrosynthesis can be regarded as a potential green technology that needs a long way to go. One aim of this book is to show the efficiencies of the different processes to get a deeper insight into the current state of the art with regard to the environmental impact of the production systems.

5 A Short and Selective History of Electrobiotechnology The roots of electrobiotechnology date back to the year 1911 and M.C. Potter at the University of Durham, UK. He reported that an electric potential arises between two platinum wires that are immersed in aqueous solutions and separated by a diaphragm, if one of these remains sterile while the other faces the presence of microorganisms [34]. Of note, only the presence of some microorganisms (e.g. Escherichia coli, Saccharomyces cerevisiae) led to the formation of electrical potentials of up to 0.5 V. However, for a long time, these findings were only considered to be a scientific peculiarity, without any scientific or even technical relevance [35]. Therefore, research on these phenomena was scarce (e.g. Cohen et al. [36]) and only revived in the decade from 1960 to 1970 in the course of the NASA space program. Here, the driver for the development was the promise of the combined cleaning of water from human feces and urine and the exploitation thereof for electricity generation [37, 38]. This was the first time that microbial electrochemistry was considered to be of technical relevance and the archetype of microbial electrochemical technology was born: the MFC. Figure 2 shows the principles of a MFC. The endeavor of developing MFCs was terminated at that time because the development of photovoltaics was highly successful. Hence, spacecraft were powered by the sun and feces from astronauts and cosmonauts were seen as falling stars when burned up in the Earth’s atmosphere. Besides biosensors, only limited research on microbial and enzymatic electrochemistry was performed until the turn of the millennium. One main exception were gastrobots – that is, robots driven by electric energy from biomass, such as insects or fruits, that they digest in MFCs [39]. The year 2000 brought the groundbreaking discovery by Kim and colleagues that MFCs can operate on wastewater without adapations with artificial substances such as dyes [40] as redox shuttles for facilitating electron transfer (see [21]). Since then, the field of MFCs has flourished and the first projects on technical implementation were performed. The driving force for this advancement was similar to that

Electrification of Biotechnology: Status quo

9

Table 1 Comparison of the 12 principles according to Anastas and Warner [31] and the related contribution of bioelectrosynthesis

1 Prevention

2 Atom economy

3 Less hazardous chemical syntheses

Principles according to Anastas and Warner It is better to prevent waste than to treat or clean up waste after it has been created.

Synthetic methods should be designed to maximize the incorporation of all materials used in the process into the final product. Wherever practicable, synthetic methods should be designed to use and generate substances that possess little or no toxicity to human health and the environment.

4 Designing safer chemicals

Chemical products should be designed to affect their desired function while minimizing their toxicity.

5 Safer solvents and auxiliaries

The use of auxiliary substances (e.g. solvents, separation agents, etc.) should be made unnecessary wherever possible and innocuous when used. Energy requirements of chemical processes should be recognized for their environmental and economic impacts and should be minimized. If possible, synthetic methods should be conducted at ambient temperature and pressure. A raw material or feedstock should be renewable rather than depleting whenever technically and economically practicable.

6 Design for energy efficiency

7 Use of renewable feedstock

Contribution of bioelectrosynthesis The use of the immaterial reagent electron results in the avoidance of waste. Both electrochemistry and biocatalysis are highly specific and can be used to avoid the production of unwanted side products. Furthermore, different processes (e.g. microbial fuel cells or microbial electrolysis cells) exploit waste rather than producing it. The vast majority of materials in bioelectrosynthesis are incorporated in the final product, leading to high atom efficiencies. Self-sustaining microorganisms and enzymes replace toxic oxidants and/or reductants, which results in minimization of hazards. At the same time, the reaction products of the biocatalytic steps are usually nontoxic. The bio-based products of bioelectrosynthic processes can be considered to be usually biodegradable. After application of the products (e.g. bio-based plastics), they can easily degrade without negative effects to the environment. Water is mostly used in bioelectrosynthesis as the only (and safe) solvent.

The biological reactions proceed at ambient temperature and ambient pressure. Furthermore, in bioelectrosynthesis, quite high electron efficiencies (up to 90%) can be reached [32]. In bioelectrosynthesis, usually renewable feedstocks (complex mixtures containing, e.g., carbohydrates) or, even better, the greenhouse gas CO2 is used as carbon source. (continued)

10

F. Harnisch and D. Holtmann

Table 1 (continued)

8 Reduce derivatives

9 Catalysis

Principles according to Anastas and Warner Unnecessary derivatization (use of blocking groups, protection/ deprotection, temporary modification of physical/chemical processes) should be minimized or avoided if possible, because such steps require additional reagents and can generate waste. Catalytic reagents (as selective as possible) are superior to stoichiometric reagents.

10 Design for degradation

Chemical products should be designed so that at the end of their function they break down into innocuous degradation products and do not persist in the environment.

11 Real-time analysis for pollution prevention

Analytical methodologies need to be further developed to allow for real-time, in-process monitoring and control prior to the formation of hazardous substances.

12 Inherently safer chemistry for accident prevention

Substances and the form of a substance used in a chemical process should be chosen to minimize the potential for chemical accidents, including releases, explosions, and fires.

Contribution of bioelectrosynthesis Due to the high specificity of the biocatalysts, the application of upstream derivatization can hopefully be more or less completely avoided.

Bioelectrosynthesis exploits the combined features of biocatalysts and electrocatalysts. Electrocatalysts often show high stabilities and can be applied for long times. Biocatalysts are usually highly selective. Furthermore, microorganisms regenerate autonomously. The enzymes can be produced in situ. As well as any products from other bioprocesses, products from bioelectrosynthesis can very likely be degraded in the environment and do not persist in the environment. In general, hardly any hazardous products are produced in bioelectrosynthesis. Furthermore, process analytic technologies can be implemented in the reaction systems. Most of the products of and processes in bioelectrosynthesis are inherently safe, with water being used as a solvent and no or only slight overpressure.

for the NASA research: combining the cleaning of wastewater with electric energy generation, which turns an energy sink to an energy source. At the same time as the gastrobot research, the first reports on using the interaction of microorganisms and electrodes for the synthesis of chemicals were published. Hongo and Iwahara [41] were the first to report that microbial fermentations can be steered by electric current. Their publication on the example of a 10% increased yield of l-glutamic acid fermentation from glucose by Brevibacterium flavum was followed 10 years later by Gosh and Zeikus as well as Emde and Schink [42, 43]. They reported on enhanced yields in the fermentative production of

Electrification of Biotechnology: Status quo

outlet

inlet

outlet

H2O

fuel, substrate

O2, H+

Cathode

CO2, H+

Anode

Fig. 2 Principle sketch of a two-chamber microbial fuel cell. The anode is a biofilm electrode, where the oxidation of the substrate (e.g. wastewater) takes place. At the cathode, oxygen is reduced to water. Yellow arrows indicate electron flow

11

caons

anions

inlet

separator

butanol and propionate, respectively, in bioelectrochemical reactors. This work has formed the basis of current research on microbial electrosynthesis, which is not only ever expanding but also includes the conversion of CO2 into valuable products using pure and engineered microbes [25] as well as microbiomes [26]. In 2010, two groups published groundbreaking work concerning the conversion of CO2 and electrical energy to produce methane and acetate [32, 44]. The microbial production of bulk chemicals from carbon dioxide and water with electricity as the energy source motivated numerous researchers worldwide to investigate microbial electrosynthesis. The history of enzyme electrochemistry also dates back more than 50 years. The first demonstration of an amperometric enzyme electrode for the detection of glucose was described by Leland Clark in 1962 at a New York Academy of Sciences symposium [45]. In the beginning, the father of biosensors wanted to measure the reduction of oxygen with a platinum electrode to determine the oxygenation of blood. His first sensor failed because blood components were adsorbed on the surface of the electrode. Clark then had the ingenious idea of using a cellophane wrapper from a cigarette packet on his sensor. Only lowmolecular-weight substances – mainly oxygen – could reach the electrode and be measured. To calibrate his sensor, Clark added an enzyme, glucose oxidase (GOD), to the solution. Clark then developed the sensor further by entrapping concentrated GOD with another semi-permeable membrane in front of the electrode. Shortly after Clark published his enzyme electrode work, George Guilbault published a description of a potentiometric urea electrode using immobilized urease and a pH-sensitive sensor [46]. In 1982, Steckhan introduced electrochemical methods

12

F. Harnisch and D. Holtmann

to regenerate co-factors for enzyme catalysis to produce fine chemicals [47]. After this successful start, the field of electroenzymatic synthesis flourished [23]. In summary, electrobiotechnology is not a totally new field – it might only be that its time has come. Therefore, we stand on the shoulders of the mentioned pioneers and many more forefathers and foremothers in other scientific disciplines. Acknowledgements We are obliged to all authors and referees that contributed to this book, especially the members of our working groups who shouldered the load.

References 1. Roland Berger Strategy Consultants (2015) Chemicals 2035 – gearing up for growth how Europe’s chemical industry can gain traction in a tougher world. https://www.rolandberger. com/publications/publication_pdf/roland_berger_tab_chemicals_2035_20150521.pdf 2. Gulliver JS (2012) Transport and fate of chemicals in the environment - selected entries from the encyclopedia of sustainability science and technology. Springer, New York 3. Sillanpa¨a¨ M, Ncibi C (2017) A sustainable bioeconomy - the green industrial revolution. Springer 4. Sydow A et al (2014) Electroactive bacteria—molecular mechanisms and genetic tools. Appl Microbiol Biotechnol 98(20):8481–8495 5. Krieg T et al (2014) Reactor concepts for bioelectrochemical syntheses and energy conversion. Trends Biotechnol 32(12):645–655 6. Schr€ oder U, Harnisch F, Angenent LT (2015) Microbial electrochemistry and technology: terminology and classification. Energy Environ Sci 8(2):513–519 7. Logan BE (2008) Microbial fuel cells. Wiley, Hoboken 8. Scott K, Yu EH (2015) Microbial electrochemical and fuel cells. Woodhead Publishing, Sawston 9. Rasmussen M, Abdellaoui S, Minteer SD (2016) Enzymatic biofuel cells: 30 years of critical advancements. Biosens Bioelectron 76:91–102 10. Minteer SD, Liaw BY, Cooney MJ (2007) Enzyme-based biofuel cells. Curr Opin Biotechnol 18(3):228–234 11. Hiegemann H et al (2016) An integrated 45L pilot microbial fuel cell system at a full-scale wastewater treatment plant. Bioresour Technol 218:115–122 12. Logan BE (2010) Scaling up microbial fuel cells and other bioelectrochemical systems. Appl Microbiol Biotechnol 85(6):1665–1671 13. Mu Y et al (2011) Dehalogenation of iodinated X-ray contrast media in a bioelectrochemical system. Environ Sci Technol 45(2):782–788 14. Wang A-J et al (2011) Efficient reduction of nitrobenzene to aniline with a biocatalyzed cathode. Environ Sci Technol 45(23):10186–10193 15. Pous N et al (2015) Monitoring and engineering reactor microbiomes of denitrifying bioelectrochemical systems. RSC Adv 5(84):68326–68333 16. Lu L et al (2014) Microbial metabolism and community structure in response to bioelectrochemically enhanced remediation of petroleum hydrocarbon-contaminated soil. Environ Sci Technol 48(7):4021–4029 17. Gregory KB, Lovley DR (2005) Remediation and recovery of uranium from contaminated subsurface environments with electrodes. Environ Sci Technol 39(22):8943–8947 18. Kristiawan M (2017) Integration of basic knowledge models for the simulation of cereal foods processing and properties. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_10

Electrification of Biotechnology: Status quo

13

19. Cao X et al (2009) A new method for water desalination using microbial desalination cells. Environ Sci Technol 43(18):7148–7152 20. Turner AP (2013) Biosensors: sense and sensibility. Chem Soc Rev 42(8):3184–3196 21. Guo W et al (2017) Synergizing 13C metabolic flux analysis and metabolic engineering for biochemical production. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_2 22. Meng D-C, Chen G-Q (2017) Synthetic biology of polyhydroxyalkanoates (PHA). Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_3 23. Wagemann K, Tippk€otter N (2017) Biorefineries: a short introduction. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_4 24. De Tissera S et al (2017) Syngas biorefinery and syngas utilization. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_5 25. Rais D, Zibek S (2017) Biotechnological and biochemical utilization of lignin. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_6 26. Kosman J, Juskowiak B (2017) Bioanalytical application of peroxidase-mimicking DNAzymes: status and challenges. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/ 10_2017_7 27. Rouleau S et al (2017) RNA G-Quadruplexes as key motifs of the transcriptome. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_8 ¨ zilgen M (2017) How to decide on modeling details: risk and benefit assessment. Adv 28. O Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_9 29. Ahmad MH et al (2017) Fluorescence spectroscopy for the monitoring of food processes. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_11 30. Singh N, Herzer S (2017) Downstream processing technologies/capturing and final purification: opportunities for innovation, change, and improvement. A review of downstream processing developments in protein purification. Adv Biochem Eng Biotechnol. https://doi. org/10.1007/10_2017_12 31. Anastas PT, Warner JC (1998) Green chemistry: theory and practice. Oxford University Press, New York 32. Nevin KP et al (2010) Microbial electrosynthesis: feeding microbes electricity to convert carbon dioxide and water to multicarbon extracellular organic compounds. MBio 1(2): e00103–e00110 33. Ni Y, Holtmann D, Hollmann F (2014) How green is biocatalysis? To calculate is to know. ChemCatChem 6(4):930–943 34. Potter MC (1911) Electrical effects accompanying the decomposition of organic compounds. Proc R Soc Lond B 84:260–276 35. Schr€ oder U (2011) Discover the possibilities: microbial bioelectrochemical systems and the revival of a 100-year-old discovery. J Solid State Electrochem 15(7):1481–1486 36. Cohen B (1931) The bacterial culture as electrical half-cell. J Bacteriol 21:18–19 37. Canfield JH, Goldner BH (1964) Research on applied bioelectrochemistry. NASA Technical Report Magna Corporation, Anaheim, p 127 38. Ellis GE, Sweeny EE (1963) Biochemical fuel cells. In: NASA Technical Report The Marquardt Corporation 39. Wilkinson S, Campbell C (1996) Green bug robots - renewable environmental power for miniature robots. In: Proceedings of 4th IASTED international conference, robotics and manufacturing, Honolulu 40. Kim B et al (2001) A biofuel cell using wastewater and active sludge for wastewater treatment. International patent: WO0104061 41. Hongo M, Iwahara M (1979) Application of electro-energizing method to l-glutamic acid fermentation. Agric Biol Chem 43(10):2075–2081 42. Ghosh B, Zeikus J (1987) Electroenergization for control of hydrogen transformation in acetone butanol fermentations. In: Abstracts of papers of the American Chemical Society. Amer Chemical Soc 1155 16TH St, NW, Washington, DC 20036

14

F. Harnisch and D. Holtmann

43. Emde R, Schink B (1990) Enhanced propionate formation by Propionibacterium freudenreichii subsp. freudenreichii in a three-electrode amperometric culture system. Appl Environ Microbiol 56(9):2771–2776 44. Villano M et al (2010) Bioelectrochemical reduction of CO2 to CH4 via direct and indirect extracellular electron transfer by a hydrogenophilic methanogenic culture. Bioresour Technol 101(9):3085–3090 45. Heineman WR, Jensen WB (2006) Leland C. Clark Jr. (1918–2005). Biosens Bioelectron 21 (8):1403–1404 46. Guilbault GG, Montalvo Jr JG (1969) Urea-specific enzyme electrode. J Am Chem Soc 91 (8):2164–2165 47. Wienkamp R, Steckhan E (1982) Indirect electrochemical regeneration of NADH by a Bipyridinerhodium (I) complex as electron-transfer agent. Angew Chem Int Ed 21 (10):782–783

Adv Biochem Eng Biotechnol (2019) 167: 15–38 DOI: 10.1007/10_2017_34 © Springer International Publishing AG 2017 Published online: 26 October 2017

Extracellular Electron Transfer and Biosensors Francesca Simonte, Gunnar Sturm, Johannes Gescher, and Katrin Sturm-Richter

Abstract This chapter summarizes in the beginning our current understanding of extracellular electron transport processes in organisms belonging to the genera Shewanella and Geobacter. Organisms belonging to these genera developed strategies to transport respiratory electrons to the cell surface that are defined by modules of which some seem to be rather unique for one or the other genus while others are similar. We use this overview regarding our current knowledge of extracellular electron transfer to explain the physiological interaction of microorganisms in direct interspecies electron transfer, a process in which one organism basically comprises the electron acceptor for another microbe and that depends also on extended electron transport chains. This analysis of mechanisms for the transport of respiratory electrons to insoluble electron acceptors ends with an overview of questions that remain so far unanswered. Moreover, we use the description of the biochemistry of extracellular electron transport to explain the fundamentals of biosensors based on this process and give an overview regarding their status of development and applicability.

F. Simonte, G. Sturm, and K. Sturm-Richter (*) Department of Applied Biology, Institute for Applied Biosciences, Karlsruhe Institute of Technology, Fritz-Haber Weg 2, 76131 Karlsruhe, Germany e-mail: [email protected] J. Gescher Department of Applied Biology, Institute for Applied Biosciences, Karlsruhe Institute of Technology, Fritz-Haber Weg 2, 76131 Karlsruhe, Germany Department of Microbiology of Natural and Technical Interfaces, Institute of Functional Interfaces, Karlsruhe Institute of Technology, Eggenstein-Leopoldshafen, Germany

16

F. Simonte et al.

Graphical Abstract

e-

Signal

e-

An aly te s

e-

id ox

d ize

e-

es

tl ut

Sh

ed uc

d re

oxidized Shuttles

e-

reduced

Keywords c-type cytochromes, Direct interspecies electron transfer (DIET), Electron transfer network, Extended respiratory chain, Geobacter, Microbial fuel cell, Shewanella Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 The Geobacter Solution: Direct Electron Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Pili and the Conductivity of Geobacter Biofilms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 The Shewanella Electron Transfer Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 What We Know About Extracellular Electron Transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Biosensors: Application of Microbe–Electrode Interactions as Sensor Systems . . . . . . . . . . . 5.1 Nonaxenic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Axenic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Future Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

16 18 21 22 25 26 27 30 32 32

1 Introduction The interaction of microorganisms with electrodes as their electron acceptor is based on one of the oldest respiratory processes on earth. Ferric iron was and still is a very abundant potential electron acceptor [1]. In fact, it is the fourth most abundant element in the earth’s crust. Nevertheless, the dissimilatory (respiratory)

Extracellular Electron Transfer and Biosensors

17

reduction of iron necessitates certain adaptations by microorganisms because iron occurs in soil sediments in the form of insoluble minerals such as ferrihydrite and hematite. Hence, the reduction of ferric iron is dependent on the evolution of an extended respiratory chain to the cell surface. Usually, the activity of enzymes is highly specific and follows the “lock and key” principle. In the case of ferric iron respiration, evolution has selected for rather unspecific terminal reductases that can reduce a broad range of solid-phase electron acceptors if they have a redox potential that sustains the electron transfer process. This ability is necessary because environmental ferric iron minerals have an uncountable number of different shapes, which makes a specific lock–key interaction unsuitable. The redox potential of the electron acceptor determines the electron transfer rate. Using electrodes as electron acceptors offers the ability to steer the respiratory rate and thereby the overall metabolic activity of electrode-attached organisms. Prokhorova et al. conducted experiments with a mixed-species biofilm of Geobacter sulfurreducens, Geobacter metallireducens, and Shewanella oneidensis grown on an activated carbon electrode [2]. With a sweep voltammetry experiment, during which the potential of the anode was slowly decreased and thereafter increased again, the authors revealed that electron transfer by the organisms was possible down to a redox potential of 200 mV against a normal hydrogen electrode (NHE). Half-maximal electron transfer rates were achieved at 80 mV; in other words, the organisms respired half as fast as at this potential. In the model genera studied so far, dissimilatory reduction of iron and electrodes seems to be dependent on the activity of c-type cytochromes [3]. These proteins harbor one or more heme cofactors in a prosthetic group that is covalently attached to the protein backbone. The central coordinated Fe atom within the heme groups can switch between oxidized Fe(III) and reduced Fe(II) states, thereby enabling electron transfer. Using a bioinformatic approach, Sturm and colleagues tried to correlate the number of c-type cytochrome-encoding genes in the genomes of organisms from the proteobacteria phylum with the ability to reduce ferric iron or an electrode [4]. Although the average number of c-type cytochrome-encoding genes is 13, known model organisms for iron or electrode reduction such as S. oneidensis (41) or G. sulfurreducens (103) contain a multiple of this value. Hence, within certain boundaries, it is possible to conclude from the number of ctype cytochrome encoding-genes whether the genome belongs to an organism capable of thriving on minerals and electrodes. Using a biofilm on an electrode surface, Firer-Sherwood and colleagues revealed a fundamental aspect of c-type cytochrome-based extracellular electron transfer [5]. The biofilm was composed of single purified c-type cytochromes that are thought to play a role within the electron transport chain to the cell surface in S. oneidensis. These individual cytochromes form monolayers on the surface of certain electrodes. Using protein film voltammetry experiments, the authors could identify at which potential the redox-active proteins were able to accept and deliver electrons from or to the electrode. Interestingly, the experiments revealed that all analyzed cytochromes operated in overlapping redox potential windows. This allows fast transfer of respiratory electrons from the cytoplasmic membrane to

18

F. Simonte et al.

the terminal electron acceptor. Subsequently, it was shown that these overlapping redox-potential windows not only allow fast unidirectional electron transfer to the cell surface, but are also the basis for a dynamic network of interacting electron transport chains in S. oneidensis (see Sect. 3). Moreover, at the scale of microorganisms within a redox-active biofilm, these proteins allow electron transfer over hundreds of micrometers within conductive biofilms or the interspecies transfer of electrons in syntrophic communities in which organisms share their electron donors. Hence, the unspecificity of electron transfer is key to a multitude of anoxic redox reactions that are the foundation for applications in bioelectrochemical systems, but also shape whole ecosystems and determine their productivity. In this chapter, we assemble knowledge regarding electron transfer in model organisms and in new species that can interact with ferric iron or anode surfaces. We highlight the two common routes of electron transfer, which occur either via direct contact between the electron acceptor and redox-active enzymes on the cell surface or via the use of endogenously produced shuttling substances [6]. In recent years, it has been shown that electron transfer chains can also operate in the reverse direction if a suitable electron donor (e.g., a cathode or the electron chain of another organism) is available. We focus on parts that are necessary for reverse electron transfer. Moreover, we use this assembly of knowledge on extracellular electron transfer as the foundation for proposed application of these reactions for biosensors.

2 The Geobacter Solution: Direct Electron Transfer The Geobacteraceae family is one of the best-studied groups of microorganisms capable of extracellular electron exchange reactions. Electron transfer to minerals and electrodes necessitates direct contact between the cells and the electron acceptor because the microorganisms do not produce an endogenous mediator for indirect/mediated electron transfer [6, 7]. Because cell culturing and genetic manipulation of G. sulfurreducens is well established, the organism is the model bacterium for the study of extracellular respiration among the Geobacteraceae. Several studies using G. sulfurreducens have revealed some of the factors that play a role in direct extracellular electron transfer. As mentioned before, c-type cytochromes are the key players in this process [8–11], as can be judged by their abundance and diversity [9, 12]. The genome of G. sulfurreducens encodes for more than 100 c-type cytochromes. Among these, 65 homologs have been identified within the genome of the closely related strain G. metallireducens [13]. A well-conserved c-type cytochrome in Geobacter species is PpcA. This triheme periplasmic cytochrome [9] is localized to the periplasm of the cell and predicted to transfer electrons from the cytoplasmic membrane to the outer membrane by heme groups that are oriented in parallel or perpendicular to each other [14]. For the inner membrane, there is evidence for at least two transfer pathways of electrons into the periplasm. The multiheme c-type cytochrome ImcH is required for respiration to extracellular electron acceptors with high redox potentials (above 0.1 V versus

Extracellular Electron Transfer and Biosensors

19

the standard hydrogen electrode, SHE [15]), whereas CbcL, an inner membrane protein consisting of b- and c-type cytochrome domains, is chosen for the reduction of acceptors with redox potentials lower than 0.1 V [16]. MacA is a cytochrome that is associated with the inner membrane and exhibits hydrogen peroxide reductase activity. It can transfer electrons to PpcA [17]. The precise role of MacA is under debate, as the expression of macA is upregulated only under Mn(IV) oxide reduction conditions; however, a knockout of macA results in impaired growth on soluble and insoluble Fe(III), but not on Mn(IV) oxide [8]. Investigation of the expression of omcB (an outer membrane cytochrome gene) in a growth experiment with ferric citrate revealed that omcB transcript and protein levels were reduced in a macA mutant. This suggests that MacA and OmcB may function in the same or similar pathways of electron transfer, or that MacA might be a regulatory factor for omcB expression and not directly involved in electron transfer to the cell surface. OmcB is embedded in the outer membrane, with a part of the protein exposed to the cell surface [18]. Deletion of omcB leads to impaired reduction of both soluble and insoluble iron species [19]. Recently, the mechanism by which G. sulfurreducens transfers electrons across the outer membrane was elucidated in more detail and a striking similarity to the S. oneidensis mtr gene cluster was identified [20, 21] (see Sect. 3). In G. sulfurreducens, two homologous gene clusters encode for a trans-outer membrane porin cytochrome complex (Pcc), consisting of a periplasmic c-type cytochrome (OmaB/C), a porin-like protein (OmbB/C), and a reductase (OmcB/C) in the outer membrane. It was shown by Liu and colleagues that the Pcc protein complex is able to transfer electrons across a liposomal membrane and reduce ferric citrate and ferrihydrite similarly to the MtrABC complex in S. oneidensis [22, 23]. However, although function and localization of the Pcc complex are very similar to the outer membrane complex in S. oneidensis, the amino acid sequence differs considerably, which indicates independent evolution of (at least) two ways of transferring electrons across the outer membrane [20, 22]. As mentioned, OmcB is an important protein in the outer membrane, but it is just one of many outer membrane and extracellular cytochromes known to be involved in direct electron transfer from the cell to an extracellular electron acceptor. Several cell surface c-type cytochromes have been identified in G. sulfurreducens and extensively studied. In addition to OmcB, OmcE and OmcT are also localized to the outer membrane. When omcE is deleted, G. sulfurreducens can no longer reduce Fe(III) oxide but is still able to reduce soluble electron acceptors, including Fe(III) citrate. OmcT is predicted to encode a hexaheme outer membrane cytochrome with roughly 60% amino acid sequence identity to OmcS and also contributes to the reduction of iron oxide [11]. Another important and thoroughly studied c-type cytochrome involved in extracellular respiration is OmcS, a hexaheme c-type cytochrome [24] associated with the pili of G. sulfurreducens [25]. Like OmcE and OmcT, OmcS is required for growth on insoluble, but not soluble, iron species [11]. However, the deletion of OmcS has no effect on electron transfer to an electrode, indicating that there is a

20

F. Simonte et al.

different mechanism involved in electron transfer to anode surfaces [24]. OmcS is also essential for accepting electrons in direct interspecies electron transfer (DIET) within cocultures of G. sulfurreducens and G. metallireducens, a process that is discussed in more detail in Sect. 2.1 [26]. OmcZ is another c-type cytochrome and is necessary for respiration to electrodes but not to insoluble Fe(III) oxides [27]. Although OmcZ is essential for highdensity current production, it is not important for cells grown under currentconsuming (i.e., electron accepting) conditions. The omcZ gene showed lower transcript abundance in current-consuming compared with current-producing biofilms [28]. Furthermore, deletion of omcZ had no influence on electron transfer from the electrode to G. sulfurreducens cells [28]. These results suggest that the mechanism(s) for reverse electron transfer from the electrodes to G. sulfurreducens cells might be different from the mechanism of electron transfer from cells to electrodes. As mentioned, DIET is another process whereby a cell transfers electrons to an external acceptor other than iron or to an anode. In this syntrophic relationship, one microorganism receives electrons from a neighboring species via direct electrical connections, without the involvement of electron carriers such as hydrogen or formate [29]. DIET was first observed in cocultures of G. metallireducens and G. sulfurreducens [26], but was also demonstrated for aggregates of G. metallireducens and methanogens such as Methanosarcina barkeri and Methanosaeta harundinacea [30, 31]. The Geobacter cocultures were mainly studied in media with ethanol as electron donor and fumarate as electron acceptor. G. metallireducens is capable of metabolizing ethanol but cannot use fumarate as an electron acceptor, whereas G. sulfurreducens cannot metabolize ethanol but can reduce fumarate. Recent DIET studies with G. metallireducens and a G. sulfurreducens strain that is unable to metabolize acetate showed cell growth and metabolism of ethanol similar to those found for a wild-type coculture, indicating that the electrons transferred via DIET are sufficient to support growth of G. sulfurreducens. Transcriptomic data demonstrate low expression of genes associated with hydrogen and formate uptake, giving evidence that it is unlikely that these molecules are the interspecies electron transfer carriers. The cocultures form electrically conductive biofilm aggregates that are macroscopically visible. Further studies showed that OmcS is essential for direct electron exchange by G. sulfurreducens in DIET cocultures. Additionally, the pilA gene seems to be involved in effective DIET, but it is still not clear whether the nanowires (see Sect. 2.1) participate in direct exchange of electrons or are important for attachment of the cells during aggregate formation [26]. Also, the full diversity of organisms that can participate in DIET is not yet known. For example, Pelobacter carbinolicus, a close relative of G. metallireducens, is incapable of DIET but dependent on hydrogen or formate as interspecies electron carrier [32]. In addition to the importance of cytochromes (and pili, see Sect. 2.1) for extracellular electron transfer in Geobacter, a role for extracellular polysaccharides as attachment sites for peripheral redox proteins was demonstrated by Rollefson et al. [33]. Deletion mutants in the extracellular anchoring polysaccharide (xap)

Extracellular Electron Transfer and Biosensors

21

gene cluster did not show any residual electron transfer activity on electrodes and were unable to reduce ferric iron or restore this capability over time. This indicates that sugar molecules might be important as extracellular anchors for c-type cytochromes in extracellular electron transfer.

2.1

Pili and the Conductivity of Geobacter Biofilms

In addition to cytochromes and other proteins, the electrically conductive type IV pili are known to play an important role in effective extracellular respiration by Geobacter species [34, 35]. Other microorganisms use type IV pili for twitching motility. However, Geobacter pili were thought to be biological filaments with a metallic-like conductivity attributed to overlapping π–π orbitals of aromatic amino acids, comparable to synthetic conductive polymers [34, 35]. Nevertheless, the hypothesis of metallic-like conductivity has been steadily questioned. For example, Yates and colleagues recently reported that the conductivity of living Geobacter cells is based on redox conductivity and that the conclusions of Malvankar and colleagues were based on a misleading experimental setup [36]. Malvankar and colleagues later questioned whether Yates and colleagues were using biological material similar to the samples used for their 2011 study [37]. The pili described as “microbial nanowires” were initially discovered in G. metallireducens biofilms grown on insoluble iron oxides, but not when grown with soluble or chelated Fe(III) as electron acceptors [38]. In G. sulfurreducens, the pilus fibers are covered with the cytochrome OmcS, but the spacing between individual cytochromes is too far for electron transfer along the pili (100–200 nm distance on the nanowires versus 1–2 nm spacing required for cytochrome-tocytochrome electron hopping). It was therefore hypothesized that the electrons are transported via the pili to OmcS proteins, which catalyze the final reduction of ferric iron [39, 40]. However, it can be questioned whether the antibody-based detection of OmcS leads to complete identification of all proteins along the fibers and how strong the interaction of OmcS with the pili has to be in order to prohibit loss of the outer membrane cytochrome during sample generation. The main structural protein of the pili is encoded by pilA and deletion of this gene inhibits iron oxide reduction and production of conductive biofilms by G. sulfurreducens [35] and G. metallireducens [41]. As mentioned before, pili are also essential for DIET in both organisms [26, 42]. The PilA protein is the structural component of the nanowires. The N-terminus of the PilA sequence of G. sulfurreducens and other organisms is well conserved, whereas the C-terminus of G. sulfurreducens PilA differs from that of other species such as Pseudomonas, Vibrio, or Neisseria and is atypically rich in aromatic amino acids. It was hypothesized by Malvankar and colleagues that the tight packing of aromatic amino acids might be the basis for the hypothesized metallic-like conductivity. Even if the conductivity is redox-based, these aromatic amino acids seem to be necessary for electron transfer [34]. Interestingly, the expression of

22

F. Simonte et al.

Pseudomonas aeruginosa PilA in G. sulfurreducens results in poorly conductive pili, although OmcS is localized properly [20]. Substituting the five most distal aromatic amino acids with alanine leads to the production of pili with properly localized OmcS, but the conductivity of filament preparations is greatly diminished and cells of the mutant strain are no longer able to reduce anodes or insoluble Fe (III) oxides. These results demonstrate that filaments decorated with OmcS are not sufficient for effective electron transfer and that the pili themselves have to be conductive [43]. Substitution of phenylalanine and tyrosine residues at the C-terminus of the PilA protein with tryptophan, an amino acid that promotes electron transport more efficiently, increased the conductivity of the nanowires 2000-fold. Interestingly, the diameter of the filaments was halved to 1.5 nm, probably because of the more hydrophobic surface structure [44]. In applications such as bioelectrochemical systems, nanowires are an efficient strategy for promoting current production because biofilm conductivity permits electron transfer to an anode at distances up to 400 μm. It was demonstrated that G. sulfurreducens biofilms are more conductive if the strain produces more pili. For instance, strain KN400 is a rare variant of the Geobacter strain DL-1 that was selected for high current production; it produced biofilms that were five times more conductive and had a much higher abundance of PilA [34, 45]. Conductive pili are not only of interest for microbial electron transfer reactions, but also as “natural” electronic materials that have advantages compared with chemically produced fibers. Microbial nanowires are cheap, nontoxic, selfrenewing, and highly chemically stable. This allows new research and development directions regarding material development. Also, the strong dependence of conductivity on pH could be applied in devices such as pH biosensors.

3 The Shewanella Electron Transfer Solution Organisms from the genus Shewanella within the γ-proteobacteria can also serve as model organisms for the investigation of bacterial extracellular electron transfer (EET). It seems that these organisms, compared to Geobacteraceae, have developed different solutions for the same problem of transferring electrons to insoluble electron acceptors. The best studied organism from the genus is S. oneidensis MR-1, which shows an astonishing versatility of terminal electron acceptors (TEAs) that can be used by the organism in anoxic environments. The organism is capable of using organic compounds (fumarate, TMAO, DMSO, humic acids, and others) as well as inorganic compounds [NO3 , NO2 , Mn(VI), Fe(III)] as TEAs during anaerobic respiration [46–48]. Sturm et al. developed an electron transfer network hypothesis in which the electron transfer chains to most of these electron acceptors are interwoven by a central electron transfer hub consisting of the two c-type cytochromes STC (small tetraheme cytochrome) and FccA (Fig. 1) [49]. The starting point of most of these electron transport pathways is CymA (cytoplasmic

Extracellular Electron Transfer and Biosensors

23

Fig. 1 Periplasmic electron transfer reactions. Proposed electron transfer routes for CymAderived electrons to different periplasmic and outer membrane terminal reductases. Diamonds indicate the number of heme groups of the single c-type cytochromes involved

membrane protein A). This menaquinol oxidase is a 16.5 kDa tetraheme protein that is anchored to the cytoplasmic membrane and faces the periplasm. The crucial role of CymA within extracellular electron transfer reactions was discovered early in S. oneidensis research by Myers and colleagues in 1997. They were able to show the integral role of CymA by studying deletion mutants within the corresponding gene that were unable to use a variety of anaerobic electron acceptors, including ferric iron and manganese oxides [50]. Current experimental evidence allows the formulation of the following model regarding electron transfer to the cell surface: Electrons derived from cellular metabolism enter the quinone pool and are transferred to CymA. The tetraheme STC, the flavotetraheme, and catabolic fumarate reductase FccA function as a

24

F. Simonte et al.

CymA oxidase, receive electrons, and transfer them to the decaheme cytochrome MtrA. MtrA, the periplasmic part of the outer membrane Mtr complex, transfers the electrons to the cell surface-localized decaheme c-type cytochrome MtrC, which functions as the terminal reductase. The interaction of the periplasmic component MtrA and the outer membrane protein MtrC is possible via MtrB. This β-barrel protein probably forms a pore in the outer membrane that allows MtrA and MtrC to come into close contact to facilitate electron transfer. MtrC, as the last component of the electron transport chain, transfers the electrons to the terminal electron acceptor Fe3+. The same pathway is also used for the reduction of an electrode. The extracellular electron transport chains of S. oneidensis are connected to a variety of periplasmic routes leading to periplasmic reduction targets such as NO3 , NO2 , and fumarate. S. oneidensis also expresses a multitude of additional cytochromes during anaerobic respiration [49, 51, 52]. The function of most of these cytochromes is not yet known, but it is possible to estimate a metabolic or evolutionary benefit for this behavior. These cytochromes can operate as a capacitor, which enables S. oneidensis to store electrons even in the absence of a suitable electron acceptor. In natural habitats, the local concentration of electron acceptors can be low. A “charged” cell would be able to load its electron cargo onto acceptors as soon as they are close enough to allow electron transfer. Moreover, the cytochrome-based periplasmic capacitor interconnects the individual respiratory chains, leading to a phenomenon called the electron transfer hub. Using this hub, the charged cell has the ability to transfer electrons either simultaneously or consecutively to distinct reduction targets. Of note, the heme content of a single S. oneidensis MR-1 cell allows about 700,000 electrons to be stored [49]. Details of the underlying biochemical behavior of the proteins involved are still under debate. Evidence for static formation of protein/protein complexes between CymA and FccA was obtained from experiments using a quartz crystal microbalance with dissipation (QCM-D) [53]. By contrast, diffusion-based electron shuttling behavior was suggested by NMR-based biochemical characterization of STC, a key component in diverse respiratory chains in S. oneidensis [54]. The electron transport routes described so far are the basis for direct electron transfer from cellular bound proteins to extracellular terminal electron acceptors. This process requires local proximity. In addition to this mechanism, S. oneidensis is able to transfer electrons over a certain distance using endogenous shuttles. These shuttles are flavin-based molecules that are produced intracellularly and then secreted to the medium. Depending on the strain, the amount of secreted flavins could reach micromolar concentrations [55–57]. These shuttles are capable of taking up electrons and diffusing through the medium until reaching a suitable electron acceptor to discharge their cargo onto. Figure 2 shows an idealized illustration of the influence of secreted shuttles in a media-exchange experiment. The process of electron uptake and discharge can be repeated several times by diffusing between the electron donor (the cell) and the electron acceptor. This behavior and the reduction potential of these redox cycles can be measured via cyclic voltammetry (CV) [59, 60]. CV is a common tool in the field of bioelectrochemistry and is used intensely for determining redox properties of

Extracellular Electron Transfer and Biosensors

25

Fig. 2 Idealized illustration of a media exchange experiment adopted from Brutinel and Gralnick [58]. The scheme shows the influence of flavins secreted to the medium with respect to chronoamperiometric behavior. The already secreted flavins in re-used cell-free medium facilitate rapid current production by the cultured cells, whereas the cells first need to produce new flavins if fresh medium is added to the experiment

redox-active molecules such as c-type cytochromes [5, 60–62]. Results from these works show the influence of the presence of flavin-based shuttles with respect to the redox properties of the electron transfer abilities of whole cell systems. Recent studies revealed a possible secondary (or primary) function of the secreted flavins. These molecules bind to outer membrane cytochromes (OMCs) after secretion and support extracellular electron transfer to iron oxides or electrodes. Endogenously produced flavins bind to OmcA, and riboflavin (RF) and flavin mononucleotide (FMN) bind to MtrC, establishing a one-electron reaction based on the formation of a semiquinone-like state that facilitates electron transfer between reduced cytochrome-bound hemes and the electron acceptor [63–66].

4 What We Know About Extracellular Electron Transfer The previous sections highlighted decades of research with Shewanella and Geobacter strains. Even though many research groups all over the world work with these strains and their deletion mutants, there are still several open questions. One question is whether the function of flavin in extracellular electron transfer in S. oneidensis is as an electron shuttle or as a cofactor for outer membrane cytochromes. Moreover, it is still debated whether there is a means by which S. oneidensis can direct electron transfer reactions within the periplasm. Another highly discussed but unresolved issue is how long-range electron transfer is conducted by Geobacter cells and within Geobacter biofilms. One opinion is that electron transfer is carried out via parts of Geobacter cells that have metallic-like conductivity, but the opposite opinion is that electron transfer is based on electron hopping between discrete redox sites [67].

26

F. Simonte et al.

These questions are important, but appear rather small if one considers how many microorganisms seem to be capable of electron transfer to solid-phase electron acceptors and how little we know about how the electron transfer is conducted. Koch and Harnisch recently published a metaresearch article with a list of organisms for which interaction with an electrode was detected [68]. For many of these organisms, microbe–electrode transfer is still described on the phenomenological level, without any idea of the biochemical mechanism. Grampositive organisms are of interest here because they must function differently to Gram-negative model organisms. Some groups have tried to elucidate the mechanism by which Thermincola and Clostridia species might be able to conduct electron transfer. For Thermincola, evidence indicates that c-type cytochromes might be the key players [69]. Nevertheless, how the thick murein layer is conductively bridged is not known. Interestingly, two Thermincola strains are currently used by research groups. Although T. potens produces monolayer biofilms on anodes [70], T. ferriacetica produced biofilms several micrometers thick that generated higher current densities than G. sulfurreducens [71]. Future research will hopefully elucidate whether electron transfer conducted by other microorganisms is a variation of known pathways or whether evolution has solved the same problem via different approaches that depend on the cellular morphology of the biocatalyst or the environmental conditions in which the organisms conduct this process.

5 Biosensors: Application of Microbe–Electrode Interactions as Sensor Systems One application of the biochemical reactions discussed above is in biosensors based on extracellular electron transport reactions. The field of biosensor development is gaining more and more momentum. The trend for the development of biosensor systems started with the invention of Clark in 1962 [72]. He presented the first biosensor, which incorporated immobilized enzymes on an electrode, forming the so-called enzyme electrodes. That sensor was the basis of numerous variations, with the use of many other immobilized (oxidase) enzymes. The principle of the enzyme electrode is still used and commercialized as glucose sensors for the measurement of glucose concentration in blood. The need for rapid measurements has increased drastically, especially for medical applications but also in many other fields such as pollution monitoring, which has led to fast development of diverse biosensors [73, 74]. A biosensor consists of a biological recognition element that is associated with or integrated within a transducer that provides an indication or signal, which can be optical, electrochemical, thermometric, piezoelectric, magnetic, or micromechanical [75–77]. Electrochemical sensors can be subdivided further based on the change in the measured voltage between the electrodes

Extracellular Electron Transfer and Biosensors

27

(potentiometric), the change in the measured current at a given applied voltage (amperometric), or the change in the ability of the sensing material to transport charge (conductometric). A modern modification of these bioelectrochemical systems (BES) are microbial fuel cell-based biosensors. Because of their versatility in construction and application modes, this section focuses on BES-based biosensors. Selected applications, differentiated by nonaxenic and axenic setups, are presented.

5.1 5.1.1

Nonaxenic Systems Biological Oxygen Demand

The application of microorganisms as the sensing or recognition element allows the detection of bioactive substances or the toxicity of environmental pollutants. Thus, microbial biosensors have been mainly developed for environmental monitoring applications [78]. The most popular application of BES-based biosensors is the determination of the biological oxygen demand (BOD) of wastewater, effluents, and polluted waters that contain mixtures of reduced organic carbon sources [79, 80]. The conventional methods for BOD analysis usually require 5 or 7 days of incubation at 20  1 C in the dark [81, 82] and technicians with experience and skill to achieve reproducible results. In comparison, BES-based biosensors operate with drastically reduced analysis times (minutes to hours). Furthermore, the new techniques possess long-term stability and high reusability rate, which makes them suitable for field online monitoring applications [83, 84]. Among these BES-based biosensors, amperometry is the most commonly used measuring method. The used transducer is mainly a dissolved oxygen (DO) probe consisting of an Ag/AgCl anode and a gold or platinum cathode covered with an oxygen-permeable membrane [85–87]. Karube et al. first described a BES-based BOD sensor utilizing isolated microorganisms from soil [86]. The aerobic microorganisms are immobilized with collagen fibers as the recognition element between the porous membrane and the gas-permeable membrane of the DO electrode (see Fig. 3a). DO is consumed by the immobilized cells for oxidation of the organic compounds in the sample solution. The remaining oxygen diffuses through the gas-permeable membrane and reduces at the cathode of the DO electrode. As a result, the current of the electrode decreases with time until a steady state is reached. The steady state indicates equilibrium between the bacterial consumption rate and the diffusion of oxygen from solution [86, 88]. To compare BOD values obtained using the conventional method and the BOD biosensor, a proper calibration solution needs to be selected. For the conventional BOD method, a solution containing a mixture of glucose and glutamic acid (GGA) corresponding to a concentration of 205 mg/L BOD is used. The GGA standard is not always adequate for the calibration of microbial BOD sensors because the composition of real wastewater is complex and the components mostly have low oxidation rates. Thus, defined synthetic wastewaters

28

F. Simonte et al.

Fig. 3 (a) Clark dissolved oxygen probe with immobilized microorganisms. (b) Microbial fuel cell setup with direct or mediated electron transfer

recommended by the Organization for Economic Cooperation (OECD) are often used [87]. It is also possible to develop calibration solutions based on the main components found in the analyzed water samples. There are two main measuring methods for correlating the signal with the BOD in the sample solution for BES-based BOD sensors. One approach is the steadystate method, which uses the current difference between the steady states prior to and after sample addition. The other method, referred to as initial rate method or kinetic method, is based on the initial rate of current change directly after sample addition. This measuring method is less time-consuming than the steady state method [87, 89]. Since 1977, various studies have been presented based on the principle of monitoring DO consumption using a Clark DO electrode with immobilized microorganisms [90–93]. However, BOD sensors that use a membrane and a DO probe are subject to disadvantages such as membrane fouling, short-term stability, calibration drift, and dependency on oxygen saturation [94]. These disadvantages can be overcome by the use of extracellular electron transfer in bioelectrochemical systems such as microbial fuel cells (MFC). These systems consist of an anoxic compartment with an anode (negative electrode) and an oxic compartment with a cathode (positive electrode), separated by a proton exchange membrane (see Fig. 3b). In the anoxic compartment, microorganisms degrade organic matter and generate electrons and protons [95, 96]. The respiratory electrons are transported onto the anode surface, either directly via an extended

Extracellular Electron Transfer and Biosensors

29

electron transport chain or mediated by extracellular mediators, and measured electrochemically. The use of mediators such as ferricyanide leads to measurements almost independent of the DO in the medium and results in generation of higher currents than for DO electrode-based biosensors [80, 97–100] . When the mediator is given in excess, errors occurring through dilution of the samples can be further reduced. Another advantage of the mediated MFC-based BOD sensor, compared to DO probe sensors, is that it allows an increase in the concentration of microorganisms without rapid depletion of the electron acceptor (e.g., oxygen) and, consequently, a decrease in the time required to degrade significant amounts of organic material [101]. Accordingly, many MFC-based mediator-less BOD measurement methods have been developed. Kim et al. [84] showed that an enriched microbial consortium of electrochemically active organisms could be used within an MFC to treat wastewater but also as a BOD sensor. The BOD values could be either determined by reading the maximum current or by calculating the coulombic response [84, 102]. Furthermore, Chang et al. [83] used MFCs for BOD analysis within the linear range of organic contaminants up to 100 mg/mL. Moreover, the operational stability of the MFC-based BOD sensors could be maintained for at least 5 years compared to the 7–140 days reported for DO electrode BOD biosensors [83, 103]. Nevertheless, the response time still varied from 1 h to several hours. Kumlangham et al. [104] proposed a novel design for their MFC-based BOD sensor. A continuous anaerobic bioreactor was integrated into the system to supply a fresh and stable anaerobic consortium to the anode compartment after each sample measurement. Thereby, the metabolic recovery time was avoided and the measurement time could be significantly reduced to 3–5 min. The linearity of MFC sensor response to glucose standard solutions ranged from 1 to 25 g/L and the limit of detection was 25 mg/L [104]. Further improvements regarding simplicity and cost reduction were achieved by Di Lorenzo et al., who presented a single-chamber MFC with an air cathode that was independent of aeration, recycling, or chemical regeneration of the catholyte [105]. The biosensor output of artificial wastewater had a linear relationship to BOD concentrations up to 350 mg/L. The measurements were characterized by high reproducibility and stability over 7 months. The measuring time corresponded to a hydraulic retention time (HRT) of around 40 min [105].

5.1.2

Toxins

In addition to their application for determination of biologically degradable organic matter, BES-based sensors can be used to determine biologically active toxins and pollutants in waters. Provided the MFC works at saturated feed concentrations and parameters such as pH, salinity, temperature, and anode potential are kept constant, any variation in the current output can be attributed to the presence of toxic contamination in the feed stream. In other words, inhibition of the microorganisms by toxins and pollutants leads to decreased metabolic rates, which can be detected

30

F. Simonte et al.

as lower electron transfer rates [106–109]. Thus, the MFC can act as a biosensor for biologically active substances in water samples. The microbial toxicity biosensors fulfill the need for unspecific detection covering a wide range of water samples. This setup has many advantages compared with alternative toxicity sensor approaches using living organisms such as fish, protozoa, algae, bivalves, and daphnia [110–113]. These methods require long incubation times (up to several weeks) and have a low reproducibility and stability. Moreover, the self-sustainable nature of the MFCs enables long-term online monitoring and avoids the need for an external transducer and power source. The use of mixed bacterial communities allows the robust detection of a broad range of toxic pollutants [106, 114]. Kim et al. [107] used an MFC to study the toxicity of some organophosphorous compounds and heavy metals such as Pb(II) and Hg(II). The anode was previously colonized by electrochemically active microorganisms. Inhibition was seen at concentrations as low as 1 mg/L. This system could not discriminate between the different toxic substances but it was able to monitor contamination with toxic substances [107]. The work of Patil et al. [115] compared the toxic effects of several biocides [sulfamethaxozole and sulfadiazin, chloramine B, Cu(II), Ag(I), Pb(II), and Hg(II)] in MFC setups running either with electrochemically active biofilms or with planktonic cells. The planktonic cells were supplemented with anthraquinone-2-sulfonate (500 μM) as a redox mediator. The authors could show that the planktonic cells in the MFC setup were more affected by the toxins than the biofilm. According to these experiments, MFC-based biosensors with exoelectrogenic biofilms are less sensitive than MFCs with planktonic cells [115]. The first microfabricated toxicity sensor based on MFC technology was developed by Da´vila et al. in 2011 [116]. It features a working volume of 144 μL per compartment and detects the presence of toxic compounds by a remarkably low generation of electricity by electrically active bacteria, here G. sulfurreducens. In this study, the toxin tested was formaldehyde, which was added to the anode compartment stepwise to give final concentrations of 0.1–4%. In all cases, the sensor detected the presence of the toxin with a steep drop in current production. It has to be mentioned that all tested concentrations irreversibly inactivated the biofilm [116].

5.2

Axenic Systems

BES-based BOD sensors benefit from the unspecific mixture of diverse microorganisms to respire a broad range of carbon sources. In contrast, specific biosensors for single analytes often require the use of microorganisms with a specific substrate spectrum. The group of Tront presented laboratory MFC-based biosensors for acetate and lactate by implementing the microorganisms G. sulfurreducens und S. oneidensis MR-1, respectively [117, 118]. The biosensors presented by Tront and colleagues were suitable for detecting and quantifying the carbon source in the medium because the current production by the organisms was proportional to the

Extracellular Electron Transfer and Biosensors

31

concentrations of the respective carboxylic acids (0–2.3 mM for acetate and 0–41 mM for lactate). Holtmann and colleagues developed an MFC-based bioactivity sensor (BAS) to monitor microbial activities relevant for effective wastewater treatment. The BAS uses biomass directly from the treatment plant and measures signals that are proportional to substrate degradation. Simultaneously, the sensor can also provide information on inhibiting effects on the biomass [119]. In 2006, the group used a BAS to monitor the metabolic activity during cultivation of microorganisms. Measurement is based on the detection of an electron flow, induced by reduced metabolic intermediates or products during growth. The secreted reduced or electroactive products are capable of oxidation at the platinum anode, whereas oxygen is reduced to water at the cathode. Using the BAS, the group could demonstrate different activity signals depending on the different cultivated microorganisms and cultivation conditions [120]. A different approach was used by Golitsch et al. [121]. For the first time, genetically engineered bacteria were used as a very selective recognition element for an MFC-based biosensor. The group was able to control the expression of relevant proteins for extracellular electron transfer. The work was based on the previously engineered strain of S. oneidensis ΔOMC [122], which is deficient in the transport of electrons to extracellular electron acceptors or anodes, because all outer membrane cytochromes (OMCs) were deleted. The genes for the terminal reductase MtrF, the periplasmic cytochrome MtrA, and the integral outer membrane protein MtrB were set under the control of the arabinose inducible promoter PBAD. Thus, it was possible to steer extracellular electron transfer rates by modifying the inducer concentration. A linear proportionality of current production and inducer concentration between 0.1 and 1 mM was achieved [121]. The most prominent advantage of the strain S. oneidensis MtrFAB is its versatility, as it could be used as platform strain for the detection of various analytes just by replacing the promoter region. Antibiotics, toxins, sugars, short chain fatty acids, etc. could be used as possible analytes The genotype and phenotype of the engineered strain S. oneidensis MtrFAB are shown in Fig. 4. A similar approach was used by Webster and colleagues 1 year later in 2014 [123]. The authors described a genetically engineered biosensor for the detection of arsenic. The exoelectrogenic strain S. oneidensis ΔmtrB [55] was used as sensor strain and further modified by complementation with a plasmid encoding a copy of mtrB. The expression of mtrB was set under the control of an arsenic-inducible promoter (Pars) and its downstream transcriptional regulator (arsR). MtrB is essential for electrode reduction [55]. Consequently, a linear increase in mean peak current was observed over a range of 0–100 μM arsenite [123].

32

F. Simonte et al. e-

Fe(III)

Fe(II)

MtrF

MtrB outer membrane MtrA

e-

periplasm

S. oneidensis MtrFAB feoA

PBAD araC

mtrF

mtrA

mtrB

SO_1775

Fig. 4 Relevant phenotype and genotype of the biosensor strain S. oneidensis MtrFAB [121]

6 Future Challenges This review has described the broad range of applications of extracellular electron transfer in the field of biosensor development. Nevertheless, to the best of our knowledge, only one sensor system based on microbial electron transfer to electrodes has made it to the market (TOX/BOD System HABT-3000; http://korbi.en. ecplaza.net/). Apparently, the robustness, range, and specificity toward detectable analytes need to be increased and measuring times decreased. Moreover, it is necessary to build small and cheap bioelectrochemical systems that sustain the cost-efficient use of this technology.

References 1. Vargas M, Kashefi K, Blunt-Harris EL, Lovley DR (1998) Microbiological evidence for Fe (III) reduction on early earth. Nature 395:65–67 2. Prokhorova A, Sturm-Richter K, Doetsch A, Gescher J (2017) Resilience, dynamics and interactions within a multi-species exoelectrogenic model biofilm community. Appl Environ Microbiol 83(6):e03033–e03016

Extracellular Electron Transfer and Biosensors

33

3. Richter K, Schicklberger M, Gescher J (2012) Dissimilatory reduction of extracellular electron acceptors in anaerobic respiration. Appl Environ Microbiol 78:913–921 4. Sturm G, Dolch K, Richter K, Rautenberg M, Gescher J (2012) Metal reducers and reduction targets. A short survey about the distribution of dissimilatory metal reducers and the multitude of terminal electron acceptors. Microbial metal respiration: from geochemistry to potential applications. Springer-Verlag, Heidelberg, pp 129–159 5. Firer-Sherwood M, Pulcu GS, Elliott SJ (2008) Electrochemical interrogations of the Mtr cytochromes from Shewanella: opening a potential window. J Biol Inorg Chem 13:849–854 6. Nevin KP, Lovley DR (2002) Mechanisms for Fe(III) oxide reduction in sedimentary environments. Geomicrobiol J 19:141–159 7. Straub KL, Schink B (2003) Evaluation of electron-shuttling compounds in microbial ferric iron reduction. FEMS Microbiol Lett 220:229–233 8. Aklujkar M, Coppi MV, Leang C et al (2013) Proteins involved in electron transfer to Fe(III) and Mn(IV) oxides by Geobacter sulfurreducens and Geobacter uraniireducens. Microbiology 159:515–535 9. Butler JE, Young ND, Lovley DR (2010) Evolution of electron transfer out of the cell: comparative genomics of six Geobacter genomes. BMC Genomics 11:40 10. Holmes DE, Chaudhuri SK, Nevin KP et al (2006) Microarray and genetic analysis of electron transfer to electrodes in Geobacter sulfurreducens. Environ Microbiol 8:1805–1815 11. Mehta T, Coppi MV, Childers SE, Lovley DR (2005) Outer membrane c-type cytochromes required for Fe(III) and Mn(IV) oxide reduction in Geobacter sulfurreducens. Appl Environ Microbiol 71:8634–8641 12. Ding YHR, Hixson KK, Giometti CS et al (2006) The proteome of dissimilatory metalreducing microorganism Geobacter sulfurreducens under various growth conditions. Biochim Biophys Acta Proteins Proteomics 1764:1198–1206 13. Aklujkar M, Krushkal J, DiBartolo G, Lapidus A, Land ML, Lovley DR (2009) The genome sequence of Geobacter metallireducens: features of metabolism, physiology and regulation common and dissimilar to Geobacter sulfurreducens. BMC Microbiol 9(1):109 14. Morgado L, Saraiva IH, Louro RO, Salgueiro CA (2010) Orientation of the axial ligands and magnetic properties of the hemes in the triheme ferricytochrome PpcA from G. sulfurreducens determined by paramagnetic NMR. FEBS Lett 584:3442–3445 15. Zacharoff L, Chan CH, Bond DR (2016) Reduction of low potential electron acceptors requires the CbcL inner membrane cytochrome of Geobacter sulfurreducens. Bioelectrochemistry 107:7–13 16. Levar CE, Chan CH, Mehta-Kolte MG, Bond DR (2014) An inner membrane cytochrome required only for reduction of high redox potential extracellular electron acceptors. MBio 5 (6):e02034–e02014 17. Seidel J, Hoffmann M, Ellis KE et al (2012) MacA is a second cytochrome c peroxidase of Geobacter sulfurreducens. Biochemistry 51:2747–2756 18. Qian XL, Reguera G, Mester T, Lovley DR (2007) Evidence that OmcB and OmpB of Geobacter sulfurreducens are outer membrane surface proteins. FEMS Microbiol Lett 277:21–27 19. Leang C, Coppi MV, Lovley DR (2003) OmcB, a c-type polyheme cytochrome, involved in Fe(III) reduction in Geobacter sulfurreducens. J Bacteriol 185:2096–2103 20. Liu X, Tremblay PL, Malvankar NS, Nevin KP, Lovley DR, Vargas M (2014) A Geobacter sulfurreducens strain expressing Pseudomonas aeruginosa type IV pili localizes OmcS on Pili but is deficient in Fe(III) oxide reduction and current production. Appl Environ Microbiol 80:1219–1224 21. Liu YM, Fredrickson JK, Zachara JM, Shi L (2015) Direct involvement of ombB, omaB, and omcB genes in extracellular reduction of Fe(III) by Geobacter sulfurreducens PCA. Front Microbiol 6:–1075

34

F. Simonte et al.

22. Liu YM, Wang ZM, Liu J et al (2014) A trans-outer membrane porin-cytochrome protein complex for extracellular electron transfer by Geobacter sulfurreducens PCA. Environ Microbiol Rep 6:776–785 23. Hartshorne RS, Reardon CL, Ross D et al (2009) Characterization of an electron conduit between bacteria and the extracellular environment. Proc Natl Acad Sci U S A 106:22169–22174 24. Qian XL, Mester T, Morgado L et al (2011) Biochemical characterization of purified OmcS, a c-type cytochrome required for insoluble Fe(III) reduction in Geobacter sulfurreducens. BBA-Bioenergetics 1807:404–412 25. Leang C, Qian XL, Mester T, Lovley DR (2010) Alignment of the c-type cytochrome OmcS along pili of Geobacter sulfurreducens. Appl Environ Microbiol 76:4080–4084 26. Summers ZM, Fogarty HE, Leang C, Franks AE, Malvankar NS, Lovley DR (2010) Direct exchange of electrons within aggregates of an evolved syntrophic coculture of anaerobic bacteria. Science 330:1413–1415 27. Nevin KP, Kim BC, Glaven RH et al (2009) Anode biofilm transcriptomics reveals outer surface components essential for high density current production in Geobacter sulfurreducens fuel cells. PLoS One 4:e5628 28. Strycharz SM, Glaven RH, Coppi MV et al (2011) Gene expression and deletion analysis of mechanisms for electron transfer from electrodes to Geobacter sulfurreducens. Bioelectrochemistry 80:142–150 29. Lovley DR (2011) Reach out and touch someone: potential impact of DIET (direct interspecies energy transfer) on anaerobic biogeochemistry, bioremediation, and bioenergy. Rev Environ Sci Biotechnol 10:101–105 30. Rotaru AE, Shrestha PM, Liu F, Markovaite B, Chen S, Nevin KP, Lovley DR (2014) Direct interspecies electron transfer between Geobacter metallireducens and Methanosarcina barkeri. Appl Environ Microbiol 80:4599–4605 31. Rotaru AE, Shrestha PM, Liu FH et al (2014) A new model for electron flow during anaerobic digestion: direct interspecies electron transfer to Methanosaeta for the reduction of carbon dioxide to methane. Energ Environ Sci 7:408–415 32. Rotaru AE, Shrestha PM, Liu FH, Ueki T, Nevin K, Summers ZM, Lovley DR (2012) Interspecies electron transfer via hydrogen and Formate rather than direct electrical connections in cocultures of Pelobacter carbinolicus and Geobacter sulfurreducens. Appl Environ Microbiol 78:7645–7651 33. Rollefson JB, Stephen CS, Tien M, Bond DR (2011) Identification of an extracellular polysaccharide network essential for cytochrome anchoring and biofilm formation in Geobacter sulfurreducens. J Bacteriol 193:1023–1033 34. Malvankar NS, Vargas M, Nevin KP et al (2011) Tunable metallic-like conductivity in microbial nanowire networks. Nat Nanotechnol 6:573–579 35. Reguera G, McCarthy KD, Mehta T, Nicoll JS, Tuominen MT, Lovley DR (2005) Extracellular electron transfer via microbial nanowires. Nature 435:1098–1101 36. Yates MD, Strycharz-Glaven SM, Golden JP, Roy J, Tsoi S, Erickson JS, El-Naggar MY, Barton SC, Tender LM (2016) Measuring conductivity of living Geobacter sulfurreducens biofilms. Nat Nanotechnol 11:910–913 37. Malvankar NS, Rotello VM, Tuominen MT, Lovley DR (2016) Reply to ‘Measuring conductivity of living Geobacter sulfurreducens biofilms’. Nat Nanotechnol 11:913–914 38. Childers SE, Ciufo S, Lovley DR (2002) Geobacter metallireducens accesses insoluble Fe (III) oxide by chemotaxis. Nature 416:767–769 39. Lovley DR (2012) Long-range electron transport to Fe(III) oxide via pili with metallic-like conductivity. Biochem Soc Trans 40:1186–1190 40. Malvankar NS, Tuominen MT, Lovley DR (2012) Lack of cytochrome involvement in longrange electron transport through conductive biofilms and nanowires of Geobacter sulfurreducens. Energ Environ Sci 5:8651–8659

Extracellular Electron Transfer and Biosensors

35

41. Tremblay PL, Aklujkar M, Leang C, Nevin KP, Lovley D (2012) A genetic system for Geobacter metallireducens: role of the flagellin and pilin in the reduction of Fe(III) oxide. Environ Microbiol Rep 4:82–88 42. Shrestha PM, Rotaru AE, Aklujkar M et al (2013) Syntrophic growth with direct interspecies electron transfer as the primary mechanism for energy exchange. Environ Microbiol Rep 5:904–910 43. Vargas M, Malvankar NS, Tremblay PL et al (2013) Aromatic amino acids required for Pili conductivity and long-range extracellular electron transport in Geobacter sulfurreducens. MBio 4(2):e00105–e00113 44. Tan Y, Adhikari RY, Malvankar NS et al (2016) Synthetic biological protein nanowires with high conductivity. Small 12:4481–4485 45. Yi HN, Nevin KP, Kim BC, Franks AE, Klimes A, Tender LM, Lovley DR (2009) Selection of a variant of Geobacter sulfurreducens with enhanced capacity for current production in microbial fuel cells. Biosens Bioelectron 24:3498–3503 46. Burns JL, DiChristina TJ (2009) Anaerobic respiration of elemental sulfur and thiosulfate by Shewanella oneidensis MR-1 requires psrA, a homolog of the phsA gene of Salmonella enterica serovar Typhimurium LT2. Appl Environ Microbiol 75:5209–5217 47. Gralnick JA, Vali H, Lies DP, Newman DK (2006) Extracellular respiration of dimethyl sulfoxide by Shewanella oneidensis strain MR-1. Proc Natl Acad Sci U S A 103:4669–4674 48. Myers CR, Nealson KH (1988) Bacterial manganese reduction and growth with manganese oxide as the sole electron-acceptor. Science 240:1319–1321 49. Sturm G, Richter K, Doetsch A, Heide H, Louro RO, Gescher J (2015) A dynamic periplasmic electron transfer network enables respiratory flexibility beyond a thermodynamic regulatory regime. ISME J 9:1802–1811 50. Myers CR, Myers JM (1997) Cloning and sequence of cymA a gene encoding a tetraheme cytochrome c required for reduction of iron(III), fumarate, and nitrate by Shewanella putrefaciens MR-1. J Bacteriol 179:1143–1152 51. Beliaev AS, Klingeman DM, Klappenbach JA et al (2005) Global transcriptome analysis of Shewanella oneidensis MR-1 exposed to different terminal electron acceptors. J Bacteriol 187:7138–7145 52. Schuetz B, Schicklberger M, Kuermann J, Spormann AM, Gescher J (2009) Periplasmic electron transfer via the c-type cytochromes MtrA and FccA of Shewanella oneidensis MR-1. Appl Environ Microbiol 75:7789–7796 53. McMillan DGG, Marritt SJ, Firer-Sherwood MA et al (2013) Protein-protein interaction regulates the direction of catalysis and electron transfer in a redox enzyme complex. J Am Chem Soc 135:10550–10556 54. Alves MN, Neto SE, Alves AS et al (2015) Characterization of the periplasmic redox network that sustains the versatile anaerobic metabolism of Shewanella oneidensis MR-1. Front Microbiol 6:665. https://doi.org/10.3389/fmicb.2015.00665 55. Coursolle D, Baron DB, Bond DR, Gralnick JA (2010) The Mtr respiratory pathway is essential for reducing flavins and electrodes in Shewanella oneidensis. J Bacteriol 192:467–474 56. Marsili E, Baron DB, Shikhare ID, Coursolle D, Gralnick JA, Bond DR (2008) Shewanella secretes flavins that mediate extracellular electron transfer. Proc Natl Acad Sci U S A 105:3968–3973 57. von Canstein H, Ogawa J, Shimizu S, Lloyd JR (2008) Secretion of flavins by Shewanella species and their role in extracellular electron transfer. Appl Environ Microbiol 74:615–623 58. Brutinel ED, Gralnick JA (2012) Shuttling happens: soluble flavin mediators of extracellular electron transfer in Shewanella. Appl Microbiol Biotechnol 93:41–48 59. Baron D, LaBelle E, Coursolle D, Gralnick JA, Bond DR (2009) Electrochemical measurement of electron transfer kinetics by Shewanella oneidensis MR-1. J Biol Chem 284:28865–28873

36

F. Simonte et al.

60. Ross DE, Flynn JM, Baron DB, Gralnick JA, Bond DR (2011) Towards electrosynthesis in Shewanella: energetics of reversing the Mtr pathway for reductive metabolism. PLoS One 6 (2):e16649 61. Breuer M, Rosso KM, Blumberger J, Butt JN (2015) Multi-haem cytochromes in Shewanella oneidensis MR-1: structures, functions and opportunities. J R Soc Interface 12:20141117 62. Clarke TA, Edwards MJ, Gates AJ et al (2011) Structure of a bacterial cell surface decaheme electron conduit. Proc Natl Acad Sci U S A 108:9384–9389 63. Edwards MJ, White GF, Norman M et al (2015) Redox linked Flavin sites in extracellular Decaheme proteins involved in microbe-mineral electron transfer. Sci Rep 5:11677 64. Okamoto A, Hashimoto K, Nealson KH, Nakamura R (2013) Rate enhancement of bacterial extracellular electron transport involves bound flavin semiquinones. Proc Natl Acad Sci U S A 110:7856–7861 65. Okamoto A, Nakamura R, Nealson KH, Hashimoto K (2014) Bound Flavin model suggests similar electron-transfer mechanisms in Shewanella and Geobacter. ChemElectroChem 1:1808–1812 66. Okamoto A, Kalathil S, Deng X, Hashimoto K, Nakamura R, Nealson KH (2014) Cellsecreted flavins bound to membrane cytochromes dictate electron transfer reactions to surfaces with diverse charge and pH. Sci Rep 4 67. Ding M, Shiu HY, Li SL et al (2016) Nanoelectronic investigation reveals the electrochemical basis of electrical conductivity in Shewanella and Geobacter. ACS Nano 10:9919–9926 68. Koch C, Harnisch F (2016) Is there a specific ecological niche for electroactive microorganisms? ChemElectroChem 3:1282–1295 69. Carlson HK, Iavarone AT, Gorur A et al (2012) Surface multiheme c-type cytochromes from Thermincola potens and implications for respiratory metal reduction by Gram- positive bacteria. Proc Natl Acad Sci U S A 109:1702–1707 70. Wrighton KC, Thrash JC, Melnyk RA et al (2011) Evidence for direct electron transfer by a Gram-positive bacterium isolated from a microbial fuel cell. Appl Environ Microbiol 77:7633–7639 71. Lusk BG, Parameswaran P, Popat SC, Rittmann BE, Torres CI (2016) The effect of pH and buffer concentration on anode biofilms of Thermincola ferriacetica. Bioelectrochemistry 112:47–52 72. Clark Jr LC, Lyons C (1962) Electrode systems for continuous monitoring un cardiovascular surgery. Ann N Y Acad Sci 102:29–45 73. Setford SJ, Newman JD (2005) Enzyme biosensors. vol 17, pp 29–60 74. Turner AP, Karube I, Wilson GS (1987) Biosensors fundamentals and applications. Oxford Science Publications, Oxford, Engalnd 75. Lazcka O, Del Campo FJ, Munoz FX (2007) Pathogen detection: a perspective of traditional methods and biosensors. Biosens Bioelectron 22:1205–1217 76. Luong JH, Male KB, Glennon JD (2008) Biosensor technology: technology push versus market pull. Biotechnol Adv 26:492–500 77. Su L, Jia W, Hou C, Lei Y (2011) Microbial biosensors: a review. Biosens Bioelectron 26:1788–1799 78. Belkin S (2003) Microbial whole-cell sensing systems of environmental pollutants. Curr Opin Microbiol 6:206–212 79. Kumlanghan A, Kanatharana P, Asawatreratanakul P, Mattiasson B, Thavarungkul P (2008) Microbial BOD sensor for monitoring treatment of wastewater from a rubber latex industry. Enzyme Microb Technol 42:483–491 80. Nakamura H, Suzuki K, Ishikuro H et al (2007) A new BOD estimation method employing a double-mediator system by ferricyanide and menadione using the eukaryote Saccharomyces cerevisiae. Talanta 72:210–216 81. APHA (1998) Standard methods for the examination of waters and wastewaters, vol 20. American Public Health Association, Washington

Extracellular Electron Transfer and Biosensors

37

82. SIS (1979) Water analysis - determination of biochemical oxygen demand, BOD, of water dilution method (Svensk standard SS 02 81 43 E), vol 1. The Swedish Standards Institution, Stockholm 83. Chang IS, Jang JK, Gil GC, Kim M, Kim HJ, Cho BW, Kim BH (2004) Continuous determination of biochemical oxygen demand using microbial fuel cell type biosensor. Biosens Bioelectron 19:607–613 84. Kim BH, Chang IS, Gil GC, Park HS, Kim HJ (2003) Novel BOD (biochemical oxygen demand) sensor using mediator-less microbial fuel cell. Biotechnol Lett 25:541–545 85. Clark Jr LC (1956) Monitor and control of blood and tissue oxygen tensions. Trans Am Soc Artif Intern Organs 2(1):41–48 86. Karube I, Matsunaga T, Mitsuda S, Suzuki S (1977) Microbial electrode BOD sensors. Biotechnol Bioeng XIX:1535–1547 87. Liu J, Mattiasson B (2002) Microbial BOD sensors for wastewater analysis. Water Res 36:3786–3802 88. Ponomareva ON, Arlyapov VA, Alferov VA, Reshetilov AN (2011) Microbial biosensors for detection of biological oxygen demand (a review). Appl Biochem Microbiol 47:1–11 89. Liu J, Bjornsson L, Mattiasson B (2000) Immobilised activated sludge based biosensor for biochemical oxygen demand measurement. Biosens Bioelectron 14(12):883–893 90. Chee G, Nomura Y, Karube I (1999) Biosensor for the estimation of low biochemical oxygen demand. Anal Chim Acta 379:185–191 91. Kulys J, Kadziauskiene K (1980) Yeast BOD sensor. Biotechnol Bioeng XXII:221–226 92. Marty JL, Olive D, Asano Y (1997) Measurement of BOD: correlation between 5-day BOD and commercial BOD biosensor values. Environ Technol 18:333–337 93. Riedel K, Renneberg R, Kühn M, Scheller F (1988) A fast estimation of biochemical oxygen demandusing microbial sensors. Appl Microbiol Biotechnol 28:316–318 94. Kang KH, Jang JK, Pham TH, Moon H, Chang IS, Kim BH (2003) A microbial fuel cell with improved cathode reaction as a low biochemical oxygen demand sensor. Biotechnol Lett 25:1357–1361 95. Grzebyk M, Poz´niak G (2005) Microbial fuel cells (MFCs) with interpolymer cation exchange membranes. Sep Purif Technol 41:321–328 96. Rabaey K, Verstraete W (2005) Microbial fuel cells: novel biotechnology for energy generation. Trends Biotechnol 23:291–298 97. Tkac J, Vostiar I, Gorton L, Gemeiner P, Sturdik E (2003) Improved selectivity of microbial biosensor using membrane coating. Application to the analysis of ethanol during fermentation. Biosens Bioelectron 18:1125–1134 98. Trosok SP, Driscoll BT, Luong JHT (2001) Mediated microbial biosensor using a novel yeast strain for wastewater BOD measurement. Appl Microbiol Biotechnol 56:550–554 99. Yoshida N, Hoashi J, Morita T, McNiven S, Nakamura H, Karube I (2001) Improvement of a mediator-type biochemical oxygen demand sensor for on-site measurement. J Biotechnol 88:269–275 100. Pasco NF, Baronian K, Jeffries C, Hay J (2000) Biochemical mediator demand - a novel rapid alternative for measuring biochemical oxygen demand. Appl Microbiol Biotechnol 53:613–618 101. Morris K, Zhao H, John R (2003) The use of a mixed microbial consortium in a rapid ferricyanide mediated biochemical oxygen demand assay. Trans Ecol Environ:65 102. Gil G-C, Chang I-S, Kim BH, Kim M, Jang J-K, Park HS, Kim HJ (2003) Operational parameters affecting the performannce of a mediator-less microbial fuel cell. Biosens Bioelectron 18:327–334 103. Pasco NF, Weld RJ, Hay JM, Gooneratne R (2011) Development and applications of whole cell biosensors for ecotoxicity testing. Anal Bioanal Chem 400:931–945 104. Kumlanghan A, Liu J, Thavarungkul P, Kanatharana P, Mattiasson B (2007) Microbial fuel cell-based biosensor for fast analysis of biodegradable organic matter. Biosens Bioelectron 22:2939–2944

38

F. Simonte et al.

105. Di Lorenzo M, Curtis TP, Head IM, Scott K (2009) A single-chamber microbial fuel cell as a biosensor for wastewaters. Water Res 43:3145–3154 106. Di Lorenzo M, Thomson AR, Schneider K, Cameron PJ, Ieropoulos I (2014) A small-scale air-cathode microbial fuel cell for on-line monitoring of water quality. Biosens Bioelectron 62:182–188 107. Kim M, Sik Hyun M, Gadd GM, Joo Kim H (2007) A novel biomonitoring system using microbial fuel cells. J Environ Monit 9:1323–1328 108. Stein NE, Hamelers HMV, van Straten G, Keesman KJ (2012) On-line detection of toxic components using a microbial fuel cell-based biosensor. J Process Control 22:1755–1761 109. Wang B, Barahona M, Buck M (2013) A modular cell-based biosensor using engineered genetic logic circuits to detect and integrate multiple environmental signals. Biosens Bioelectron 40:368–376 110. Hernandez Leal L, Soeter AM, Kools SA et al (2012) Ecotoxicological assessment of grey water treatment systems with Daphnia magna and Chironomus riparius. Water Res 46:1038–1044 111. Matsunaga T, Takeyama H, Nakao T, Yamazawa A (1999) Screening of marine microalgae for bioremediation of cadmium- polluted seawater. J Biotechnol 70:33–38 112. Qu R, Wang X, Liu Z, Yan Z, Wang Z (2013) Development of a model to predict the effect of water chemistry on the acute toxicity of cadmium to Photobacterium phosphoreum. J Hazard Mater 262:288–296 113. Zhang H, Cao H, Meng Y, Jin G, Zhu M (2012) The toxicity of cadmium (Cd(2)(+)) towards embryos and pro-larva of soldatov’s catfish (Silurus Soldatovi). Ecotoxicol Environ Saf 80:258–265 114. Lee H, Yang W, Wei X, Fraiwan A, Choi S (2015) A microsized microbial fuel cell based biosensor for fast and sensitive detection of toxic substances in water 115. Patil S, Harnisch F, Schroder U (2010) Toxicity response of electroactive microbial biofilms a decisive feature for potential biosensor and power source applications. ChemPhysChem 11:2834–2837 116. Davila D, Esquivel JP, Sabate N, Mas J (2011) Silicon-based microfabricated microbial fuel cell toxicity sensor. Biosens Bioelectron 26:2426–2430 117. Tront JM, Fortner JD, Plotze M, Hughes JB, Puzrin AM (2008) Microbial fuel cell biosensor for in situ assessment of microbial activity. Biosens Bioelectron 24:586–590 118. Tront JM, Fortner JD, Plotze M, Hughes JB, Puzrin AM (2008) Microbial fuel cell technology for measurement of microbial respiration of lactate as an example of bioremediation amendment. Biotechnol Lett 30:1385–1390 119. Holtmann D, Sell D (2002) Detection of the microbial activity of aerobic heterotrophic, anoxic heterotrophic and aerobic autotrophic activated sludge organisms with an electrochemical sensor. Biotechnol Lett 24:1313–1318 120. Holtmann D, Schrader J, Sell D (2006) Quantitative comparison of the signals of an electrochemical bioactivity sensor during the cultivation of different microorganisms. Biotechnol Lett 28:889–896 121. Golitsch F, Bücking C, Gescher J (2013) Proof of principle for an engineered microbial biosensor based on Shewanella oneidensis outer membrane protein complexes. Biosens Bioelectron 47:285–291 122. Bücking C, Popp F, Kerzenmacher S, Gescher J (2010) Involvement and specificity of Shewanella oneidensis outer membrane cytochromes in the reduction of soluble and solidphase terminal electron acceptors. FEMS Microbiol Lett 306:144–151 123. Webster DP, TerAvest MA, Doud DF et al (2014) An arsenic-specific biosensor with genetically engineered Shewanella oneidensis in a bioelectrochemical system. Biosens Bioelectron 62:320–324

Adv Biochem Eng Biotechnol (2019) 167: 39–86 DOI: 10.1007/10_2017_42 © Springer International Publishing AG 2017 Published online: 10 December 2017

Electron Transfer Between Enzymes and Electrodes Tanja Vidakovic-Koch

Abstract Efficient electron transfer between redox enzymes and electrocatalytic surfaces plays a significant role in development of novel energy conversion devices as well as novel reactors for production of commodities and fine chemicals. Major application examples are related to enzymatic fuel cells and electroenzymatic reactors, as well as enzymatic biosensors. The two former applications are still at the level of proof-of-concept, partly due to the low efficiency and obstacles to electron transfer between enzymes and electrodes. This chapter discusses the theoretical backgrounds of enzyme/electrode interactions, including the main mechanisms of electron transfer, as well as thermodynamic and kinetic aspects. Additionally, the main electrochemical methods of study are described for selected examples. Finally, some recent advancements in the preparation of enzymemodified electrodes as well as electrodes for soluble co-factor regeneration are reviewed.

T. Vidakovic-Koch (*) Max Planck Institute for Dynamics of Complex Technical Systems, Magdeburg, Germany e-mail: [email protected]

40

T. Vidakovic-Koch

Renewable feedstock

+/-

e-

Bioanode

Green Catalysts

Biocathode

Graphical Abstract

e-

-/+

Valuable products and/or Electricity

eRenewable energy

Efficient electron transfer (DET or MET)

e-

Keywords Direct electron transfer, Electrochemical co-factor regeneration, Electrochemical methods, Kinetics, Mediated electron transfer, Porous electrodes, Redox enzymes Contents 1 Types of Electron Transfer Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Some Mechanistic Aspects of Enzyme/Electrode Electron Transfer . . . . . . . . . . . . . . . . . 2 Thermodynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Mechanisms of Bioelectrochemical Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Reaction Kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Balance Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Methods of Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Stationary Polarization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Cyclic Voltammetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Electrochemical Impedance Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Parameter Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Toward the Development of Electrobiotechnological Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Development of Enzyme-Modified Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Electrochemical Regeneration of Soluble Co-Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix: List of Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Greek ........................................................................ Super- and Sub-scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41 43 44 47 47 50 51 55 57 60 62 66 72 72 76 78 79 80 81 81 82

Electron Transfer Between Enzymes and Electrodes

41

1 Types of Electron Transfer Mechanisms The natural cycle of oxidoreductases involves several substrates. Normally, two substrates and two products are required, which corresponds to a so-called bi-bi reaction in the general nomenclature of enzyme reactions [1]. A convenient example of a bi-bi reaction is glucose oxidation by the glucose-oxidase enzyme, in accordance to: D
glucose þ O2 ! D
glucono
1, 5
lactone þ H2 O2

ð1Þ

Here, glucose and oxygen are the two substrates of this enzyme. The co-factor of glucose-oxidase is flavin adenine dinucleotide (FAD). FAD is usually deeply buried into the protein backbone, which normally excludes any direct electron transfer between FAD and the electrode. It is regenerated by its normal electron acceptor oxygen or artificial electron acceptors. Another enzyme that catalyzes glucose oxidation is nicotinamide adenine dinucleotide (NAD) dependent glucose 1-dehydrogenase: D
glucose þ NADþ ! D
glucono
1, 5
lactone þ NADH þ Hþ

ð2Þ

In this example, the two enzyme substrates are glucose and NAD+ (or its phosphorylated derivate nicotinamide adenine dinucleotide phosphate [NADP+]). NAD+ is also the soluble enzyme co-factor, for which reason it is also the co-substrate. A similar case is NAD (NADP) dependent alcohol dehydrogenases, where alcohol is one substrate and NAD+ or NADP+ is the second one. Some enzymes require more than two substrates. Examples of interest for electrobiotechnological applications are P450 monooxygenases. The most common reaction catalyzed by these enzymes is the insertion of molecular oxygen chemoselectively into the inert C-H bond of an organic substrate (RH), as follows: RH þ NADPH þ Hþ þ O2 ! ROH þ NADPþ þ H2 O

ð3Þ

Here, one of substrates (NADPH) is a common soluble co-factor of many enzymes. During biotransformations (Eqs. 1, 2, and 3), oxidoreductases participate actively in the electron transfer between different substrates. The redox changes on the enzyme side take place mainly on co-factors. Some of the previously mentioned enzymes have only one co-factor (glucose oxidase, NAD dependent glucose 1-dehydrogenase), but many oxidoreductases have more than one. For example, cellobiose dehydrogenase has two co-factors (FAD and heme), whereas P450 often has three (FAD, flavin mononucleotide (FMN), and heme). If more co-factors are available, the electron transfer between enzyme substrates is accompanied by an interprotein electron transfer between co-factors. Usually, one of the co-factors is a catalytic center, whereas the other participates only in the electron transfer.

42

T. Vidakovic-Koch

Fig. 1 Schematic representation of a porous enzymatic electrode structure. Two mechanisms of enzyme/electron conductive support electron transfer are shown: mediated electron transfer (MET) and direct electron transfer (DET). CC current collector, CL catalyst layer, Medi i ¼ ox, red (oxidized and reduced forms of a mediator), S substrate, P product. Reprinted from [2] with permission from Elsevier

In electrobiotechnology, one of the natural enzyme substrates is replaced by an artificial substrate (a so-called mediator) or an electrode itself. Alternatively, the natural enzyme substrates are regenerated electrochemically with the help of artificial substances or an electrode. The latter case applies if the natural enzyme substrate cannot be replaced (e.g. a co-substrate). The resulting electron transfer mechanisms are termed mediated electron transfer (MET) or direct electron transfer (DET). A schematic representation of a porous enzymatic electrode, showing its main components and the two main types of electron transfer mechanisms between enzymes and electron-conductive support, is shown in Fig. 1. The different electron transfer mechanisms are further described in examples of glucose oxidase and NAD-dependent formate dehydrogenase. In a glucose-oxidase natural enzyme substrate, oxygen is replaced by an artificial substrate, which is then electrochemically regenerated. An advantage of using a substrate other than oxygen is the avoidance of H2O2 formation, which is one of the products in the natural cycle of glucose oxidase. As is well known, hydrogen peroxide can inhibit glucoseoxidase and influences its long-term stability. In the presence of a mediator, Eq. (1) can be rewritten as follows: D
glucose þ Medox ! D
glucono
1, 5
lactone þ Medred

ð4Þ

In the next step, the mediator is regenerated electrochemically: Medred ! Medox þ 2Hþ þ 2e
where Medox and Medred are the oxidized and reduced forms of the mediator.

ð5Þ

Electron Transfer Between Enzymes and Electrodes

43

The mediated regeneration of the NADH co-substrate on an example of formate dehydrogenase, can be represented as follows: CO2 þ NADH þ Hþ ! HCOOH þ NADþ NADþ þ Medred ! NADH þ Medox þ Medox þ þ Hþ þ 2e
! Medred

ð6Þ ð7Þ ð8Þ

Alternatively to Eq. (7), NAD+ can be directly reduced at the electrode surface, in which case a DET occurs: NADþ þ Hþ þ 2e
! NADH

ð9Þ

If NADH (NADPH) or its oxidized forms are enzyme substrates or co-substrates, its electrochemical regeneration or replacement by an artificial substrate is of high interest in electrobiotechnological applications. The main reasons for this are the high prices of these substrates, their difficult regeneration, and their tendency to decompose over time. The replacement of NADH (NADPH) is not straightforward because enzymes are optimized from nature for these electron donors. In some cases, protein engineering can tailor protein properties with respect to artificial substrates, as for example with P450 monooxygenase [3].

1.1

Some Mechanistic Aspects of Enzyme/Electrode Electron Transfer

The scientific basis of electron transfer between an enzyme and an electrode surface is described by the Marcus theory [4]. The electron transfer rate between two species participating in a redox process depends on the driving force in terms of the Gibbs free energy change of the reaction (i.e. the potential difference), the distance between redox centers and the reorganization energy. If these conditions are favorable, then DET is possible, as reported in several cases. For example, different peroxidases show DET (cytochrome c peroxidase, horseradish peroxidase) [5, 6]. Laccase and bilirubin-oxidase are also able to exchange electrons directly with an electrode surface [7]. DET has also been reported for Ni-Fe hydrogenases [8]. This DET is favored because of the presence of iron-sulfur clusters, which act as an intraelectron transfer chain; they channel electrons from the Ni-Fe catalytic center to the protein surface, which reduces the distance between the catalytic center and the electrode surface. Furthermore, DET has been reported for bifunctional enzymes such as cellobiose dehydrogenase, where the heme co-factor can exchange electrons with the electrode [9]. The efficiency of DET depends strongly on the enzyme orientation on the surface. For example, oriented enzyme immobilization on the electrode surface, with the help of self-assembled monolayers (SAMs), results in more efficient electron transfer rates [4]. Other possibilities for oriented enzyme immobilization

44

T. Vidakovic-Koch

were demonstrated by Zimmermann et al. [10], who reconstituted horseradish peroxidase (HRP) apo-enzyme on covalently attached heme co-factors on SAMs of a modified gold surface. However, this approach did not ultimately lead to improved electron transfer rates. Although there is experimental evidence that isolated FAD can exchange electrons with an electrode [11], native enzymes with FAD cofactors (glucose-oxidase, cellobiose dehydrogenase) are normally not able to exchange electrons with the electrode surface. This is due to the isolated protein shell, which blocks electron transfer between the enzyme and electrode. FAD-dependent enzymes are able to perform DET only after partial removal of the protein shell by, for example, deglycosylation [12]. Yet, in number of cases, DET for GOx used in porous enzymatic electrodes with carbon nanotubes was reported. Some possible artefacts that lead to the conclusion of GOx DET were summarized elsewhere [13]. In free diffusive mediators, the electron transfer mechanism is similar to that of natural enzyme substrates. Namely, the mediator diffuses in and out of the enzyme, enabling the electron shuttle. If the mediator is immobilized, the mechanism of MET is not so clear. Some mechanistic interpretations of MET were discussed by Habermüller et al. [4], who emphasized the need for a degree of free diffusional movement of the redox mediator for efficient electron transfer. This can be achieved by a so-called electron-hopping mechanism, where the electron transfer distance between an enzyme co-factor and an electrode is reduced by dividing the overall electron transfer process into a sequence of electron-hopping reactions between adjacent redox mediator molecules. This approach might be used for redox relays in modified conducting polymers serving as an immobilization matrix. Some examples are modified polypyrrole matrices in ferrocene or osmium complexes. Additionally, some level of free diffusional movement of redox mediators can be achieved by the covalent attachment of the redox mediator via long and flexible spacer chains, either to the electrode surface, a suitable matrix, or the outer surface of the enzyme. The latter case uses the so-called whipping mechanism of MET.

2 Thermodynamics Let us consider the following well-known reaction, which is catalyzed by the glucose-oxidase: D
glucose þ O2 ! H2 O2 þ D
glucono
1, 5
lactone

ð10Þ

This biochemical redox reaction can be separated into two half-reactions, as follows: D
glucono
1, 5
lactone þ 2Hþ þ 2e
! D
glucose O2 þ 2Hþ þ 2e
! H2 O2

ð11Þ ð12Þ

Electron Transfer Between Enzymes and Electrodes

45

To satisfy the stoichiometry, the half-reactions (Eqs. 11 and 12) contain protons in addition to electrons. Normally, protons are excluded from biochemical reactions when the pH value (e.g. pH 7) is specified. Also following the electrochemical convention, both half-reactions are written as reduction reactions. The standard half-cell electrode potentials of Eqs. 11 and 12 can be calculated based on the thermodynamic data, with the hydrogen reaction as a reference half-cell reaction (H2 gas activity and H+ activity are assumed to be 1). The Gibbs free energy change can be calculated as follows: Δr Go ¼

n X

νi Δf Gio

ð13Þ

i¼1

Here, νi is a stoichiometric coefficient of the component i in the biochemical reaction (e.g. Eqs. 11 and 12), having positive signs for the products of the reaction and negative for the reactants. The standard Gibbs free energies of formation (Δf Gio) for selected biochemically relevant components are summarized in Table 1. Standard half-cell electrode potentials can be further calculated by applying the relationship between the standard electrode potential Eo and the standard Gibbs free energy of the reaction. This can be derived from the first law of thermodynamics, assuming a reversible process at a constant temperature and pressure in which both mechanical work and electrical work are done [16], and assuming that the electrical work can be expressed as the work required to move charge Q ¼ nF by the potential difference E: Δr Go ¼
nFEo

ð14Þ

Because biochemical reactions occur under conditions that are far away from standard conditions, electrode potentials at pH 7 instead of standard electrode

Table 1 Standard Gibbs free energies of formation from the elements

Substance Acetaldehyde CO2 D-glucono-1,5 lactone D-glucose Ethanol FAD FADH2 Formaldehyde H2O H2O2 Methanol NAD+ NADH a

ΔfGo/kJ mol
1
139.00a
394.36
905.92a
917.2
181.54a
79.83a
118.57a
102.53
237.13
134.03 (
136.69)
166.27
39.89a
17.24a

Calculated based on data in [14]

Reference [14] [15] [14] [15] [14, 15] [14] [14] [15] [14] [14, 15] [15] [14] [14]

46

T. Vidakovic-Koch

potentials are normally reported. Assuming a proton concentration of 10
7 and ideal solutions, electrode potentials at pH 7 (Eo, #) can be recalculated from the values under standard conditions (Eo) using the Nernst equation: Eo, # ¼ Eo þ νHþ ∙ 2:303

RT pH nF

ð15Þ

The values of half-cell electrode potentials at pH 7 are summarized in Table 2. The observed values in electroenzymatic systems are close to the calculated values in Table 2 only if the bioelectrochemical reaction is reversible and if it follows the DET mechanism. An example of such a reaction is the interconversion of H+ and H2 catalyzed by hydrogenase [17]. The data from Table 2 can be used for the calculation of reversible cell potentials, assuming standard concentrations of all involved species and standard conditions. Positive reversible cell potential values are characteristics of spontaneous systems, such as fuel cells (so-called galvanic systems); negative values are the characteristics of electrolytic systems (electrical energy has to be supplied to the cell in order to run the reaction in the desired direction). As can be seen in the example of methanol oxidation (Fig. 2), the oxidation reaction always takes place on the anode side and the reduction reaction on the cathode side, independent of the mode of operation (fuel cell or electrolyzer). However, the charge (plus/minus) of the anode and cathode depends on the mode of operation: in electrolysis mode, the anode is more positive than the cathode; in the fuel cell operation, it is opposite. As it is clear from the discussion, reversible cell potentials are theoretical values, which are independent of the type of

Table 2 Electrode potentials at pH 7 of selected biochemical reactions Half-reaction CO2 + 2H+ + 2e
¼ Formate Gluconate- + 2e
+ 3H+ ¼ D-glucose + H2O 2H+ + 2e
¼ H2 6CO2 + 24e
+ 24H+ ¼ D-glucose + 6H2O Formate + 3H+ + 2e
¼ Formaldehyde + H2O CO2 + 6H+ + 6e
¼ Methanol + H20 D-glucono-1,5 lactone + 2H+ + 2e
¼ D-glucose NAD+ + H+ + 2e
¼ NADH Formaldehyde + 2H+ + 2e
¼ Methanol FAD + 2H+ + 2e
¼ FADH2 Acetaldehyde + 2H+ + 2e
¼ ethanol O2 (g) + 2H+ + 2e
¼ H2O2 O2 (g) + 4H+ + 4e
¼ 2H2O H2O2 + 2H+ + 2e
¼ 2H2O a

SHE, standard hydrogen electrode Calculated based on data in Table 1

b

Eo,# vs. SHE/Va
0.524
0.47
0.414
0.404
0.400
0.396
0.356
0.324 (
0.32)
0.182
0.213 (
0.2)
0.197 0.294 0.816 1.335

Reference b

[15] [15] b b b b b

[15]

b b

[15] [15] b

[15] b

Electron Transfer Between Enzymes and Electrodes

47

Fig. 2 Half-cell reactions and electrode and reversible cell potentials for methanol to CO2 conversion (and vice versa)

catalyst (metal or different enzymes). The values that are measured experimentally under open circuit conditions (no current flow through the cell) are denoted as open cell potential (ocp) values. The ocp values close to theoretical values were experimentally measured only for fast reversible reactions (e.g. hydrogen reaction) [17] or at high temperatures (which are not applicable for biotechnological systems). Different mediators are used in combination with different cofactors (enzymes). An overview of some mediators used with FAD and NAD cofactors and their respective electrode potentials are provided in Table 3. The electrode potential values depend on the type of ions present in the solution, meaning that for most mediators only the formal potential values are reported.

3 Kinetics 3.1

Mechanisms of Bioelectrochemical Reactions

Bioelectrochemical reactions comprise biochemical and electrochemical steps. Additionally, mass transfer of the involved soluble substances to and from the electrode surface has to be considered. Assuming that only substrates and products of a bioelectrochemical reaction are soluble, the mass transfer from the bulk to the electrode surface can be represented as follows: !

Substratebulk diff Substrate

ð16Þ

Product diff Productbulk

ð17Þ

!

Here, the substrate and product are any soluble enzyme substrate and product concentrations in the bulk and at the electrode surface.

48

T. Vidakovic-Koch

Table 3 Mediators typically used in connection with FAD and NAD cofactors and their respective formal redox potentials Mediator With FAD cofactor 2-methyl-1-4-naphtoquinone on PLL Os polymers Pyrroloquinoline quinone p-Benzoquinone 8-hydroxyquinoline-5-sulfonic acid Phenazine methosulfate Tetrathiafulvalene Poly(vinilferrocene) Ferrocene monocarboxylic acid With NAD cofactor Nile blue Methylene greenb Poly(brilliant cresyl blue) Poly(methylene blue)

Formal redox potential vs. SHEa/V

pH

Reference


0.07 0.01 (0.09) 0.01 (0.04) 0.11 (0.08) 0.6 0.305 0.32 0.419 0.5 0.527

7 5 7 7 (7.2) 7 5 6 7 7 7

[18] [19, 20] [19, 21] [11, 22] [23] [24] [25] [26] [27] [28]


0.15 0/0.15 0.09 0.1

7 6 7 6

[29] [30] [31] [32]

E vs. Ag/AgCl ¼ E vs. SHE-0.197 (V); E vs. SCE ¼ E vs. SHE-0.24 (V) Methylene green shows two redox peaks

a

b

Biochemical steps are related to reactions between an enzyme and a substrate, while electrochemical steps are related to enzyme regeneration steps. Due to the variety of enzymes, different mechanisms of enzyme/substrate reactions are operative. In the following bi-bi mechanism, is shown in more detail because it is common to many oxidoreductases. A bi-bi reaction can be represented as follows: Substrate 1 þ Substrate 2 ¼ Product 1 þ Product 2

ð18Þ

This reaction holds for both the natural cycle of enzymes as well as for bioelectrochemical regeneration. The reaction (Eq. 18) commonly follows a double-displacement or Ping-Pong mechanism, where Substrate 1 binds first to the enzyme (Eq. 19). In this step, the enzyme is reduced/oxidized and Product 1 is formed. In the second step, the reduced/oxidized form of the enzyme binds Substrate 2 (Eq. 20). As a result, the enzyme is regenerated and Product 2 is formed. In electrobiotechnological applications, Substrate 2 is an artificial substrate (a mediator) or it is an electrode itself. Assuming MET, one can write the following: Ered þ Substrate 1 $ Eox þ Product 1 Eox þ Medox $ Ered þ Medred Here, Ered and Eox are the reduced and oxidized forms of an enzyme.

ð19Þ ð20Þ

Electron Transfer Between Enzymes and Electrodes

49

The kinetics of biochemical (enzymatic) steps are commonly described by the Briggs–Haldane mechanism, which is a more generalized formulation of the Michaelis–Menten mechanism, including explicitly the rate constants of forward and backward reactions of enzyme substrate (ES) formation [33]. Assuming that the overall reaction is a reduction of Substrate 1 to Product 1, one can write the following: $

Substrate 1 þ Ered k1 =k
1 ES ES

$ kcat

Product 1 þ Eox

ð21Þ ð22Þ

The kinetics of MET enzyme regeneration can be also described by the Briggs– Haldane mechanism: $

Eox þ Medred km1 =k
m1 EM EM

! km2

Medox þ Eox

ð23Þ ð24Þ

In most publications, the steps in Eqs. 23 and 24 are lumped together, showing only one effective rate constant (km), as follows: *

Eox þ Medred km Medox þ Ered

ð25Þ

The electrochemical regeneration of the mediator can be represented as follows: *

Medox þ ne
kme Medred

ð26Þ

In the simplest and most favorable case reaction (Eq. 26) involves transfer of only one electron. In DET, instead of Eq. (25), one can write the following: $

Eox þ ne
ke =k
e Ered

ð27Þ

The reaction in Eq. 27 involves transfers of more than one electron, which are not likely to occur all at once; therefore, this step can be split into several one-electron steps. Assuming a two-electron process and the absence of intermediate chemical steps, one can write the following: $

Eox þ e
ke1 =k
e1 Int

ð28Þ

Int þ e
ke2 =k
e2

ð29Þ

$

Ered

50

3.2

T. Vidakovic-Koch

Reaction Kinetics

The rate expressions for single steps in a DET or MET mechanism can be formulated based on the law of mass action, which assumes that the reaction rates are proportional to the activities (concentrations) of involved species. This approach is only valid for elementary steps in the mechanism. For lumped steps, empirical kinetic expressions can be also used. The rate expressions for the biochemical steps (Eqs. 21 and 22) can be formulated as follows: r 1 ¼ k1 ΓEred ðtÞ cR ð0; tÞ
k
1 ΓES ðtÞ r 2 ¼ k2 ΓES ðtÞ

ð30Þ ð31Þ

Here, ki (i ¼ 1,
1,2) are rate constants with units (mol
1 m3 s
1, s
1, s
1). Γi (i ¼ Ered,ES) are surface concentrations of the reduced form of the enzyme and enzyme substrate complex with units (mol m
2geo). For the DET mechanism, enzyme regeneration involves direct electron transfer between the enzyme and the electrode surface. The kinetics of steps (Eqs. 28 and 29) can be described by the empirical Butler-Volmer (BV) formulation or the more theoretical Marcus relationship (not discussed here). For the Butler-Volmer equation (Eq. 28), assuming an elementary step, one can write the following: r DE

1

    αF βF η ΓInt ðtÞ ¼ ke10 exp
η ΓEox ðtÞ
k
e10 exp RT RT |fflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflffl} ke1

ð32Þ

k
e1

Here, η is defined as η ¼ E
Eoe1, # , with E being the electrode potential, Eoe1, # , standard electrode potential of the step (Eq. 28) at the studied pH, and ki (i ¼ e1,
e1) electrochemical rate constants with units s
1. In a similar manner, the rate of the second electrochemical step (Eq. 29) can be formulated. The DET kinetics can be simulated using the BV approach; however, the values of some parameters, such as transfer coefficients (α,β), can deviate a lot from typical values in electrochemical kinetics. Transfer coefficient values of approximately 0.2 and below have been reported [6, 33]. The rate expression for the enzyme mediator step (Eq. 25), assuming an irreversible reaction and both enzyme and the mediator adsorbed at the surface, can be formulated as follows: r EM ¼ km ΓEox ðtÞ ΓMedred ðtÞ

ð33Þ

The electrochemical rate of the mediator regeneration (Eq. 26) can be described by the BV equation. In general, any of the biochemical or electrochemical steps in DET or MET mechanisms can be rate determining. For a DET mechanism with the steps in Eqs. 21, 22, 28 and 29, where electrochemical steps are considered to be

Electron Transfer Between Enzymes and Electrodes

51

irreversible for the sake of simplicity, one can show that under steady-state conditions the reciprocal value of the overall reaction rate (rDET_SS) is as follows: 1 r DET

SS

  1 1 cR ð0; SSÞ þ K M ¼ þ ΓE k e k2 cR ð0; SSÞ

ð34Þ

Here, ΓE is a total surface enzyme concentration (mol m
2geo), and ke and KM are overall constants of electrochemical (Eqs. 28 and 29) and biochemical (MichaelisMenten constant) (Eqs. 21 and 22) steps, defined as follows: ke1 ke2 ke1 þ ke2 k
1 þ k2 KM ¼ k1 ke ¼

ð35Þ ð36Þ

Equation 34 demonstrates the series connection between the “resistances” of the electrochemical and biochemical steps. The overall control of the reaction is influenced by the overpotential, showing in general electrochemical control at lower overpotentials and biochemical at higher overpotentials. Because the second term in Eq. 34 contains the reactant concentration, the possibility of a mass transfer limitation at higher overpotentials is given. If biochemical and mass transfer resistances are of the same order of magnitude, mixed-control conditions can be observed experimentally. This is nicely demonstrated in the example of a hydrogen reaction catalyzed by Desulfovibrio vulgaris Miyazaki F (DvMF) [NiFe]hydrogenase (Fig. 3). This reaction shows very fast kinetics, which quickly reach limiting current conditions. The limiting currents are more dependent on the concentration than the rotation rate, indicating predominant biochemical reaction control. In the direction of hydrogen evolution, no dependence on the rotation rate is observed. The small influence of the hydrogen concentration is probably caused by the reversibility of the hydrogen reaction. Based on Faraday’s law, in electrochemical systems, the reaction rate is proportional to the measured current. The same proportionality holds for the rate of the bioelectrochemical reaction and the measured current. For DET, the steady-state current density value is proportional to the steady-state reaction rate (Eq. 34) as follows: jDET

SS

¼ nFr DET

SS

ð37Þ

Here, n is the number of exchanged electrons (in the present example, n ¼ 2).

3.3

Balance Equations

The overall reaction rate (Eq. 34) is concentration dependent. To get the concentration profile of the reactant (product) mass, the balance equation has to be solved.

52

T. Vidakovic-Koch

Fig. 3 Influence of the rotation rate on steady-state polarization curves for a hydrogen reaction catalyzed by DvMF [NiFe]-hydrogenase at pH 7 and two different H2 concentrations (reference electrode: SCE, temperature 20 C, electrode surface area 0.0314 cm2). Unpublished results, courtesy of Dr. Olaf Rüdiger, Max Planck Institute for Chemical Energy Conversion

Assuming an electroenzymatic reaction at the flat electrode surface (Fig. 4a), it reads:   ∂ gk , α þ c α v k ∂cα ðz; tÞ ¼
∂t ∂zk

ð38Þ

where gk, α is a diffusion flux of the component α (mol m
2geo s
1), vk is the average molar velocity (m s
1), and zk (k ¼ 1,2,3) is the space coordinates. Assuming the concentration changes only in one zk direction, with a constant value of vk velocity and validity of Fick’s law, one obtains the following: 2

∂cα ðz; tÞ ∂ cα ðz; tÞ ∂cα ðz; tÞ ¼ Dα, H2 O
v ∂t ∂z2 ∂z

ð39Þ

Here, 5 Dα, H2 O is a binary diffusion coefficient of component α in water. Equation 39 is applicable for the description of concentration profiles of components reacting on a rotating disc electrode (RDE). This equation can be further simplified by neglecting the convective contribution (the last term in Eq. 39) to the concentration change. This is usually the case in the thin layer close to the electrode surface—the

Electron Transfer Between Enzymes and Electrodes

53

Fig. 4 (a) Schematic representation of a thin-film enzymatic electrode with (b) steady-state surface concentration of H2O2 and (c) geometrical current densities (calculated based on [33]) and (d) schematic representation of a porous enzymatic electrode with (e) steady-state H2O2 concentration profiles along the porous catalyst layer and (f) geometrical current density. Parts e and f are reprinted from [5] with permission from Elsevier

so-called Nernstian diffusion layer—where the velocity falls below 1% of its maximal value in the bulk and mass transport is mainly dependent on diffusion [34]. (A schematic representation of concentration profiles at time zero and infinitely long time are shown in Fig. 4a.)

54

T. Vidakovic-Koch

To solve this equation, the initial and boundary conditions are required. For RDE, they are as follows: z ¼ 0 Dα, H2 O

R X ∂cα ðz; tÞ ναm r m ðtÞ jz ¼ 0 ¼ ∂z m¼1

z ¼ δD, α cα ðδD, α ; tÞ ¼ cα, 1

ð40Þ ð41Þ

Here, R is the number of reaction steps where the component α is involved, ναm is the stoichiometric coefficient of component α in the respective step, rm(t) is the rate of the surface reaction (e.g., Eq. 30), and δD, α is the thickness of the diffusion layer of component α, defined as follows: δD, α ¼ 1:61ðDα, H2 O Þ1=3 ν1=6 w
1=2

ð42Þ

where, ν is the kinematic viscosity of the solution and ω is the electrode rotation rate. The solution of Eq. 39 gives a concentration profile over time and space coordinates. The calculated steady-state values of the substrate surface concentrations (cSubstrate(0,t)) as a function of applied electrode potential for three different bulk concentrations are shown in Fig. 4b. The surface concentrations deviate a bit from bulk concentrations, indicating the small influence of mass transport conditions. The presented profiles are calculated based on the model presented in [33], which describes the kinetics of hydrogen peroxide reduction by horseradish peroxidase on a flat enzyme-modified RDE. The experimental and calculated steady-state polarization curves [33] are shown in Fig. 4c. Most technical systems use porous electroenzymatic electrodes (Fig. 4d). In such cases, Eqs. 38 and 39 are valid for the mass transfer description in the solution. Depending on conditions, one can also expect concentration changes inside of the enzymatic electrode, in which case the following mass balance equation can be formulated [5, 35]: ε

2 R X ∂cα ðz; tÞ ∂ cα ðz; tÞ ¼ Dαeff þ a ναm r m , H2 O ∂t ∂z2 m¼1

ð43Þ

Here, ε is the volume fraction of the liquid phase inside of the enzymatic electrode porous structure (m3L/m3geo), Dαeff, H2 O , effective diffusivity and “a” (m2act =m3geom ) is the internal surface of the electroenzymatic electrode where the reactions (rm) take place per its geometrical volume. The calculated concentration profiles of H2O2 in a porous electroenzymatic electrode are shown in Fig. 4e [5]. The linear decrease of the concentration in the diffusion layer and the exponential decrease in the catalyst layer (CL) can be observed. The profiles are almost independent of potential; strong reactant depletion in the CL is seen at all potentials. For dynamic operation, in addition to mass balance, a charge balance equation has to be defined [5, 36]:

Electron Transfer Between Enzymes and Electrodes

cDL

R X ∂Eðz; tÞ ¼
ιðz; tÞ þ aF nm r m ∂t m¼1

55

ð44Þ

where, cDL is double layer capacity F m
2act, ι is the local current density (A m
2act), and nm is the number of exchanged electrons in the reaction step m. In a DET mechanism, the steps in Eqs. 28 and 29 will contribute electrons, while in the MET case step (Eq. 26) will contribute electrons. The electrode potential value (E(z, t)) is defined as follows: Eðz; tÞ ¼ ϕECL ðz; tÞ
ϕICL ðz; tÞ

ð45Þ

Here, ϕECL ðz; tÞ, ϕICL ðz; tÞ are potential field profiles in the electron and ion conducting phases respectively; the superscript CL refers to the catalyst layer. The potential profiles in electron and ion conducting phases follow from:   CL ∂ CL ∂ϕE ðz; tÞ γ
aιðz; tÞ 0¼ ∂z E ∂z   ∂ ∂ϕ CL ðz; tÞ γ ICL I 0¼
aιðz; tÞ ∂z ∂z

ð46Þ ð47Þ

Alternatively, it can be assumed that there is no potential distribution throughout the catalyst layer. In this case, the electrode potential, E(t), is only a function of time and is equal to the applied potential corrected for ohmic drop resistance. EðtÞ ¼ Eapl ðtÞ
RΩ jðtÞ

ð48Þ

Here, j is the global current density value (A m
2geo) and RΩ (Ω m2geo) is the electrolyte resistance. The global current density (experimentally measurable) can be obtained by integration of the local current values over the electrode thickness (in a one-dimensional system) according to the following [5, 36]: Z

L

jðtÞ ¼

aιðz; tÞdz

ð49Þ

0

where, L is the electrode thickness. The calculated and experimental current densities for the same example of a porous HRP electrode are shown in Fig. 4f [5].

4 Methods of Study Electrochemical methods are very convenient to study electron transfer between enzymes and electrodes. The main advantage of these methods over other methods of measurements (e.g. spectroscopic) is the possibility of direct measurement of the reaction rate in the form of electrical current (Eqs. 37 or 49). In this section, some

56

T. Vidakovic-Koch

basic electrochemical methods of measurement are demonstrated using selected examples with relevance to electrobiotechnological applications. For further reading, please consider [37, 38]. In general, an electrochemical cell and a device (normally a potentiostat/ galvanostat) that can monitor, apply, or control the potential or current input/output variables of an electroenzymatic system are needed to perform an electrochemical measurement. There are two types of electrochemical cells: two- and threeelectrode setups. Two-electrode cells are used in technical applications, whereas three-electrode setups are mainly used for research purposes. In the two-electrode cell, one electrode is the anode and the second electrode is the cathode. On the anode side, an oxidation reaction takes place, whereas reduction occurs on the cathode. The cell potential is always expressed as the difference between cathode and anode potentials. In a three-electrode setup, the electrodes are termed the working (WE), counter (CE), and reference electrodes (RE). The electrochemical reaction takes place on the WE, which is the subject of the study. This electrochemical reaction can be either oxidation or reduction, while the complementary reaction will take place on the CE. The third electrode, the RE, is used to monitor or set the potential of the WE. The potential difference between the WE and the RE is the electrode potential. Common reference electrodes are the saturated calomel electrode (SCE) and the silver/silver chloride electrode (Ag/AgCl). In the two-electrode setup, the current is measured between the cathode and anode; in the three-electrode setup, it is measured between the WE and CE (in both cases, they are termed cell currents). Electrode/cell potentials and cell currents can be set and measured with a potentiostat or galvanostat, which are fundamental tools in electrochemistry. A potentiostat or galvanostat can operate in a potentiostatic/voltastatic or galvanostatic mode of operation. In the potentiostatic mode, the potential difference between the WE and RE (electrode potential) is set, while the cell current is measured. In the voltastatic mode, the cell potential (the potential difference between the cathode and anode) is set, while the cell current is measured (Fig. 5a). During galvanostatic operations, the cell current is set, while electrode potential (EWE_RE) or cell potential (UC_A) for three- and two-electrode setups, respectively, are measured. The schematic representations of two- and threeelectrode setups and control and measured variables for voltastatic/potentiostatic modi of operation are shown in Fig. 5. As can be seen in Fig. 5b, the EWE_RE includes the ohmic drop in the solution between the tip of a so-called Luggin capillary and the WE electrode ( jcellRΩ). Some devices are able to compensate for the ohmic drop online with the help of positive feedback or a current-interrupt technique. This is especially important if significant currents flow through the cell at significant ohmic resistances. In electrobiotechnological applications, currents are usually low, but the ohmic resistances can be quite large because neutral and close-to-neutral solutions have lower conductivity due to low proton concentrations compared to strong acidic or alkaline solutions. It can be calculated that a 0.1 M phosphate buffer at pH 5 has an ohmic resistance of approximately 3 mΩ m2 [33].

Electron Transfer Between Enzymes and Electrodes

57

Fig. 5 Schematic representation of (a) two- and (b) three-electrode setups showing the control and measured variables in (a) voltastatic and (b) potentiostatic modi of operations

Compared to nonelectrical variables, current and electrode/cell potentials can be more easily modulated. Therefore, there is a broad range of electrochemical methods where electrical input variables are varied in a form of step, linear, or sinusoidal signals. The most common are stationary polarization, cyclic/linear sweep voltammetry, and electrochemical impedance spectroscopy.

4.1

Stationary Polarization

In stationary polarization, a step change of potential or current is applied as an input, while a transient change of the current or potential is recorded as an output. A record of current change over time is termed chronoamperometry, while a record of the potential change over time is termed chronopotentiometry. Figure 6 shows experimental data obtained by using a porous enzymatic electrode employing HRP as a WE in a three-electrode set-up in a potentiostatic mode of operation (EWE_RE is set, while Icell is measured). As can be seen, the current response contains both transient and steady-state parts. Although the transient part of the response can yield valuable information about the system, in stationary polarization only the steady-state part of the response is evaluated. By plotting steady-state potential versus steady-state current values, one can obtain a so-called polarization curve. These data are shown in Fig. 6b. The current points (Fig. 6b) measured after 60 and 120 s do not show significant differences, indicating steady-state conditions. Stationary polarization is also used for the characterization of electroenzymatic fuel cells or electrolyzers. In this case, the potential difference between cathode and anode (UC-A, cell potential) is set and the cell current is measured over time. It is common to present this data in the form of a cell potential versus cell current plot (Fig. 7a) (x-y axes are switched compared to a three-electrode set-up). The electrochemical behavior of three different electroenzymatic reactors operating under the same conditions (temperature, glucose concentration, flow rates) can be

58

T. Vidakovic-Koch

Fig. 6 Potentiostatic stationary polarization measurements on an example three-electrode setup. (a) Change of the input and response variables over time and (b) polarization curves. WE: HRP modified porous enzymatic electrode, CE: platinum wire, RE: SCE. Room temperature, 0.1 M phosphate buffer, pH ¼ 6. Adapted from [39]

Fig. 7 Potentiostatic stationary polarization measurements on an example of an enzymatic fuel cell (two-electrode set-up): (a) polarization curves, (b) electrode potentials and ohmic drop in an electrolyte, and (c) power density and efficiency curves. C: GOx-HRP porous enzymatic electrode, A: GOx/TTF porous enzymatic electrode, RE: SCE. Room temperature, 0.1 M phosphate buffer, pH ¼ 6, flow rate 10 mL min
1; (a) is reprinted from [40, 41]

Electron Transfer Between Enzymes and Electrodes

59

visualized in Fig. 7a. From these data, it is difficult to rationalize why the performance of Reactor 3 is better than of Reactor 1. The fuel cell or electrolyzer cell potential current characteristics can be described with the following equation: X     X Ucell ¼ Ec
EA
jRi ¼ EoC, # þ ηC
E0A, # þ ηA
ð50Þ jRi Equation (50) shows clearly that the overall performance is influenced by losses on the cathode and anode sides as well as ohmic losses. Providing that an RE can be inserted into a fuel cell or electrolyzer, it becomes possible to measure anode and cathode potentials during a fuel cell’s or electrolyzer’s operation. Depending on the position of the RE, the measured values will contain contributions of the ohmic resistance of the electrolyte between the electrodes. This contribution can be estimated based on the value of electrolyte resistance (RΩ) obtained either by calculations or measurements (e.g. from electrochemical impedance spectroscopy [EIS]; see below). The potentials of single electrodes as well as ohmic resistance under fuel cell conditions are shown in Fig. 7b. Compared to Fig. 7a, b offers some additional information. It follows that the contribution of the ohmic resistances in the electrolyte to observed losses is low. It is also clear that the cathode contributes more to potential losses than the anode (~600 mV on the cathode side compared to ~200 mV on the anode side are needed to achieve max current densities). The cathode in Reactor 3 can reach higher current densities than in the other two reactors, indicating a higher number of electrochemically active enzymes. The overall electrochemical performance is influenced by the enzymatic cathode. For fuel cell operation, it is also important to know which power density (P ¼ UC_A  Icell/Ageo) can be provided by the fuel cell. This information can be obtained from polarization curves by plotting P versus the cell current density ( jcell ¼ Icell/Ageo) (Fig. 7c). The power density current curve shows a characteristic maximum, which corresponds to the maximal efficiency of the fuel cell (Fig. 7c). The total fuel cell efficiency can be easily calculated as follows: ηtotal ¼

Δr Go U cell nF jcell Ageo ∙ ∙ Δr Ho Δr Go nFGGA, in |fflffl{zfflffl} |fflfflffl{zfflfflffl} |fflfflfflffl{zfflfflfflffl} ηth

ηec

ð51Þ

ηfuel

where GGA, in is flow rate of the glucose solution at the inlet, ηth is thermodynamic efficiency, ηec is electrochemical efficiency, and ηfuel is fuel utilization efficiency. The total efficiency in Fig. 7c is calculated by assuming a fuel cell surface area of 1 m2. Smaller surface areas will result in lower efficiency (~0.001% for 1 cm2 of surface area). As can be seen in Fig. 7c, the electrochemical efficiency is decreasing with an increase of the current density, while at the same time the fuel utilization is increasing. The tradeoff between these two trends determines the overall efficiency.

60

4.2

T. Vidakovic-Koch

Cyclic Voltammetry

In cyclic voltammetry (CV), cell potential (UC-A) or electrode potential (EWE-RE) as an input is changed between the starting and maximum levels at a constant sweep rate (Fig. 8a) and cell current is recorded as a response (Fig. 8b). The common sweep rate values are between 10
3 and 103 V s
1. The cycle shown in Fig. 8a can be repeated n times; usually, the potential is cycled until the current stops changing. CV is more often used for characterization of WE in a three-electrode setup, but it can be used also for two-electrode setup characterizations. In a three-electrode setup, the potential difference EWE-RE is set. This potential difference also contains the contribution of the ohmic drop in the electrolyte. Because the electrochemical reaction is driven only by the ohmic drop’s free potential difference, it is important to correct the ohmic drop’s contribution during a CV experiment. The reason can be easily seen in Fig. 8a. The red line shows the time change of the ohmic drop’s corrected EWE_RE potential, which is calculated assuming an electrolyte resistance of 3 mΩ m2; deviations from the expected linear change of the potential can be observed, dEWE_RE/dt is no longer constant, which is a main assumption of CV measurements. The data in a CV experiment are normally presented in the form of current-potential characteristics, which is called cyclic voltammogram (Fig. 8c). The yellow rectangle in Fig. 8c corresponds to the behavior of a so-called ideally polarizable electrode showing only double-layer charging (no Faradaic reaction). The measured current then corresponds to pure capacitive current, where in this case holds as follows: I WE
CE ¼ I capacitive ¼ cDL ∙ v Ageo

ð52Þ

with v being the sweep rate (V s
1). The behavior of a real porous enzymatic electrode shows some deviations compared to the ideally polarizable electrode, especially at more positive and more negative potentials. In the potential region from approximately 0.2 to 0.4 V, the behavior corresponds almost to the expected behavior of a pure capacitor and can be used for the estimation of double-layer capacitance. Cyclic voltammetry experiments are fast and simple to perform. They provide good orientation on system properties, such as the potential region where the reaction takes place, and they provide an orientation on rate-determining steps. Additionally, CV can be also used for estimation of the enzyme surface coverage (enzyme total surface concentration). The surface coverage of most DET enzymes at electrode surfaces is very low, for which reason redox peaks showing direct electron exchange between an enzyme and an electrode cannot always be observed. To exemplify the surface coverage determination, determination of covalently adsorbed PQQ mediator coverage on a cysteamine-modified gold electrode is chosen (Fig. 8d) [42]. The well-defined redox peaks can be observed in Fig. 8d. The areas under these peaks are integrated using the straight lines as a background. The values of 2.42 and 2.48 μA V were estimated for the anodic and cathodic peaks,

Electron Transfer Between Enzymes and Electrodes

61

Fig. 8 Cyclic voltammetry measurements on an example of a three-electrode setup. (a) Input EWE_RE potential change over time. (b) Response cell current change over time. (c) Cyclic voltammetry of a porous HRP enzymatic electrode without Faradaic reaction, 20 mV s
1. (d) Cyclic voltammogram of an electrode showing a Faradaic reaction related to the surface adsorbed species (PQQ modified Au electrode at pH 7.2, 20 mV s
1). (e) Comparison between stationary polarization and CV in a DET mechanism at pH 6 and room temperature. (f) Comparison between stationary polarization and CV (5 mV s
1) in a MET mechanism. The blue line corresponds to CV (5 mV s
1) of a tetrathiafulvalene (TTF) dispersed in Nafion film modified graphite electrode, pH ¼ 7.2, 37 C. Further conditions: 0.1 M phosphate buffer. (d) is reprinted from [42] with permission from the author, (e) is reprinted from [39]. (f) is reprinted from [43] with permission from Elsevier

respectively. These areas are divided by sweep rate, number of exchanged electrons, electrode geometric surface area, and the Faraday constant. The PQQ coverage of 1.33  10
9 mol cm
2geo and 1.37  10
9 mol cm
2geo is calculated. The roughness of the gold electrode is estimated to be approximately 25.

62

T. Vidakovic-Koch

If the reactant is added to the reaction mixture, the Faradaic current will add up to the capacitive current. Subtraction of the capacitive currents measured without the reactant in the reaction mixture from the total current in presence of reactant will roughly result in the Faradaic current of the reaction. Although the capacitive current is dependent on the sweep rate, the activation current should be independent. This holds only if there is no mass transfer limitation in the system. If the latter is the case, the Faradaic current will also depend on the sweep rate because the concentration profiles throughout the catalyst layers will require some time to establish. Therefore, the currents obtained by subtraction of the capacitive contribution will still be sweep rate dependent; in the limiting case of very slow sweep rates, they will approximate the steady-state currents. This is demonstrated in Fig. 8d, where the currents obtained by the subtraction of CV currents and steady-state responses are compared. The same can be followed in an example of MET (Fig. 8f). Tetrathiafulvalene (TTF) was used as a mediator, which starts to “dissolve” in the absence of glucose at more positive potentials (black curve in Fig. 8f). If the direction of the potential is reversed, a small reduction peak can be observed. This behavior can be attributed to the solubility of TTF+ formed in an aqueous solution. If the TTF is captured close to the surface, such as in a Nafion polymer (blue line), almost symmetrical oxidation/reduction peaks can be observed (Fig. 8f, blue line). One of disadvantages of CV is the need for a complicated mathematical description in multistage processes.

4.3

Electrochemical Impedance Spectroscopy

In EIS, the input signal is a potential (EWE-RE or UC-A) or current (Icell), which is applied in the form of periodic cosine (sine) perturbation around a steady-state potential/current value. For the cosine perturbation, assuming a linear response, it will follow that: Input : ΔEWE
RE ðtÞ ¼ EWE
RE ðtÞ
EWE
RE, SS ¼ A cos ðωtÞ  A  iωt e þ e
iωt ¼ 2 Output : ΔI cell ðtÞ ¼ I cell ðtÞ
I cell, SS  A Y ðωÞeiωt þ Y ð
ωÞe
iωt ¼ A ∙ jY j ∙ cos ðωt þ φÞ ¼ 2

ð53Þ

ð54Þ

Here, Y(ω) is a linear frequency response function (FRF), which corresponds to electrochemical admittance (reciprocal of electrochemical impedance (Z
1(ω)), |Y| is an amplitude gain, φ is a phase shift, and ω is an angular frequency in

Electron Transfer Between Enzymes and Electrodes

63

rad s
1. Y(ω) is defined as the ratio between output and input signals. This, it will follow from Eqs. 53 and 54 that: Y ðωÞ ¼ jY j ∙ eiφ

ð55Þ

Similarly, one can write for the electrochemical impedance: ZðωÞ ¼ jZ j ∙ eiφ ¼ ReðZ Þ þ iImðZ Þ

ð56Þ

Keeping in mind that the electrochemical impedance is a complex number, it is common to plot its values as a Nyquist plot (
Im(Z ) over Re(Z )) or Bode plot (log| Z| and φ) over the log of frequency. Usually in these plots, the frequency ( f ) in s
1 (Hz) is used instead of the angular frequency ω, where: ω ¼ 2πf

ð57Þ

The frequencies of the input are changed between millihertz and megahertz. The whole impedance spectrum covers the whole range of frequencies. Impedance spectroscopy was originally invented for the analysis of electrical circuits. Most electrochemical systems behave similarly to such equivalent circuits. The simplest example is an ideally polarizable electrode, which in terms of electrical circuit can be represented as a serial connection of an electrolyte resistance and a capacitance (Fig. 9a). In a more general case, an electrochemical reaction takes place at the electrode surface, causing a charge transfer resistance. For electroenzymatic reactions, the charge transfer resistance (resistance of electrochemical reaction) is accompanied by the resistance of enzymatic reactions. For example, in the case of DET (Eq. 34), these two resistances are connected in serial. The overall resistance of the electrochemical and biochemical reactions is denoted as RR in Fig. 9b. Additionally, the mass transfer of reactants or products contributes an additional resistance (RMT). The reaction and mass transfer resistances are connected in parallel with double-layer capacitance. Furthermore, there is an ohmic resistance of the electrolyte, which is related to the ohmic drop in the electrolyte between the tip of the Luggin capillary and the WE.

Fig. 9 Equivalent circuits of (a) an ideally polarizable electrode (only electrolyte resistance and double-layer capacitance) and (b) a general electrochemical system showing electrolyte, reaction, and mass transfer resistance

64

T. Vidakovic-Koch

The total impedance of the ideally polarizable electrode can be calculated as follows: Z¼

1
RΩ |{z} iωcDL |fflffl{zfflffl} Impedance of Impedance of electrolyte double layer

ð58Þ

Here, it can be easily seen that: ReðZÞ ¼ RΩ and ImðZ Þ ¼


1 ωcDL

ð59Þ

The magnitude and the phase shift can be calculated accordingly: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ReðZÞ2 þ ImðZÞ2 ImðZÞ φ ¼ a tan ReðZÞ

jZ j ¼

ð60Þ ð61Þ

The impedance of an ideally polarizable electrode is calculated based on Eqs. 58, 59, 60 and 61, with the double-layer capacitance value estimated from the CV of a porous enzymatic electrode (Fig. 8c) and the electrolyte resistance value based on a typical electrolyte resistance in a three-electrode set-up at pH 5 (~3 mΩ m2) [33]. The calculated data are shown in Fig. 10 as Bode and Nyquist plots. As can be seen in Fig. 10a, at high frequencies, the behavior is dominated by the electrolyte resistance (the phase shift is 0 and the magnitude value corresponds to the assumed resistance value); at low frequencies, the behavior is dominated by the capacitive behavior, with a phase shift of
90 being typical for the capacitor. In case of an electroenzymatic reaction, in the absence of mass transfer resistance, the overall impedance will be as follows: Z¼

þ RΩ |{z} Impedance of electrolyte

RR 1
iRR ωcDL |fflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflffl} Impedance of double layer and electroenzymatic reaction in parallel

ð62Þ

where: RR 1 þ R2R ω2 c2DL RR ωcDL ImðZÞ ¼
1 þ R2R ω2 c2DL

ReðZ Þ ¼ RΩ þ

ð63Þ ð64Þ

Electron Transfer Between Enzymes and Electrodes

65

Fig. 10 Calculated (a) magnitude and (b) phase shift of impedance of an enzymatic electrode without and with DET Faradaic reaction at E ¼ 0.744 V; experimentally determined (c) magnitude and (d) phase shift of hydrogen peroxide reduction on an HRP-modified graphite RDE at E ¼ 0.744 V; and calculated Nyquist plots using simplified Eqs. 63, 64 and 69 (e) and (f) a more rigorous overall impedance expression [33]. Conditions in (e) and (f): pH 5, 0.08 mM H2O2. (f) is reprinted from [33] with permission from Elsevier

The reaction impedance RR can be generally expressed as:  RR ¼

∂j ∂E


1

ð65Þ

66

T. Vidakovic-Koch

For a simple electrochemical reaction: ox þ e
¼ red

ð66Þ

     αF ð1
αÞF o o ðE
E Þ
exp
ðE
E Þ j ¼ jo exp RT RT

ð67Þ

with:

Assuming zero overpotential (E
Eo ¼ 0), it can be easily shown that  RR ¼

F jo RT


1

ð68Þ

For an electroenzymatic process, reaction impedance can be determined in a similar way from Eq. 37: RR ¼

2 2 αF RTnFΓE k 2 cR k DE

ðk2 cR þ kDE ðK M þ cR ÞÞ2

!
1 ð69Þ

Because Eq. 34 is derived by applying steady-state approximation, the expression in Eq. 69 will be strictly valid only for low frequencies (ω!0) and in the absence of mass transfer limitations. Still, for the sake of demonstration, it was used for calculation of the overall impedance of DET enzymatic electrode. The results in Fig. 10a and b show influence of the electroenzymatic reaction only in the low frequency range. The calculated data show good qualitative agreement with experimental data in Fig. 10c and d. The response of DET electroenzymatic electrodes in the Nyquist plot (Fig. 10e) shows the typical semicircle behavior that is characteristic for electrodes with a Faradaic reaction in the absence of mass transfer limitations. The more rigorously calculated DET impedance, based on the dynamic response of the DET electrode and taking into account mass transfer limitations, together with experimental data, are shown in Fig. 10f. This dependence is more complicated and involves constants of transient steps, which are not obvious in the simplified expression (Eq. 69) (for details, please see [33]). Therefore it has greater sensitivity for rate constants of steps, which show faster dynamics and can be used for their determination. In this respect, EIS shows higher sensitivity for model discrimination compared to stationary polarization. It could have been shown that only the model with Michaelis Menten kinetics can describe experimental data for all range of experimental conditions [33].

4.4

Parameter Determination

In this section, the parameter determination based on discussed electrochemical methods is described. The main parameters of concern are kinetic rate constants

Electron Transfer Between Enzymes and Electrodes

67

(enzymatic and electrochemical) and enzyme surface concentration. An additional parameter that appears to be important is the surface-to-volume area a, which accounts for the electrochemically active surface area in porous electrodes.

4.4.1

Kinetic Rate Constant Determination

The expression in Eq. 34 gives us a way to determine kinetic rate constants based on current potential measurements (e.g. from steady-state data for HRP-catalyzed hydrogen peroxide reduction at different concentrations, as shown in Fig. 11a). By replacing the reaction rate in Eq. 34 with the current density value divided by the number of electrons and Faraday’s constant, it follows that: nF jDET

¼ SS

  1 1 cR ð0; SSÞ þ K M þ ΓE k e k2 cR ð0; SSÞ

ð70Þ

Fig. 11 Determination of kinetic rate constants from stationary polarization measurements for an example of hydrogen peroxide reduction on HRP-modified graphite RDE. (a) Polarization curves for different hydrogen peroxide concentrations, (b) double reciprocal plot based on Eq. 70, and (c) nonlinear fit of intercept values from double reciprocal plot versus steady-state potential values. Conditions: pH 5, 0.1 M phosphate buffer, fixed delay 1 min, room temperature, rotation rate 400 rpm. (a) is based on data in [33]

68

T. Vidakovic-Koch

By plotting the j nF (at Ess ¼ const) against the reciprocal value of the reactant DET SS concentration (c1R), one should obtain a family of straight lines (Fig. 11b), with slope h i equal to Γ1E KkM2 and intercepts to Γ1E k1e þ k12 . Providing that the total enzyme surface concentration is known, one can easily obtain the KkM2 ratio from the calculated slope value. Alternatively, by adopting the KkM2 value based on homogeneous enzymatic catalysis, one can determine the total enzyme surface concentration. The values of intercept can be further plotted against the steady-state electrode potential values (Fig. 11c) because the keconstant contains the potential dependence (  αF ke ¼ ke0 exp
RT ESS
Eo, # . As expected, an exponential dependence is obtained. For the determination of kinetic parameters, nonlinear fit can be used (the nonlinear fit function and corresponding parameter values based on experimental data from Fig. 11a can be seen in Fig. 11c). In this way, providing again that the total enzyme concentration value is known, the rate constants of the electrochemical step (ke0 and α) and enzymatic step (k2) can be obtained based on rate expression (Eq. 70). Another strategy for the determination of rate constants (and other parameters) is based on global optimization of a large set of experimental data obtained under different operating conditions (concentrations, temperature, pH) and possibly by using different methods of measurements (steady-state and dynamic measurements; e.g. EIS). An objective function can be defined using the least square or weighted least square method (this should be the case when experimental data are not equally reliable). In the least square method, the sum of squared residuals is minimized, where the residuals are differences between the experimental values and the fitted values. In the weighted least square method, squared residuals are multiplied by corresponding weights. An example of the weighted least square objective function is [5]: m X n

X   2   jsim cR, j ; Ek ; p
jexp cR, j ; Ek wjk

ð71Þ

j¼1 k¼1

where p stands for all parameters that have to be determined (e.g. in Eq. 70, ΓE, ke0, KM, k2), wjk are weights, and subscripts “sim” and “exp” correspond to simulated and experimental data, respectively. The comparison of model parameters obtained by the linearization method, followed by fitting of the exponential curve (Fig. 11a–c) and global optimization by using weighted least square objective function on the same set of experimental data, are shown in Table 4. As can be seen in Table 4, the linearization approach gives similar values of model parameters. Some deviations can be ascribed to shortcomings of linearization by using the double reciprocal plot as well as neglecting the mass transfer limitations of the reactant. The latter effect has been taken into account into a model [33] that was used for calculation of the simulated data in a global optimization approach. The total enzyme surface concentration was a fitting parameter in the

Electron Transfer Between Enzymes and Electrodes

69

Table 4 Comparison of parameter values obtained using two different approaches Parameter ( p) α k2, s
1 KM, mol m
3 ke0, s
1 ΓE, 10
9 mol m
2

Approach 1 (Fig. 13a–c) 0.21 1.299 0.14 33.72 2.20a

Approach 2: Global optimization with objective function Eq. (71) [33] 0.17 1.898 0.21 37.80 2.20

a

Adopted value

global optimization method. As a starting value, the total enzyme surface concentration of 1.2 nmol m
2 was used. This value was obtained from the slope of the linearized rate expression (Eq. 70), by adopting a Kk2M value of 1.7  104 (mol m
3)
1 s
1 for the enzyme in the solution, as suggested by Andreu et al. [6]. The total enzyme surface concentration in Approach 1 (Table 4) was assigned the same as in Approach 2.

4.4.2

Determination of Enzyme Surface Concentration

Obviously, knowledge of the total enzyme concentration value is essential for the determination of rate constants of both enzymatic and electrochemical steps. In some cases, the redox process at the enzyme redox site(s) can be evidenced electrochemically; therefore, the determination of the total enzyme surface concentration under electrochemical conditions appears possible. Typically, a cyclic voltammetry experiment is utilized in absence of reactive species and is called noncatalytic (nonturnover) cyclic voltammetry. In such an experiment, the redox center of the enzyme is oxidized/reduced electrochemically during the potential sweep, showing corresponding redox peaks in the cyclic voltammogram. As an example, a cyclic voltammogram of cytochrome c peroxidase modified pyrolithic graphite electrode is shown in Fig. 12 [44, 45]. As can be seen in Fig. 12, an almost reversible redox peak at a potential assigned to the reduction potential of cytochrome c peroxidase compound I appears. At low enzyme surface coverage, this redox peak can be masked by the electrode capacitive contributions, which might necessitate a background current subtraction (e.g. in [46]). The enzyme’s total concentration can be calculated from the charge under the redox peak, by assuming a number of exchanged electrons in accordance to the following [33]: R E2 ΓE ¼

E1

j ∙ dE

nF ∙ v

ð72Þ

70

T. Vidakovic-Koch

Fig. 12 Cyclic voltammogram of cytochrome c peroxidase modified pyrolithic graphite electrode in 20 mM phosphate buffer, pH 6.1, 4 C. Reprinted from [33] with permission from Springer Verlag

Here, E1 and E2 are potential limits of cyclic voltammetry experiment, n is the number of exchanged electrons, and v is the sweep rate in V s
1. The total enzyme surface concentration of cytochrome c peroxidase on modified pyrolithic graphite electrode, assuming the 2e
process, was determined to be 6.2 pmol cm
2. Usually, there is no evidence of enzyme redox activity on the electrode surface in the absence of a reactive species. In such a case, nonturnover voltammetry is not helpful for the total enzyme surface concentration determination. Also, as demonstrated in an example of stationary polarization, total enzyme surface concentration cannot be directly obtained from the experimental polarization curves. The problem can be circumvented by the combination of stationary polarization and EIS measurements. As it follows from Eq. 69, the real part of the EIS at low frequencies (which corresponds to the diameter of the semicircle) depends on kinetic rate constants and total enzyme surface concentration. Therefore, the combination of

Electron Transfer Between Enzymes and Electrodes

71

these two methods (or other electrochemical methods) can enable “direct” determination of the enzyme surface concentration. An estimation of the total enzyme surface concentration can be obtained by assuming the same kinetic activities as for enzymes under homogenous conditions, where the well-established values exist in literature. This approach was discussed by Andreu et al. [6]. They obtained ΓE values in the range from 0.09 to 0.14 pmol cm
2 for pH range from 4.1 to 8 and for HRP, by adopting Kk2M value of 1.7  104 (mol m
3)
1 s
1 for enzymes in the solution. The authors commented that the obtained values represent lower limits of the total enzyme surface concentrations because the activity of immobilized enzymes is probably reduced compared to the homogenous case of enzymes in the solution. These values are only approximately 1% of the calculated value of a fully packed enzyme monolayer of approximately 10 pmol cm
2 based on enzyme crystallographic data [6]. The authors concluded that the real value of the total enzyme surface concentration is somewhere between approximately 0.1 and 10 pmol cm
2, where the values closer to the lower limit are supported by the absence of nonturnover activity in cyclic voltammetry. Other values reported in the literature for this enzyme are in the range from 0.1 to 40 pmol cm
2 [6, 47, 48].

4.4.3

Other Parameters

Another parameter of interest is the surface-to-volume area a, which appears important in porous enzymatic electrodes. In DET, the parameter a depends on the internal surface area of electron conductive support in contact with an enzyme. This can be estimated based on electron conductive material loading and BET surface area, as well as the thickness of the porous electrode. Carbon nanomaterials, which are often employed as electroconductive supports, have BET surface areas in the range between 200 and 600 m2 g
1. For Vulcan XC 72, the BET surface area is 250 m2 g
1. The addition of binder materials, which are necessary for porous electrodes, might reduce this area significantly. In fuel cell research, a BET area reduction up to ten times, depending on binder-to-carbon ratios, was observed [49]. Assuming a similar effect of the binder in the present case and Vulcan XC 72 loading of 3.0 mg cm
2 and 53-μm [5] electrode thickness, one can calculate the internal surface area of 1.4  107 m2act/m3geo. Assuming further similar coverage as in the case of the flat electrode (~1% of the full monolayer) and molecular weight of HRP of 44 kDa, one can calculate the enzyme loading of such an electrode of approximately 33 mg cm
2geo. For comparison, the value that was obtained in [5] by using model-based analysis was approximately 6 mg cm
2geo. The difference can be a consequence of overestimation of the internal surface area of the carbon support. However, it may also indicate that even lower enzyme coverage than on the flat surface is obtained in porous electrodes.

72

T. Vidakovic-Koch

5 Toward the Development of Electrobiotechnological Processes This section reviews some recent advances in the preparation of enzyme-modified electrodes. This mainly concerns research activities where the enzyme co-factor is an inherent part of the enzyme. Additionally, some challenges in the electrochemical regeneration of soluble co-factors (co-substrates; e.g. NAD, NAD(P)) are discussed.

5.1

Development of Enzyme-Modified Electrodes

A technical enzymatic electrode can be defined as a composite structure containing different layers such as a catalyst layer and/or diffusion layer and current collectors. Thereby, enzymatic catalysts are integrated into the catalyst layer by immobilization or by entrapment behind, for example, a dialysis membrane. Although the latter approach is simpler [50], the performance of such electrodes is lower (~50 μA cm
2 [50] compared to ~700 μA cm
2 [51] for the case of enzymatic electrodes employing hydrogenases). In Fig. 13a, a schematic representation of an enzymatic fuel cell containing two porous enzymatic electrodes is shown. The anode comprises a porous catalyst layer containing electron conductive nanoparticles, a mediator, binder, and the enzyme GOx (Fig. 13c) [40]. Stainless steel was used as a current collector. The cathode is a so-called gas diffusion electrode, with a catalyst layer applied on the gas diffusion layer. The graphite flow field serves as a current collector. The catalyst layer on the cathode side contains electron conductive nanoparticles, binder, and enzymes. Both anode and cathode catalyst layers contain porous electron conductive matrixes, which are formed out of carbon, gold nanoparticles, or polymer materials [51–54]. Most recent studies have concentrated on the use of carbon nanomaterials [54]. To form porous structures, carbon nanoparticles have to be pressed in pallets or bonded using a suitable binder. Such binder materials include polyvinylidene fluoride (PVDF) and gelatin [2]. PVDF is a highly nonreactive fluoropolymer that is chemically inert in the presence of different solvents, acids, and bases, while gelatin is a biocompatible natural polymer. In both cases, the binder is mixed with electron conductive nanomaterial and a mediator (for MET electrodes only) and left to dry. This procedure is applicable for immobilized mediators supplied as nanomaterials or polymers, as well as for dissolved mediators. The choice of mediator depends on the enzyme cofactor, as well as on the purpose of the electroenzymatic system. For example, soluble mediators might complicate downstream processing of the reaction mixture; they are less preferred for implantable devices (e.g. enzymatic fuel cells for in vivo operation) [2]. In combination with GOx (FAD cofactor), different mediators have been used. Some examples are ferrocene [55], tetrathiafulvalene (TTF) [35, 43], 8-hydroxyquinoline-5-sulfonic

Electron Transfer Between Enzymes and Electrodes

73

Fig. 13 (a) A schematic representation of an enzymatic fuel cell, (b) legend, (c) structure of the catalyst layer on the anode side, and (d) a catalyst layer structure on the cathode side, and SEM cross-sections of enzymatic electrodes with (e) Vulcan/Gelatin binder and (b) Vulcan/PVDF binder. (a,b,c) were reprinted from [40], (e and f) were reprinted from [39]

74

T. Vidakovic-Koch

acid (HQS) [24], and Os redox hydrogels [19]. Beyond these mediators, Os hydrogels with redox centers attached to a polymer backbone exhibited the best performance. Their disadvantages include complicated synthesis procedures and toxicity issues associated with Os. A TTF mediator has been used either alone [26, 35, 40] or in combination with tetracyanoquinodimethanide (TCNQ), forming a charge transfer complex (CTC) that then acts as a mediator [43, 56–58]. Although CTC is electronically conductive, TTF has negligible conductivity, which is the reason why it is usually integrated into the electron conductive matrix. CTC has shown excellent performances in biosensor electrodes, high oxygen tolerance, and remarkable stability under continuous operation [56]. A further advantage is that enzymatic electrodes based on CTC do not require complicated modification procedures; they can be prepared just by mixing the respective components [59]. CTC is commercially available and has high electronic conductivity, which is beneficial for lowering the ohmic resistance within the electrode layer. The morphology of the CTC crystals can be tuned by variation of the experimental conditions [59]. They can be also prepared as nanoparticles. These strategies can be applied to tune the catalytic properties of the CTC and/or to increase the catalytically active surface area. Both TTF and TCNQ have low toxicity, which is attributed to their low solubility in water and physiological fluids. Although CTC can be employed without an electron conductive material due to its own electron conductivity, the addition of carbon nanoparticles improves the electrode performance [43]. The enzymes can be immobilized into the porous matrix together with the binder or separately. The first possibility is convenient when gelatin is used as a binder, whereas the second one is preferred when PVDF is used as a binder. (PVDF has to be dissolved in an organic solvent before mixing it with electron conductive nanoparticles and a mediator, which might not be compatible with enzymes; additionally, during drying of the porous matrix, the enzyme might denature.) In both described cases, the enzyme immobilization is physical. If enzymes are physically adsorbed in a PVDF bonded carbon nanoparticle porous matrix, only a weak interaction between a support and an enzyme is assumed. In this way, enzymes can be “wired” to different electron conducting materials (e.g. gold and carbons). Over the long term, this method of immobilization shows some loss of stability due to enzyme leaching. If the enzymes are physically immobilized by entrapment into a gel matrix formed by gelatin (collagen and polysaccharides can also be used), the stability of such electrodes is improved; however, the enzyme activity might be influenced. Because gelatin is soluble in water, to assure the mechanical stability of the catalyst layer, it has to be cross-linked with glutaraldehyde. This step improves the stability but reduces the activity of enzymes. Stability can be also improved by the covalent attachment of enzymes to the electrode surface. In this case, the electrode has to be functionalized with surface groups favoring covalent attachment with surface groups of enzymes. The enzyme immobilization procedure has significant influence on activity of porous enzymatic electrodes. The comparison of two immobilization procedures defined by use of two binders (PVDF and gelatin) in an example of DET and MET enzymes is shown in Fig. 14. For the DET enzyme, a PVDF procedure where

Electron Transfer Between Enzymes and Electrodes

75

Fig. 14 Influence of the immobilization procedure (binder) on activities of enzymes showing (a) DET mechanism (black lines ¼ gelatin binder, blue line ¼ PVDF binder; reprinted from [5] with permission from Elsevier) (b) MET mechanism (PVDF binder; reprinted from [35] with permission from Elsevier) and (c) MET mechanism (gelatin binder) at two different thicknesses of the porous electrodes (reprinted from [60] with permission from the author); (d) influence of enzyme loading on normalized activity of MET electrode and simulated concentration profiles along the spatial coordinate at two different enzyme loadings and 0.4 V (e) low enzyme loading (0.1 mg mL
1); (f) high enzyme loading (14 mg mL
1; reprinted from [35] with permission from Elsevier). Conditions: (a) H2O2 reduction on gelatin or PVDF/Vulcan/HRP enzymatic electrodes, pH 5, room temperature, stationary polarization (b) glucose oxidation on PVDF/Vulcan/GOx, and (c) on gelatin/Vulcan/GOx enzymatic electrodes, pH 7, 37 C, stationary polarization

enzymes are only physically adsorbed is more beneficial in the whole range of studied overpotentials (Fig. 14a). For MET enzymes, the limiting current region is almost independent on the electrode preparation procedure, while the kinetic region

76

T. Vidakovic-Koch

is significantly influenced. Additionally, the kinetic region is dependent on the catalyst layer thickness (Fig. 14c). Comparing the performances of thick MET electrodes prepared using PVDF (59-μm thickness) and gelatin (65-μm thickness) binders, the gelatin electrode has better performance in the kinetic region (at lower overpotentials) [35]. The performance of a thick PVDF electrode (59-μm thickness) is similar to the performance of a thin gelatin electrode (19-μm thickness). Based on modeling results, the PVDF procedure “offers” more active DET enzymes than the gelatin procedure [5, 35]. This can be rationalized by enzyme agglomeration in the presence of gelatin and the cross-linking agent glutaraldehyde [39]. The formation of enzyme agglomerates decreases the number of DET enzymes in contact with the electrode surface, leading to lower activity. An additional reason for the observed differences is mass transport limitation. Based on modeling results, a DET gelatin electrode is less limited by mass transport than a DET PVDF electrode due to the lower number of active enzymes. In MET electrodes, the differences in the kinetic region can be caused by differences in either mediator or enzyme concentrations. According to modeling results, a change in mediator concentration causes an almost parallel shift of the polarization curve on the potential scale, while a change in enzyme concentration leads to branching of the polarization curves [35]. The former effect describes the behavior of MET electrodes with respect to the preparation procedure and electrode thickness. The results indicate a lower concentration of the mediator in the MET PVDF electrode compared to an MET gelatin electrode of similar thickness. At more positive overpotentials, the activity depends on the enzyme concentration and mass transfer limitations. Although the thick MET gelatin electrode has a larger number of enzymes, it is more limited by mass transport; this leads to a lower level of enzyme utilization and finally to a similar performance compared to the thin electrode at more positive overpotentials. With an increase of enzyme loading, the normalized electrode activity is decreasing (Fig. 14d). One of reasons is the depletion of reactants along the catalyst layers (Fig. 14e, f) [35].

5.2

Electrochemical Regeneration of Soluble Co-Factors

Many enzymes of technological interest are dependent on soluble co-factors, such as nicotinamide adenine dinucleotide (NAD) and nicotinamide adenine dinucleotide phosphate (NADP) in their oxidized (NAD+, NADP+) or reduced forms (NADH, NADPH) [61]. The regeneration of these co-factors is of paramount interest with respect to the technical applications of these enzymes. In a most favorable case, their electrochemical regeneration occurs at low overpotentials, showing fast kinetics and high selectivity towards enzymatically active forms of these co-factors as well as high current efficiency. In reality, these conditions are not easily satisfied. The equilibrium electrode potential of the NAD+/NADH reaction (
0.32 V SHE) is close to the equilibrium electrode potential of the hydrogen reaction 2H+/H2 (
0.414 V SHE). Considering the fast kinetics of the

Electron Transfer Between Enzymes and Electrodes

77

hydrogen reaction on most electrode materials, it can be expected that the current efficiency for NAD+(NADP+) reduction will be significantly lowered. It was also reported that the direct reduction of NADP+ leads to one electron charge transfer and formation of a radical NADP form, followed by dimerization to an enzymatically inactive dimer form [62]. Similarly, the direct electrochemical oxidation of NADH takes place only at very high overpotentials; it is further complicated by side reactions, such as the direct anodic oxidation of reactants and electrode fouling [61]. More promising than the direct electrochemical regeneration is a mediated regeneration. A number of different mediators has been reported so far (see e.g. [33] and references therein). Most of these mediators support electrochemical regeneration of either oxidized or reduced forms of soluble co-factors. However, in few cases, electrocatalytic systems were able to regenerate both oxidized and reduced forms [33]. For example, a pyrolytic graphite edge electrode modified by a hydrophilic domain of mitochondrial NADH ubiqinone oxidoreductase was able to regenerate both forms of NAD co-factor at low overpotentials. The drawbacks of this approach were low stability and current densities. Similar findings were reported for electropolymerized neutral red [63]. In this first report, no enzymatic reaction was coupled with the regeneration system. In a later report by Arechederra et al. [64], the same polymer was used for NAD+ and NADH regenerations in a rechargeable ethanol-fueled biobattery. Other reported mediator modified electrodes support regeneration of either an oxidized or reduced form of a co-factor. Poly(methylene green) was reported as a highly efficient mediator for NADH oxidation by several groups [61, 65–67]. Poly (methylene green) modified carbon nanotubes showed high electrocatalytic activity for NADH oxidation, reaching current densities up to 5 mA cm
2 at overpotentials of approximately 0.5 V [66]. Kochius et al. [61] studied different mediators for NADH oxidation. As a most promising candidate, 2,20 -azino-bis-(3 ethyl-benzothiazoline-6-sulfonic acid) (ABTS) was identified. High substrate conversions and high space time yields were achieved. The ABTS also showed remarkable stability during prolonged cycling. Drawbacks of this mediator include the high overpotential required for the NADH oxidation as well as problems related to its separation out of the reaction mixture. NAD+ and NADP+ were reduced to NADH and NADPH using (pentamethylcyclopentadienyl- 2,2V-bipyridine aqua) rhodium (III) as a mediator [68]. The regeneration was performed in a packed bad reactor with 100% selectivity and conversion. A high space-time yield of 500–1,000 g dm
3 day
1 was achieved. The authors also suggested downstream processing (including ion exchange) to remove the mediator, nanofiltration to remove the buffer, and finally freeze-drying of the reduced nicotinamide coenzyme. Composite electrodes containing a Ru(III) complex and single-walled carbon nanotubes were reported to be active in the regeneration of electrochemically active 1,4 NADH [69] Also, it was shown that traditional carbon electrodes under certain conditions (high overpotentials) almost exclusively (98% yield) formed enzymatically active 1,4-NADH forms [70].

78

T. Vidakovic-Koch

Catalytic systems requiring NADH (NADPH) co-factor and additional substrates such as oxygen are especially challenging. This is the case in biocatalysis with monooxygenases, where the oxygen sensitivity of the mediator as well as possible direct electrochemical oxygen reduction have to be considered [71]. From this work, the authors concluded that the combination of an oxygen-independent enzyme reaction with an electrochemical reaction is more advisable. Therefore, a mediated electroenzymatic process to regenerate the NADPH in combination with a reaction catalyzed by flavin-dependent-OYE (old yellow enzymes) was investigated. High productivities up to 2.27 mM product h
1 in combination with approximately 90% electron transfer efficiency were measured [72]. In addition to approaches that concentrate on a search and optimization of a mediator for co-factor regeneration, some groups have attempted to replace natural co-factors with artificial redox mediators. For example, Cekic et al. [73] replaced NADPH with different artificial redox mediators in combination with P450cin monooxygenase. It was shown that cobalt sepulchrate, phenosafranine, safranine T, FAD, and FMN enabled artificial electron transfer from the platinum electrode to P450cin via the redox partner protein cindoxin. The highest product formation of 6.50  0.60 nmol (product) and (P450)
1 min
1 cm
2 was achieved using cobalt sepulchrate. An interaction between proteins and artificial redox mediators can be further tuned by protein engineering, as shown by Belsare et al. [3]. Furthermore, computational studies can provide a deeper understanding of the interaction of a mediator with the amino acids of an enzyme [74].

6 Conclusions The chapter summarized the fundamentals of enzyme employment in electrobiotechnological applications. The most common enzyme cofactors and mechanisms of electron transfer between an enzyme and an electrode have been introduced. The basics of thermodynamics of half-cell reactions and cell reactions were described, including the meanings of standard electrode potential, electrode potential at specified pH value, standard cell potential, and open circuit potential. Tables with Gibbs free energies of formation of selected substances and selected half-cell potential values at pH 7 for some selected examples were given. Kinetic expressions for DET and MET mechanisms were introduced and the overall reaction rate in an example of a DET mechanism were discussed. The balance equations necessary for mathematical descriptions of electroenzymatic processes on flat and porous enzymatic electrodes were introduced. This was followed by descriptions of selected electrochemical methods of study, such as stationary polarization, cyclic voltammetry, and electrochemical impedance spectroscopy. The descriptions of methods was accompanied by selected examples relevant to

Electron Transfer Between Enzymes and Electrodes

79

electrobiotechnological applications. It was also shown how kinetic and electrochemical rate constants can be determined from the measurements. The determination of total enzyme surface concentration was also discussed. Finally, some methods for the preparation of porous electroenzymatic electrodes and regeneration of soluble co-factors were introduced. It was demonstrated that the electrochemical methods, combined with proper mathematical descriptions, can provide significant insights on the reasons limiting the overall electrode behavior, contributing to better system design. However, to make these models more realistic, additional porous electrode characterizations are necessary. For example, there is a little knowledge about the real structure of porous enzymatic electrodes. Currently, only a few studies offer enough data on electrode thicknesses, porosity, and real surface area, while the data on realistic enzyme and mediator distributions are lacking. Most studies are not concerned with enzyme/ mediator/nanoparticle loadings or their utilization. Current results point out very low enzyme utilization in most publications. Because enzymes are catalysts which have to be produced and thus also create some waste, this factor has to be considered in the future if enzymatic systems are ever going to be competitive with fermentation systems. For systems with a main focus on material production, the characterization of product distribution is of paramount interest. Currently, most of these studies consider quite diluted systems. This is a significant drawback because it significantly increases the separation cost and the amount of waste liquids. To fully develop the potential of electrobiotechnological systems, electroenzymatic syntheses in more concentrated solutions have to be demonstrated. As was shown, the choice of mediator is not always simple, especially if additional substrates (e.g. oxygen) are required, as in the case of monooxygenases. Recent works indicate that the properties of enzymes toward artificial mediators can be further tuned by protein engineering. This direction will be hopefully be further explored in the future. To develop sustainable processes, issues related to the separation (recycling) and toxicity of selected mediators have to be carefully considered. Finally, to push forward new and exciting electrobiotechnological applications, more intensive interactions between different disciplines—including electrochemistry, biology, bioelectrochemistry, material science, and reaction engineering—are strongly recommended.

Appendix: List of Symbols A a Ageo c cDL D E

Amplitude (V)  Internal active surface area (m2act m
3 geo  Geometrical surface area of electrode (m2geo
3 Volumetric concentration (mol m ) Double layer capacitance (F m
2)  Diffusion coefficient of species (m2geo s
1 Electrode potential (V)

80 F f G gk I i Im(Z ) j k1, km kei KM L n P r R RΩ Re(Z ) T U vk w Y |Z| Z zk v

T. Vidakovic-Koch Faraday’s constant ¼ 96,485 (C mol
1) Frequency (Hz) Flow rate (m3 s
1)
1 Diffusion flux (k ¼ 1,2,3) (mol m
2 geo s ) Current (A) Imaginary number Imaginary part of electrochemical impedance Z (Ω m2) 
2 Current density (A mgeo Reaction constants of enzyme substrate (m3 mol
1 s
1/m2 mol
1 s
1) and enzyme mediator reactions Kinetic constant of the (s
1) Electrochemical reaction (i ¼ 1,2) Michaelis-Menten constant (mol m
3) Catalyst layer thickness (mgeo) Number of electrons Power density (W m
2) Reaction rate (mol s
1 m
2) Universal gas constant ¼ 8.314 (J mol
1 K
1) Electrolyte resistance (Ω m2) Real part of electrochemical impedance Z (Ω m2) Temperature (K) Cell potential (V) Average molar velocity (k ¼ 1,2,3) (m s
1) Rotation rate of rotating disc electrode (rad s
1) Linear frequency response function (Ω
1 m
2) Magnitude of electrochemical impedance Z (Ω m2) Electrochemical impedance (Ω m2) Space coordinate (k ¼ 1,2,3) Sweep rate (V s
1)

Greek ν η φ Γ ηi Δf Gio ΔrGo ΔrHo ϕ E, ϕ I γE , γI α,β δ ε ι ω

Stoichiometric coefficient Overpotential (V) Phase shift (o) Surface concentration (mol m
2) Efficiency (i ¼ th, ec,fuel) Standard Gibbs free energies of formation of component “i” (kJ mol
1) Standard Gibbs free energy change of reaction (kJ mol
1) Standard enthalpy change of reaction (kJ mol
1) Potentials of electron-and ion- conducting phases, respectively (V) Electron-and ion conductivities (S m
1 geo ) Transfer coefficients of electrochemical steps Diffusion layer thickness  (mgeo) Void fraction (m3 m
3 geo  Local current density (A m
2 act
1 Angular frequency (rad s )

Electron Transfer Between Enzymes and Electrodes

Super- and Sub-scripts A,C,cell act, geo CL, DL e0 ec I, E o o,# Ohm ox,red S sim, exp. SS th

Anode, cathode and cell respectively Active and geometrical respectively Catalyst layer and diffusion layer respectively Electrochemical reaction step Electrochemical Ion and electron conducting phase respectively Standard conditions At pH 7 Ohmic Oxidized and reduced states respectively Substrate Simulation and experiment respectively Steady state Thermodynamic

List of Abbreviations A Ag/AgCl C CC CE CL DET DET_SS E EM ES FAD FMN GOx HRP Int Medi (i ¼ ox,red) MET NAD NADP P RE RH S SAMs SCE SHE WE

Anode Silver/silver chloride reference electrode Cathode Current collector Counter electrode Catalyst layer Direct electron transfer Direct electron transfer steady state Enzyme Enzyme mediator complex Enzyme substrate complex Flavin adenine dinucleotide Flavin mononucleotide Glucose oxidase Horseradish peroxidase Intermediate Oxidized and reduced forms of a mediator Mediated electron transfer Nicotinamide adenine dinucleotide Nicotinamide adenine dinucleotide phosphate Product Reference electrode Organic substrate Substrate Self-assembled monolayers Saturated calomel electrode Standard hydrogen electrode Working electrode

81

82

T. Vidakovic-Koch

References 1. Copeland RA (2000) Enzyme reactions with multiple substrates. In: Enzymes: a practical introduction to structure, mechanism, and data analysis, 2nd edn. Wiley, New York, pp 350–366 2. Vidakovic-Koch T, Sundmacher K (2017) Porous electrodes in bioelectrochemistry. In: V P (ed) Encyclopedia of interfacial chemistry: surface science and electrochemistry, Elsevier, Amsterdam 3. Belsare KD et al (2017) Directed evolution of P450cin for mediated electron transfer. Protein Eng Des Sel 30(2):119–127 4. Habermüller K, Mosbach M, Schuhmann W (2000) Electron-transfer mechanisms in amperometric biosensors. Fresenius J Anal Chem 366(6):560–568 5. Do TQN et al (2014) Mathematical modeling of a porous enzymatic electrode with direct electron transfer mechanism. Electrochim Acta 137:616–626 6. Andreu R et al (2007) Direct electron transfer kinetics in horseradish peroxidase electrocatalysis. J Phys Chem B 111(2):469–477 7. Gupta G et al (2011) Direct electron transfer catalyzed by bilirubin oxidase for air breathing gas-diffusion electrodes. Electrochem Commun 13(3):247–249 8. Le´ger C et al (2002) Effect of a dispersion of interfacial electron transfer rates on steady state catalytic electron transport in [NiFe]-hydrogenase and other enzymes. J Phys Chem B 106 (50):13058–13063 9. Tasca F et al (2008) Direct electron transfer at cellobiose dehydrogenase modified anodes for biofuel cells. J Phys Chem C 112(26):9956–9961 10. Zimmermann H et al (2000) Anisotropic orientation of horseradish peroxidase by reconstitution on a Thiol-modified gold electrode. Chem Eur J 6(4):592–599 11. Vidakovic´-Koch T et al (2011) Impact of the gold support on the electrocatalytic oxidation of sugars at enzyme-modified electrodes. Electroanalysis 23(4):927–930 12. Courjean O, Gao F, Mano N (2009) Deglycosylation of glucose oxidase for direct and efficient glucose electrooxidation on a glassy carbon electrode. Angew Chem 121(32):6011–6013 13. Bartlett PN, Al-Lolage FA (2017) There is no evidence to support literature claims of direct electron transfer (DET) for native glucose oxidase (GOx) at carbon nanotubes or graphene. J Electroanal Chem 14. Alberty RA (1993) Thermodynamics of reactions of Nicotinamide adenine dinucleotide and Nicotinamide adenine dinucleotide phosphate. Arch Biochem Biophys 307(1):8–14 15. Krebs HA, Kornberg HL (1957) Energy transformations in living matter. Springer-Verlag Berlin Heidelberg, Berlin 16. Reiger PH (1994) Electrochemistry.2nd edn. Springer Science+Business Media, Dordrecht 17. Armstrong FA, Hirst J (2011) Reversibility and efficiency in electrocatalytic energy conversion and lessons from enzymes. Proc Natl Acad Sci 108(34):14049–14054 18. Togo M et al (2007) An enzyme-based microfluidic biofuel cell using vitamin K3-mediated glucose oxidation. Electrochim Acta 52(14):4669–4674 19. Mano N, Mao F, Heller A (2002) A miniature biofuel cell operating in a physiological buffer. J Am Chem Soc 124(44):12962–12963 20. Chen T et al (2001) A miniature biofuel cell. J Am Chem Soc 123(35):8630–8631 21. Kim H-H et al (2003) A miniature membrane-less biofuel cell operating under physiological conditions at 0.5 V. J Electrochem Soc 150(2):A209–A213 22. Willner I et al (1998) Biofuel cell based on glucose oxidase and microperoxidase-11 monolayer-functionalized electrodes. J Chem Soc Perkin Trans 2(8):1817–1822 23. Kuwahara T et al (2007) Properties of the enzyme electrode fabricated with a film of polythiophene derivative and its application to a glucose fuel cell. J Appl Polym Sci 104 (5):2947–2953 24. Brunel L et al (2007) Oxygen transport through laccase biocathodes for a membrane-less glucose/O2 biofuel cell. Electrochem Commun 9(2):331–336

Electron Transfer Between Enzymes and Electrodes

83

25. Bedekar AS et al (2007) Oxygen limitation in microfluidic biofuel cells. Chem Eng Commun 195(3):256–266 26. Nazaruk E et al (2008) Enzymatic biofuel cell based on electrodes modified with lipid liquidcrystalline cubic phases. J Power Sources 183(2):533–538 27. Tamaki T, Yamaguchi T (2006) High-surface-area three-dimensional biofuel cell electrode using redox-polymer-grafted carbon. Ind Eng Chem Res 45(9):3050–3058 28. Matsue T et al (1985) Electron-transfer reactions associated with host-guest complexation. Oxidation of ferrocenecarboxylic acid in the presence of .beta.-cyclodextrin. J Am Chem Soc 107(12):3411–3417 29. Yan Y-M, Yehezkeli O, Willner I (2007) Integrated, electrically contacted NAD(P) +
dependent enzyme–carbon nanotube electrodes for biosensors and biofuel cell applications. Chem Eur J 13(36):10168–10175 30. Li X et al (2008) A miniature glucose/O2 biofuel cell with single-walled carbon nanotubesmodified carbon fiber microelectrodes as the substrate. Electrochem Commun 10(6):851–854 31. Gao F et al (2007) An enzymatic glucose/O2 biofuel cell: preparation, characterization and performance in serum. Electrochem Commun 9(5):989–996 32. Yan Y et al (2006) Carbon-nanotube-based glucose/O2 biofuel cells. Adv Mater 18 (19):2639–2643 33. Vidakovic´-Koch T et al (2013) Application of electrochemical impedance spectroscopy for studying of enzyme kinetics. Electrochim Acta 110:94–104 34. Vidakovic´-Koch TR et al (2011) Nonlinear frequency response analysis of the Ferrocyanide oxidation kinetics. Part I. A theoretical analysis. J Phys Chem C 115(35):17341–17351 35. Do TQN et al (2015) Dynamic and steady state 1-D model of mediated electron transfer in a porous enzymatic electrode. Bioelectrochemistry 106(Part A):3–13 36. Vidakovic´-Koch T et al (2017) Catalyst layer modeling. In: Breitkopf C, Swider-Lyons K (eds) Springer handbook of electrochemical energy. Springer Berlin Heidelberg, Berlin, pp 259–285 37. Bard A, Faulkner L (2001) Electrochemical methods, fundamentals and applications. Wiley, New York 38. Bartlett PN (2008) In: Bartlett PN (ed) Bioelectrochemistry: fundamentals, experimental techniques and applications. Wiley, New York 39. Varnicic M et al (2014) Combined electrochemical and microscopic study of porous enzymatic electrodes with direct electron transfer mechanism. RSC Adv 4(69):36471–36479 40. Varnicˇic´ M, Vidakovic´-Koch T, Sundmacher K (2015) Gluconic acid synthesis in an electroenzymatic reactor. Electrochim Acta 174:480–487 41. Varnicˇic´ M, Vidakovic´-Koch T, Sundmacher K (2015) Corrigendum to “Gluconic acid synthesis in an Electroenzymatic reactor” [Electrochimica Acta 174 (2015) 480–487]. Electrochim Acta 176:1523 42. Ivanov I (2012) Development of a glucose-oxygen enzymatic fuel cell. Otto von Guericke University, Magdeburg, p 118 43. Ivanov I, Vidakovic´-Koch T, Sundmacher K (2013) Alternating electron transfer mechanism in the case of high-performance tetrathiafulvalene–tetracyanoquinodimethane enzymatic electrodes. J Electroanal Chem 690:68–73 44. Vidakovic-Koch T (2017) Energy conversion based on bio(electro)catalysts. In: Cornelia Breitkopf KS-L (ed) Springer handbook of electrochemical energy. Springer-Verlag Berlin Heidelberg, Berlin, pp 757–777 45. Mondal MS, Fuller HA, Armstrong FA (1996) Direct measurement of the reduction potential of catalytically active cytochrome c peroxidase compound I: voltammetric detection of a reversible, cooperative two-electron transfer reaction. J Am Chem Soc 118(1):263–264 46. Mondal MS, Goodin DB, Armstrong FA (1998) Simultaneous voltammetric comparisons of reduction potentials, reactivities, and stabilities of the high-potential catalytic states of wildtype and distal-pocket mutant (W51F) yeast cytochrome c peroxidase. J Am Chem Soc 120 (25):6270–6276

84

T. Vidakovic-Koch

47. Ruzgas T et al (1995) Kinetic models of horseradish peroxidase action on a graphite electrode. J Electroanal Chem 391(1):41–49 48. Ferapontova EE, Gorton L (2001) Effect of proton donors on direct electron transfer in the system gold electrode–horseradish peroxidase. Electrochem Commun 3(12):767–774 49. Yu H et al (2015) Influence of the ionomer/carbon ratio for low-Pt loading catalyst layer prepared by reactive spray deposition technology. J Power Sources 283:84–94 50. Lojou E´, Bianco P (2004) Membrane electrodes for protein and enzyme electrochemistry. Electroanalysis 16(13–14):1113–1121 51. De Poulpiquet A et al (2013) Exploring properties of a hyperthermophilic membrane-bound hydrogenase at carbon nanotube modified electrodes for a powerful H2/O2 biofuel cell. Electroanalysis 25(3):685–695 52. Lojou E´ et al (2008) Biocatalysts for fuel cells: efficient hydrogenase orientation for H2 oxidation at electrodes modified with carbon nanotubes. JBIC, J Biol Inorg Chem 13 (7):1157–1167 53. Luo X et al (2009) Immobilization of the hyperthermophilic hydrogenase from Aquifex aeolicus bacterium onto gold and carbon nanotube electrodes for efficient H2 oxidation. JBIC, J Biol Inorg Chem 14(8):1275–1288 54. Lojou E (2011) Hydrogenases as catalysts for fuel cells: strategies for efficient immobilization at electrode interfaces. Electrochim Acta 56(28):10385–10397 55. Yan Y, Su L, Mao L (2007) Multi-walled carbon nanotube-based glucose/O2 biofuel cell with glucose oxidase and Laccase as biocatalysts. J Nanosci Nanotechnol 7(4–1):1625–1630 56. Khan GF (1997) TTF-TCNQ complex based printed biosensor for long-term operation. Electroanalysis 9(4):325–329 57. Khan GF (1996) Construction of SEC/CTC electrodes for direct electron transferring biosensors. Sens Actuators B Chem 36(1–3):484–490 58. Khan GF, Ohwa M, Wernet W (1996) Design of a stable charge transfer complex electrode for a third-generation amperometric glucose sensor. Anal Chem 68(17):2939–2945 59. Ivanov I, Vidakovic´-Koch T, Sundmacher K (2011) Direct hybrid glucose–oxygen enzymatic fuel cell based on tetrathiafulvalene–tetracyanoquinodimethane charge transfer complex as anodic mediator. J Power Sources 196(22):9260–9269 60. Do TQN (2017) Mathematical modelling of electro-enzymatic system. Otto von Guericke University, Magdeburg 61. Kochius S et al (2014) Electrochemical regeneration of oxidised nicotinamide cofactors in a scalable reactor. J Mol Catal B Enzym 103:94–99 62. Gorton L, Domı´nguez E (2007) Electrochemistry of NAD(P)+/NAD(P)H. Encyclopedia of electrochemistry. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 63. Karyakin AA, Ivanova YN, Karyakina EE (2003) Equilibrium (NAD+/NADH) potential on poly(neutral red) modified electrode. Electrochem Commun 5(8):677–680 64. Arechederra MN, Addo PK, Minteer SD (2011) Poly(neutral red) as a NAD+ reduction catalyst and a NADH oxidation catalyst: towards the development of a rechargeable biobattery. Electrochim Acta 56(3):1585–1590 65. Karyakin AA et al (1999) Electropolymerized azines: part II. In a search of the best electrocatalyst of NADH oxidation. Electroanalysis 11(8):553–557 66. Li H, Wen H, Calabrese Barton S (2012) NADH oxidation catalyzed by electropolymerized azines on carbon nanotube modified electrodes. Electroanalysis 24(2):398–406 67. Rinco´n RA et al (2011) Flow-through 3D biofuel cell anode for NAD+
dependent enzymes. Electrochim Acta 56(5):2503–2509 68. Vuorilehto K, Lütz S, Wandrey C (2004) Indirect electrochemical reduction of nicotinamide coenzymes. Bioelectrochemistry 65(1):1–7 69. Salimi A et al (2009) Electrocatalytic reduction of NAD+ at glassy carbon electrode modified with single-walled carbon nanotubes and Ru(III) complexes. J Solid State Electrochem 13 (3):485–496

Electron Transfer Between Enzymes and Electrodes

85

70. Ali I, Soomro B, Omanovic S (2011) Electrochemical regeneration of NADH on a glassy carbon electrode surface: the influence of electrolysis potential. Electrochem Commun 13 (6):562–565 71. Tosstorff A et al (2014) Mediated electron transfer with monooxygenases—insight in interactions between reduced mediators and the co-substrate oxygen. J Mol Catal B Enzym 108:51–58 72. Tosstorff A et al (2017) Towards electroenzymatic processes involving old yellow enzymes and mediated cofactor regeneration. Eng Life Sci 17(1):71–76 73. C¸ekic¸ SZ et al (2010) Mediated electron transfer with P450cin. Electrochem Commun 12 (11):1547–1550 74. Str€ ohle FW et al (2013) A computational protocol to predict suitable redox mediators for substitution of NAD(P)H in P450 monooxygenases. J Mol Catal B Enzym 88:47–51

Adv Biochem Eng Biotechnol (2019) 167: 87–134 DOI: 10.1007/10_2017_33 © Springer International Publishing AG 2017 Published online: 14 November 2017

Enzyme-Based Electrobiotechnological Synthesis Lisa Marie Schmitz, Katrin Rosenthal, and Stephan L€ utz

Abstract Oxidoreductases are enzymes with a high potential for organic synthesis, as their selectivity often exceeds comparable chemical syntheses. The biochemical cofactors of these enzymes need regeneration during synthesis. Several regeneration methods are available but the electrochemical approach offers an efficient and quasi mass-free method for providing the required redox equivalents. Electron transfer systems involving direct regeneration of natural and artificial cofactors, indirect electrochemical regeneration via a mediator, and indirect electroenzymatic cofactor regeneration via enzyme and mediator have been investigated. This chapter gives an overview of electroenzymatic syntheses with oxidoreductases, structured by the enzyme subclass and their usage of cofactors for electron relay. Particular attention is given to the productivity of electroenzymatic biotransformation processes. Because most electroenzymatic syntheses suffer from low productivity, we discuss reaction engineering concepts to overcome the main limiting factors, with a focus on media conductivity optimization, approaches to prevent enzyme inactivation, and the application of advanced cell designs.

L.M. Schmitz, K. Rosenthal, and S. Lütz (*) Department of Biochemical and Chemical Engineering, TU Dortmund University, Dortmund, Germany e-mail: [email protected]

88

L.M. Schmitz et al.

Graphical Abstract

Keywords Cofactor regeneration, Electrochemistry, Electron mediator, Enzyme catalysis, Oxidoreductases, Reactor design, Rhodium complex Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Dehydrogenases and Reductases in Electrobiotechnological Synthesis . . . . . . . . . . . . . . . . . . . 2.1 Oxidations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Carbonyl Reductions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Ene Reductions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Oxidases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Monooxygenases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Peroxidases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Reaction Engineering Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Enzyme Stabilization by Compartmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 Reaction Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Reactor Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

88 90 90 104 108 109 111 113 115 116 119 122 127 128

1 Introduction Electroenzymatic biotechnology comprises different techniques that interface enzymes with electrodes. First, enzyme-electrodes have become important as analytical devices such as biosensors [1]. Second, electro-enzymatic systems are used in the field of energy conversion and storage for devices such as biofuel cells [2, 3] and biocapacitors [3, 4]. Third, electrochemical reactions can be used to provide the necessary redox equivalents for enzymatic biotransformations. These biotransformations, in which a cofactor or cosubstrate is (re)generated by an electrochemical reaction, are referred to as electroenzymatic syntheses [5– 7]. Although synthetic biomimetic cofactors have recently led to some regeneration

Enzyme-Based Electrobiotechnological Synthesis

89

systems that are superior to natural systems, they have not been applied in electroenzymatic processes and are therefore not considered in this review [8]. Enzymes show remarkable chemo-, regio-, and stereoselectivities and catalyze organic syntheses mostly without side reactions. In particular, biocatalytic redox reactions with oxidoreductases offer selectivities and yields that are often not achievable with chemical syntheses. The reaction range of oxidoreductases is broad, covering dehydrogenation, oxygenation, disproportionation or singleelectron transfer, and oxidative bond formation [9]. Electron supply or removal is always required when using oxidoreductases. This function is fulfilled by cofactors such as nicotinamide adenine dinucleotide [NAD(H)], nicotinamide adenine dinucleotide phosphate [NADP(H)] or enzyme-bound cofactors such as flavin adenine dinucleotide (FAD), flavin mononucleotide (FMN), and heme. To render syntheses using oxidoreductases economically feasible, regeneration of these cofactors is required. Regeneration of the desired redox state of the cofactor is dependent on the type of enzyme and its cofactor and can be achieved by several means, such as the simultaneous addition or in situ generation of a sacrificial cosubstrate. An example of this straightforward approach is the addition of H2O2 to regenerate the reactive compound I intermediate in heme peroxidase reactions. Other established and highly developed approaches deal with the in situ enzymecoupled regeneration of NAD(P)H with formate dehydrogenase (FDH) or glucose dehydrogenase (GDH). These approaches can be coupled to various enzymatic syntheses [10–13]. NADH regeneration with FDH requires formate as cosubstrate, with carbon dioxide being formed as a product. NADPH regeneration with GDH requires D-glucose to form D-glucono-δ-lactone [14]. FDH-based cofactor regeneration is thermodynamically favorable, but the high cost of enzyme production and low activities restrict its application [15]. Compared to FDH, GDH-based cofactor regeneration is less expensive and more stable, and thus more suitable for preparative scale syntheses [16]. In contrast, enzymatic syntheses coupled to electrochemical cofactor regeneration suffer from low productivities, which limit their industrial implementation [17], but present considerable advantages over enzymebased cofactor regeneration. The avoidance of byproducts and second enzymes simplifies product isolation [18]. Waste in the form of coproducts can be measured in terms of atom efficiency (calculated in grams cosubstrate used per mole electron). The calculation of atom efficiency, which is 23 for FDH-coupled and 130 for GDH-coupled regeneration, clearly shows the waste of resources for enzymecoupled regeneration approaches [13]. Electroenzymatic syntheses could thus contribute to lower waste production coupled to the simultaneous synthesis of desired product [6]. Thus, electroenzymatic syntheses are attractive, especially because of the quasi mass-free cofactor regeneration by electrons, which are the cheapest redox equivalents available [19]. However, the combination of electrochemical and enzymatic steps requires development of reaction engineering concepts. Specially designed electrochemical cells are required for electroenzymatic syntheses. The combination of isolated enzymes with electrochemical reactions is well established for electrochemical biosensors, where the biocatalyst is immobilized onto the electrode surface

90

L.M. Schmitz et al.

[20]. This arrangement is useful for analytical purposes, but is mostly unsuitable for synthetic applications. The long-term stability of enzymes and their productivity is limited in these systems because of low current densities, and only a few examples exist for preparative scale syntheses. To increase the current densities for preparative syntheses, electron transfer between electrode surface and enzyme is often carried out in the form of indirect electron transfer via a mediator molecule. This approach also enables the choice of reaction conditions more suitable for the enzymatic conversion step [21]. The focus of this chapter is productive electroenzymatic biotransformation processes. Mechanistic studies on cofactor regeneration or the development of cofactor regeneration systems in analytical cell setups are omitted. Electroenzymatic research is an interdisciplinary area covering fields in enzymology, electrochemistry, and organic synthesis, which sometimes use different data reporting systems and focus on different parameters. To ensure comparability of the presented electroenzymatic biotransformation processes, we define the set of parameters used. The catalyst consumption is given in total turnover number ttn, which is defined as the moles of product formed per mole of catalyst consumed. The maximum velocity of the catalyzed reaction at substrate saturation is Vmax. The turnover frequency (tof) is defined as the number of catalytic conversions per catalytic site and unit of time. In this chapter, tof is given in hour
1. Conversion indicates the conversion of substrate to product. The purity of chiral substances is given as enantiomeric excess (ee), which describes the degree to which one enantiomer is present in a greater amount than the other. A further criterion for the efficiency of electroenzymatic biotransformation processes is the productivity. In this chapter, productivity is given as the initial reaction rate (millimoles per liter per hour). If information about the space time yield (STY) is available, it is given in brackets. The STY describes the mass of product formed per volume and time (grams per liter per day). Parameters that are also important with respect to productivities are the conductivity, defined as the ability to conduct electricity (millisiemens per centimeter) and the current efficiency (dimensionless), defined as the ratio of used electric charge to theoretically necessary charge. Data on case studies are included in the comment sections of Tables 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, whenever suitable and available.

2 Dehydrogenases and Reductases in Electrobiotechnological Synthesis 2.1

Oxidations

Cleavage of hydrogen from substrates by dehydrogenases usually involves reduction of the cofactor NAD(P)+ to NAD(P)H. In enzyme-based systems, a second enzyme is required for cofactor regeneration. Alternatively, the production enzyme

6-Phosphogluconate dehydrogenase from pig heart Isocitrate dehydrogenase from pig heart Isocitrate dehydrogenase from pig heart (E.C. 1.1.1.41) Glucose dehydrogenase (GDH)

Ribulose 5-phosphate 2-Oxoglutarate 2-Oxoglutarate Gluconic acid

(2R,3S)-Isocitrate

(2R,3S)-Isocitrate

D-Glucose

Cyclohexanone

Cyclohexanol

6-Phosphogluconate

2-Butanon

2-Butanol

2-Hexen-1-ol

Alcohol dehydrogenase from yeast (YADH) Alcohol dehydrogenase from Thermoanaerobium brockii (TADH) HLADH

3a,4,7,7a-Tetrahydro3H-isobenzofuran-1one 4,10-Dioxa-tricyclo [5.2.1.02,6]dec-8-en-3one 2-Hexenal

meso-3,4-Dihydroxymethylcyclohexene

meso-5,6-Dihydroxy-methyl-7oxabicyclo [2.2.1]hept-2-ene

Cyclohexanone

Gluconic acid

D-Glucose

Cyclohexanol

Product D-Glucono-δ-lactone

Substrate

D-Glucose

HLADH

Alcohol dehydrogenase from horse liver (HLADH) HLADH

Enzyme Glucose dehydrogenase(GDH) from Bacillus spec. GDH

Table 1 NAD(P)+-dependent electroenzymatic oxidations catalyzed by dehydrogenases

CAV, AMAPOR CAV, AMAPOR A-Q-2-S, AMAPOR ABTS

Pd derivative

Fe(tmphen)3

Fe(tmphen)3

PDMe (BF4)2

3,4Dihydroxybenzaldehyde Ru(PD)3 (ClO4
)2 PDMe (BF4)2

Mediator None

[33]

[31]

[31]

[31]

[29]

(continued)

conversion 100% of 2 mM substrate in 10 h conversion 98% of 20 mM substrate in 2.5 h 0.005 mM h
1 after 3.8 h (ee > 99%) 0.005 mM h
1 after 3.6 h (ee > 99%) 7.2 mM h
1 (33.6 g L day
1) ttncofactor 93 ttnmediator 1,860

[28]

1.77 mM, ttncofactor 18, ttnmediator 36 4.1 mM, ttncofactor 41, ttnmediator 82

[28]

[27]

Conversion 74% of 14 mM substrate, tofmediator 30 h
1

[27]

[26]

[24]

20.7 mM h
1, ttncofactor 100

Conversion 75% of 20 mM substrate, tofcofactor 28 h
1 Conversion 99.5% of 14 mM substrate, tofmediator 35 h
1

Ref. [23]

Comments 100 mM, ttncofactor 900

Enzyme-Based Electrobiotechnological Synthesis 91

(R)-Phenylethane-1,2-diol

Glycerol dehydrogenase from Cellulomonas sp. (GDH) Aldehyde dehydrogenase from Deinococcus geothermalis

D,L-Glyceraldehyde

Substrate meso-3,4-Dihydroxymethylcyclohexene

Enzyme HLADH

Table 1 (continued) Product 3a,4,7,7a-Tetrahydro3H-isobenzofurane-1on (S)-Phenylethane-1,2diol Glyceric acid ABTS

ABTS

Mediator ABTS

17 mM ee 88%

Comments Conversion 93.5% of 48 mM substrate, ee > 99.5

[35]

[34]

Ref. [32]

92 L.M. Schmitz et al.

L-Glutamate

Cyclohexanol (1S,2S)-(+)-2Methylcyclohexanol (1S,3S)-(
)-2Methylcyclohexanol D-Lactate (R)-Mandelate (R)-Mandelate Cyclohexanol

α-Ketoglutarate

Cyclohexanone 2-Methylcyclohexanone 3-Methylcyclohexanone Pyruvate

Benzoylformate

Benzoylformate Cyclohexanone

Pyruvate

LDH from Leuconostoc mesenteroides Benzoylformate dehydrogenase (BFR) from Enterococcus faecalis BFR from E. faecalis ADH from horse liver (HLADH, E.C. 1.1.1.1) LDH

LDH

L-Glutamate

α-Ketoglutarate

Oxoglutarate Oxoglutarate

GluDH

GluDH from beef liver GluDH from beef liver

(S)-Glutamate (S)-Glutamate

MV2+ + FDR

D-Lactate

Pyruvate

MV2+, AMAPOR CoSep, AMAPOR

MV2+ + FDR

Conversion 94% of 150 mM substrate in 9 days, ee 94%, ttnenzyme 3.5  107, ttncofactor 940, ttnLipDH 5.4  105 Conversion 90% of 200 mM substrate in 14 days, ee 94%, ttnenzyme 2.2  107, ttncofactor 900, ttnFDR 7.3  106 Conversion 100% of 200 mM substrate in 7 days, ttnenzyme 1.1  107, ttncofactor 1,000, ttnFDR 7.5  106 0.003 mM h
1, ttncofactor 29,000 0.0001 mM h
1

MV2+ + LipDH

D-Lactate

(continued)

[31] [31]

[42]

[42]

[42]

[40] [41]

95% of 50 mM substrate in 18 h, ttncofactor 158 Conversion 65% of 100 mM substrate

[39]

[38]

[38] [38]

[37]

Ref. [36]

FAD, LipDH MV2+ + LipDH

Comments Conversion 72% of 25.3 mM substrate in 21 h, ttncofactor 1,400 Conversion 100% of 10 mM substrate, ttncofactor 3,300 conversion 100% of 50 mM substrate Conversion 48.9% of 50 mM substrate, ee > 99%, ttnmediator 91 Conversion 51.7% of 50 mM substrate, ee 93.1%, ttnmediator 94 Conversion 80% of 50 mM substrate

[40]

MV2+, diaphorase

MV2+, diaphorase

MV2+, diaphorase

MV2+, diaphorase MV2+, diaphorase

None

Mediator None

80% of 50 mM substrate in 30 h, ttncofactor 133

ADH

L-Glutamate dehydrogenase (GluDH) from beef liver Alcohol dehydrogenase (ADH) ADH

Product D-Lactate

dehydrogenase (LDH)

Substrate Pyruvate

D-Lactate

Enzyme

Table 2 NAD(P)+-dependent electroenzymatic reductions catalyzed by dehydrogenases with direct or indirect cofactor regeneration using methyl viologen (MV2+) in combination with diaphorase, lipoamide dehydrogenase (LipDH), ferredoxin reductase (FNR), and AMAPORs

Enzyme-Based Electrobiotechnological Synthesis 93

Substrate α-Ketoglutarate

Various ketones and alcohols Various ketones and alcohols

Enzyme GluDH from beef liver

ADH from Thermoanaerobium brockii or equine liver ADH from T. brockii or equine liver

Table 2 (continued)

[45]

MV2+ + FNR or MV2+ + diaphorase

Alcohols

Ref. [43] [45]

Comments 100% conversion of 5 mM substrate in 5 h tncofactor 200 h
1

Acetophenone

Mediator Hydrogenase

Alcohols

Product L-Glutamate

94 L.M. Schmitz et al.

Cyclohexanone

Acetophenone

Acetophenone

Acetophenone

4-Chloroacetophenone

(R)-3Methylcyclohexanone

HLADH

Alcohol dehydrogenase from Lactobacillus brevis (LbADH)

LbADH

LbADH

LbADH

Alcohol dehydrogenase from Thermus sp.

(1S,3S)-3Methylhexanol

(R)-4Chlorophenylethanol

(R)-Phenylethanol

(R)-Phenylethanol

(R)-Phenylethanol

Cyclohexanol

(S)-4-Phenyl-2butanol (S)-4-Phenyl-2butanol Cyclohexanol

4-Phenyl-2-butanone

Cyclohexanone

D-Lactate

Pyruvate

4-Phenyl-2-butanone

Product Cyclohexanol

Substrate Cyclohexanone

Alcohol dehydrogenase from Rhodococcus sp. (S-ADH) HLADH

Enzyme Alcohol dehydrogenase from horse liver (HLADH) D-Lactate dehydrogenase from Staphylococcus epidermidis (LDH) HLADH

Cp*Rh(bpy)L (polymerbound) Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

Cp*Rh(bpy)L

[Cp*Rh(bpy) Cl]+

Mediator [Rh(bpy)3]2+

1.2 mM h
1 (3.1 g L
1 day
1) (for the organic phase), de > 96%

Conversion 70% of 10 mM substrate in 5 h, ee 65% (S-product) Conversion 76% of 10 mM substrate in 5 h, ee 77% (S-product) Conversion 92% of 100 mM substrate in 45 days Conversion 100% of 100 mM substrate in 3 days Conversion 98% of 17 mM substrate, ee > 99.9%, 4.7 mM h
1 (13.5 g L
1 day
1), ttnenzyme 75,000, ttncofactor 35, ttnmediator 35 Conversion 93% of 13 mM substrate, ee > 98%, 3.1 mM h
1 (9.0 g L
1 day
1), ttnenzyme 21,000, ttncofactor 12, ttnmediator 55 Conversion 65% of 300 mM substrate in the organic phase, ee > 99.9%, 1.0 mM h
1 (3 g L
1 day
1) (for the organic phase), ttnenzyme 5,000, ttncofactor 64, ttnmediator 64 0.42 mM h
1, conversion 90% of 4 mM substrate, ee > 97.3%, ttnmediator 214

Comments Conversion ~26% of 1.12 mM substrate ttncofactor 2.9 ttnmediator 1.2 Conversion 70% in 3 h of 20 mM pyruvate, ee 93.5%, ttncofactor 7, ttnmediator 14

[51]

[52]

[50]

[50]

[50]

[41, 49]

[49]

[5]

[5]

[48]

Ref. [47]

Table 3 NAD(P)+-dependent electroenzymatic reductions catalyzed by dehydrogenases with indirect cofactor regeneration using Rh(bpy) complexes

Enzyme-Based Electrobiotechnological Synthesis 95

96

L.M. Schmitz et al.

Table 4 NAD(P)H-dependent electroenzymatic reductions catalyzed by ene reductases Enzyme Enoate reductase from Clostridium tyrobutyricum Enoate reductase from C. tyrobutyricum

Old yellow enzyme (OYE) from Thermoanaerober pseudethanolicus Pentaerythritol tetranitrate reductase (PETNR) from Enterobacter cloacae PB2 OYE from T. pseudethanolicus

Substrate (E)-2Methyl-3phenyl-2propenoate (E)-2Methyl-3phenyl-2propenoate Various substrates

Product (2R)-2-Methyl-3phenylpropionate

Mediator MV2+

(2R)-2-Methyl-3phenylpropionate

MV2+

Various products

Various substrates

Various products

2Cyclohexen1-one

Cyclohexanone

Comments Conversion 95% of 81 mM substrate in 12 h ee 98%, enzyme halflife 350 h

Ref. [54]

MV2+

–

[56]

MV2+

–

[56]

[Cp*Rh (bpy) (H2O)]2+

2.27 mM h
1 tofmediator 45 h
1

[57]

[55]

can serve as regeneration enzyme in a substrate-coupled approach [22], whereby the enzyme produces the desired compound and also regenerates the cofactor. In electrochemical regeneration systems, the cofactor can be easily regenerated with direct or mediator-promoted reactions (Fig. 1a, b). In enzyme-coupled regeneration, electron transfer with mediators is coupled to a second enzyme (Fig. 1c). Direct and indirect electrochemical regeneration avoids the need for additional regeneration enzymes. Table 1 gives an overview of the processes reported to date. Direct electrochemical regeneration of NAD+ was successfully applied for oxidizing D-glucose to D-glucono-δ-lactone by GDH in a two-compartment cell, where working and auxiliary compartments were separated by a membrane [23]. In this model system, a maximal product formation of 100 mM with a ttn of 900 for the cofactor was obtained. The general drawback of direct electrochemical regeneration of NAD+ is the requirement for high oxidation potentials of at least 900 mV versus the standard hydrogen electrode (SHE) [5]. The high potentials can cause passivation, electrode fouling, or unselective oxidation of substrates and products, which limit the applicability of direct electrochemical regeneration. These drawbacks can be overcome by using mediators for electron transfer. For example, the mediator 3,4-dihydroxybenzalaldehyde immobilized on the anode enabled a reduction in overpotential from 700 to 200 mV for NAD+ regeneration [24]. The decreased potential resulted in increased stability and activity of GDH because of the prevention of co-oxidative enzyme deactivation by high redox potentials [25]. Different phenanthrolinedione complexes are widely studied as mediators and applied for electrochemical cofactor regeneration in dehydrogenase-based

p-Ethylphenol

p-Ethylphenol

PCMH from P. putida

Ethylphenol methylenehydroxylase from P. putida (EPMH) Galactose oxidase from the fungus Fusarium NRRL 2903 (GOase)

Glycerine-3-phosphate oxidase from Pediococcus sp. D-Amino acid oxidase from Trigonopsis variabilis (D-AAO) D-Amino acid oxidase from porcine kidney (D-AAO) L-Amino acid oxidase from Crotalus adamanteus venom (LAAO) Crude enzyme extract from Escherichia coli Crude enzyme extract from Saccharomyces cerevisiae

p-Cresol

PCMH from P. putida

Dihydroxyacetone phosphate (DHAP) L-Methionine D-Alanine L-Phenylalanine

Fumarate Xylose

Glycerine-3phosphate DL-Methionine

Pyruvic acid

Phenylpyruvic acid

Succinate

Xylitol

L-Xylitol

p-Hydroxy benzaldehyde p-Hydroxy benzaldehyde

p-Cresol

(S)-1(4-Hydroxyphenyl)ethanol (R)-1(4-Hydroxyphenyl)ethanol L-Xylose

Product p-Hydroxy benzaldehyde

Substrate p-Cresol

Enzyme p-Cresol methylenehydroxylase (PCMH) from Pseudomonas alcaligenes u.a. PCMH from Pseudomonas putida

Table 5 Flavin-dependent synthesis reactions catalyzed by oxidases

None

None

Ferrocene carboxylic acid (FcCOOH) 1-Aminopropyl-10 -methyl4,40 -dipyridinium iodide 1-Aminopropyl-10 -methyl4,40 -dipyridinium iodide

PEG-ferrocene, α,ω,bismethylferrocene, PEG (20,000) PEG-ferrocene, α, ω,bismethylferrocene, PEG (20,000) PEG-ferrocene

PEG-ferrocene, α,ω,bismethylferrocene, PEG (20,000) PEG-ferrocene

PEG-ferrocene

Mediator Azurine ferrocene boric acid

[66]

0.85 mM h
1

5.5% conversion of 10 mM xylitol in 10 h

40% conversion of 2 mM succinate in 6 h

[66]

[67]

[67]

[65]

[5]

[64]

[63]

[63]

[63]

[60]

Ref. [62]

23% conversion of 390 mM of the substrate in ~17 h 0.008 mM h
1 (27 g L
1 day
1) ee > 99.9% tofFcCOOH 12.9 h
1 0.89 mM h
1

100% conversion of 10 mM pethylphenol, ee 93% (optical purity) and 99% (GC analysis) 3.7% conversion of 200 mM xylitol in 3 weeks, ttnenzyme 208,720, ttnmediator 15

~70% conversion of 10 mM pethylphenol, ee 88%

84% conversion of 0.152 mmol p-cresol in 17 h, ttnenzyme 130,000, ttnmediator 66 95% conversion of 10 mM p-cresol, ttnenzyme 400,000, ttnmediator 500

Comments 85% conversion of 0.125 mmol p-cresol

Enzyme-Based Electrobiotechnological Synthesis 97

(S)-Epoxidized styrene derivatives (S)-Epoxidized styrene derivatives 2,5Dimethylphenpol 2β-Hydroxy-1,8cineole p-Nitrophenolate

Styrene and derivatives

Styrene and derivatives

p-Xylene

pNitrophenoxydodecanoic acid

1,8-Cineole

2,3Dihydroxybiphenyl

2-Hydroxybiphenyl

2-hydroxybiphenyl-3monooxygenase (HbpA) from Pseudomonas azelaica Styrene monoxygenase (StyA) from Pseudomonas sp. VLB 120 Styrene monoxygenase (StyA) from Pseudomonas sp. VLB 120 Cytochrome P450 BM3 from Bacillus megaterium P450cin from Citrobacter braakii Cytochrome P450 BM3 from Bacillus megaterium

2β-Hydroxy-1,8cineole

Product ω-Hydroxy lauric acid ω-Hydroxy lauric acid ω-Hydroxy lauric acid

1,8-Cineole

Lauric acid

Lauric acid

Substrate Lauric acid

P450cin from Citrobacter braakii

Enzyme Fusion protein rFP450 [mRat4A1/mRatOR]L1 Cytochrome P450 BM3 from Bacillus megaterium Cytochrome P450 BM3 from Bacillus megaterium

Table 6 Electrobiotechnological synthesis reactions with monooxygenases

CoSep

CoSep, [Cp*Rh(I)(bpy) H]+ Mediator screening

None

None

CoSep, phenosafranine (PSF), safranine T (SAF), FAD and FMN [Cp*Rh(bpy)Cl]Cl

Mediator Cobalt(III)sepulchrate trichloride Cobalt(III)sepulchrate trichloride 1,10 -Dicarboxycobaltocene

[81]

0.3–1.3 mM h
1

[84]

[83]

[82]

[80]

0.07–0.22–mM–h
1 ee 98.1–99.9%

0.45 mM h
1 (CoSep, 2 mL scale) 0.01 mM h
1 (CoSep, 100 mL scale)

[79]

1.10 mM h
1 (204 mg h
1 L
1)

[78]

[77]

[76]

110 nmol nmol(P450)
1 min
1 16.4 nmol nmol(P450)
1 min
1, 0.99 mM h
1 (electron transfer to flavin) 1.8 nmol nmol(P450)
1 min
1, 0.11 mM h
1 (electron transfer to heme domain) 6.50 nmol nmol(P450)
1 min
1 cm
2, 0.78 mM h
1

Ref. [75]

Comments 0.039 mM h
1

98 L.M. Schmitz et al.

Enzyme-Based Electrobiotechnological Synthesis

99

Table 7 Electrobiotechnological synthesis without additional mediator catalyzed by peroxidases Enzyme Horse radish peroxidase (HRP) Lignine peroxidase (LiP) from Phanerochaete chrysosporium Chloroperoxidase (CPO) from Calariomyces fumago CPO from C. fumago

Substrate N,N-Dimethylaniline

Product N-Methylaniline

Comments 0.078 mM h
1

Ref. [88]

Veratryl alcohol

Veratraldehyde

3 mM h
1

[89]

Barbituric acid

5-Chlorobarbituric acid

conversion >96%, 0.17 mM h
1

[90]

Thioanisole

(R)-Methylphenyl sulfoxide

[92]

CPO from C. fumago

Thioanisole

(R)-Methylphenyl sulfoxide

CPO from C. fumago

Thioanisole

(R)-Methylphenyl sulfoxide

CPO from C. fumago

Methyl p-tolyl sulfide

(R)-Methyl p-tolyl sulfoxide

CPO from C. fumago

1-Methoxy-4(methylthio)benzene

(R)Methoxyphenyl methyl sulfoxide

CPO from C. fumago

N-MOC-L-methionine methyl ester

CPO from C. fumago

Monochlorodimedone

N-MOC-L-methionine methyl ester sulfoxide Dichlorodimedone

CPO from C. fumago

Thioanisole

8.92 mM h
1 (30 g L
1 day
1), ee 98.5%, ttnCPO 95,000 30.91 mM h
1 (104 g L
1 day
1), ee 98.5%, ttnCPO 145,000 22.29 mM h
1 (75 g L
1 day
1) , ttnCPO 123,000 76% conversion of 21 mM substrate, ee 93%, ttnenzyme 58,900 83% conversion of 21 mM substrate, ee 99%, ttnenzyme 64,400 60% conversion of 21 mM substrate, ttnenzyme 700 1.89 mM h
1 (9.5 g L
1 day
1), ttnCPO 203,100 ([CPO] 10 nM; electrode surface area 5.5 cm2; current densities 5.5 mA cm
2 ) 6.84 mM h
1 (23 g L
1 day
1), ttnCPO 83,600 ([CPO] 600 nM; electrode surface area 16.5 cm2; current densities 5.5 mA cm
2)

Methyl phenyl sulfoxide

[93]

[65]

[94]

[94]

[94]

[95]

[95]

(continued)

100

L.M. Schmitz et al.

Table 7 (continued) Enzyme CPO from C. fumago

Substrate Indole

Product Oxindole

CPO from C. fumago

Monochlorodimedone (MCD)

Dichlorodimedone

AaeUPO from Agrocybe aegerita

Ethylbenzene

1-Phenethyl alcohol

Comments 2.60 mM h
1 (8.3 g L
1 day
1), ttnCPO 39,000 ([CPO] 100 nM; electrode surface area 16.5 cm2; current densities 1.8 mA cm
2) ttnCPO 1,150,000 10.36 mM h
1 (52 g L
1 day
1) ttnAaeUPO 400,000 8.52 mM h
1 (25 g L
1 day
1)

Ref. [95]

[96]

[97]

Table 8 Selected results of the reduction of NADP+ with new mediators (adapted from [121]) Ligand Bpy 4,40 -Dimethoxy-bpy 4,40 -Dimethyl-bpy 5,50 -Dimethyl-bpy

Ep vs Ag|AgCl (mV)
688
757
721
756

Tof (h
1) 33 113 48 116

Reaction rates (mM h
1) 1.67 5.67 1.67 4.04

Reaction conditions: V ¼ 200 mL, pH ¼ 7; 50 mM phosphate buffer, 0.05 mM mediator, 1 mM NADP+ Table 9 Conductivity of different buffer for cathodic reduction of oxygen in CPO-catalyzed sulfoxidation (adapted from [93]) Buffer Sodium acetate Sodium acetate Sodium acetate Sodium acetate Sodium citrate Sodium citrate Sodium citrate

Concentration (mM) 25 25 100 100 25 25 100

Sodium sulfate (mM) 0 50 0 50 0 50 0

Conductivity (mS m
1) 1.4 10.1 5.6 13.4 3 13.1 12.2

Table 10 Results of electroenzymatic sulfoxidation with CPO from Calariomyces fumago in the presence of ionic liquids Ionic liquid MMIM Me2PO4 BMIM MDEGSO4 MMIM MeSO4 BMIM MeSO4 EMIM EtSO4

Vol (%) 2 2 2 2 2

ee (%) 98 97 96 98 95

STY (g L
1 day
1) 34.9 41.5 45.5 67.5 75.4

Reduction of cyclohexanone using pentaerythitol tetranitrate reductase

Synthesis of 2,3-dihydroxybiphenyl utilizing 2-hydroxybiphenyl 3-monooxygenase Hydroxylation of lauric acid using a P450 monooxygenase

Product and used enzyme Synthesis of (S)-styrene oxide utilizing StyA

0.04 mM h
1/ 14.5 μM in 60 min 0.7 mM h
1/ 2.8 mM P450 monooxygenase fused with a NADPH-P450 reductase using CoSep for electron transfer Indium tin oxide coated glass electrode, MV for electron transfer

Electrochemical cofactor regeneration Reaction rate/ final product concentration Comment 3D RVC electrode, regeneration of 0.8 mM h
1/ FADH2, aeration with 22.8 mmol h
1 6.9 mM or 46.8 mmol h
1 oxygen 2.2 mM h
1/ 3.5 mM 1.1 mM h
1/ Carbon felt electrode, regeneration 2.2 mM of NADH

3.10 mM h
1/ 12.4 mM

0.04 mM h
1/–

2.4 mM h
1/ 16.9 mM

Using a twofold excess of NADPH

NADPH-P450 reductase for NADPH regeneration

FDH and formate for NADH regeneration

Enzymatic cofactor regeneration Reaction rate/ final product concentration Comment 6.2 mM h
1/ FDH and formate for NADH 65.1 mM regeneration, StyB for FADH2 regeneration

Table 11 Comparison of the reaction rates and productivities of processes with electrochemical and enzymatic cofactor regeneration

[56]

[75]

[79, 124]

Ref. [122, 123]

Enzyme-Based Electrobiotechnological Synthesis 101

102 Fig. 1 Electrochemical regeneration of NAD(P)+ from NAD(P)H. (a) Direct regeneration at the anode. (b) Indirect electrochemical regeneration via a mediator. (c) Indirect electroenzymatic regeneration via enzyme and mediator

L.M. Schmitz et al.

A

O

O NH2

NH2 +

N R

eanode

NAD(P)+

NAD(P)H

B

N R

O

O NH2

NH2 reduced mediator

oxidized mediator

N R

+

N R NAD(P)+

NAD(P)H eanode

C

O

O NH2 N R NAD(P)H

NH2 oxidized enzyme

reduced enzyme

reduced mediator

oxidized mediator

+

N R NAD(P)+

eanode

oxidations [26]. The complexation of phenanthrolinediones by transition metals accelerates hydride transfer and decreases the electron density of the ligand. For example, the complexation of 1,10-phenanthroline-5,6-dione with ruthenium [Ru (PD)3(ClO4)2] resulted in a 14-fold increase in catalytic tof for NAD+ regeneration [26]. This regeneration system was used for NAD+-dependent oxidation of cyclohexanol to cyclohexanone by alcohol dehydrogenase from horse liver (HLADH), with 75% conversion of 20 mM cyclohexanol and a cofactor tof of 28 h
1. To prevent product inhibition by cyclohexanone, meso-diols were used for further experiments [27]. The oxidation of 3.5 mmol meso-3,4-dihydroxymethylcyclohexene with HLADH using the mediator N-methyl 1,10-phenthrolinium-5,6dione tetrafluoroborate [PDMe(BF4)2] resulted in 99.5% conversion, with a tof of 35 h
1. The tof was determined over the total reaction time. In the same reaction, using Ru(PD)3(ClO4)2 as mediator, a lower tof of 15 h
1 was achieved. In this case, tof was determined from the conversion after 1 h reaction time. The authors concluded that PDMe(BF4)2 has a higher turnover because it has a lower molecular weight than Ru(PD)3(ClO4)2 and thus better diffusion. Iron 3,4,7,8tetramethylphenantroline [Fe(tmphen)3] is also suitable for NAD+ regeneration and was used for oxidation of 2-hexen-1-ol with yeast alcohol dehydrogenase (YADH) and of 2-butanol with alcohol dehydrogenase of Thermoanaerobium brockii (TADH) [28]. Product concentrations of 1.77 mM 2-hexenal in 1 h and 4.1 mM 2-butenal in 2.5 h were achieved.

Enzyme-Based Electrobiotechnological Synthesis

103

NADH can also be regenerated via enzyme-coupled electron transfer with diaphorase. Diaphorase, a flavoprotein, was used for the oxidation of cyclohexanol to cyclohexanone by HLADH [29]. Full conversion of 2 mM substrate was achieved within 10 h using the mediator 1,7-phenanthrolinequinone at low potentials of
100 mV versus the saturated calomel electrode (SCE). Diaphorase was also used in combination with the mediator ferrocene, co-immobilized on a graphite felt electrode coated with a thin poly(acryl acid) film [30]. The immobilization of diaphorase led to an increased yield of NAD+ from 50 mM NADH, achieving 100% conversion in 3 h compared with 20% conversion using nonimmobilized diaphorase. Alternatively, NADP+ can be regenerated using AMAPOR (artificial mediator accepting pyridine nucleotide oxidoreductase) with appropriate mediators such as carboxamidomethyl viologen (CAV) and anthraquinone 2-sulfonate (AQ-2-S). The enzyme-coupled regeneration of NADP+ with either CAV or AQ-2-S was coupled to the conversion of (2R,3S)-isocitrate catalyzed by an isocitrate dehydrogenase [31]. The conversion resulted in comparable initial reaction rates of 0.005 mM h
1. CAV and AQ-2-S are two-electron transfer mediators, which should be more effective overall because they avoid the high barrier of the NAD radical encountered with one-electron transfer agents [28]. Two-electron transfer mediators, however, have relatively slow electron transfer rates and are unstable at basic pH, which is typically the condition for dehydrogenase catalysis [32]. Screening of mediators in terms of reversibility, stability (even under basic conditions), and tof demonstrated that the one-electron mediator 2,20 -azino-bis(3-ethylbenzothiazoline6-sulfonic acid (ABTS) is especially suitable for NAD+ regeneration [33]. Therefore, recent studies are based on ABTS for NAD+ regeneration in electroenzymatic reactions. Using ABTS, a tof of 1,200 h
1 was obtained, with a ttn of 1,860 for the mediator and 93 for the cofactor, which was the highest reported ttn for NAD+ regeneration [33]. This system was used for GDH-catalyzed oxidation of glucose, achieving a reaction rate of 7.1 mM h
1. ABTS was also used in the synthesis of (3aR,7aS)-3a,4,7,7a-tetra-hydro-3H-isobenzofurane-1-one by HLADH, leading to an ee of 99.5% [32]. The synthesis required two oxidation steps from 48 mM meso3,4-dihydroxymethylcyclohex-1-ene, yielding 93.5% conversion within 46.5 h. The long reaction time was a result of the low specific enzyme activity of 4.2 U mg
1, mass transfer limitation of the mediator from the electrode to the bulk solution, and weak product inhibition. To circumvent product inhibition, a membrane reactor with solvent extraction can be used [34]. This system was applied for the oxidative resolution of racemic phenylethane-1,2-diol by glycerol dehydrogenase to yield enantiopure (S)-phenylethane-1,2-diol. In a recent study of the electroenzymatically driven conversion of glyceraldehyde into glyceric acid by aldehyde dehydrogemase from Deinococcus geothermalis, concentrations up to 17 mM with an ee of 88% [35] were achieved. Overall, the data show the suitability of coupling electrochemical regeneration of NAD+ to enzymatic processes. Omission of additional coenzymes and cosubstrates is a major advantage of electroenzymatic reactions.

104

2.2

L.M. Schmitz et al.

Carbonyl Reductions

The regeneration of NAD(P)H from NAD(P)+, which is required as cofactor for dehydrogenase-based reductions, necessitates selective hydride addition to form the enzymatically active reducing agent. Electrochemical regeneration is a two-step reaction in which one electron is first transferred to NAD(P)+ to generate a radical species (Fig. 2). After protonation, a second electron is transferred. Dimerization of the radicals formed in the first step leads to enzymatically inactive cofactors. Suitable electrode modification might reduce intermolecular radical coupling. Gold amalgam electrodes were modified with cholesterol to prevent cofactor dimerization [36]. The authors explained this observation by postulating simultaneous two-electron transfer and protonation. With this system, enzymatic reduction of 25.3 mM pyruvate to 18.2 mM D-lactate by D-lactate dehydrogenase (LDH) with direct NADH regeneration was carried out. A ttn of 1,400 for the cofactor and a conversion of 72% within 21 h were achieved (Table 2). Direct cofactor regeneration was also used for reduction of α-ketoglutarate catalyzed by glutamate dehydrogenase (GluDH) [37]. Vanadia-silica mixed gels, made of vanadium oxytripropoxide and tetramethyl orthosilicate, were used as encapsulation matrices for GluDH. Encapsulation in an electrically conductive sol–gel matrix improved enzyme stability and increased the conductivity of the reaction medium. The increased conductivity enhanced the efficiency of NADH regeneration during biosynthesis of L-glutamate. Conversion and ttn were three times higher compared to those obtained for conversion in less conductive silica gels. The complete conversion of 10 mM α-ketoglutarate into L-glutamate was achieved within 2.5 h with a ttn of 3,300 for NADH. In this case, 3 μM NADH was used and it was assumed that the dimerization was negligible for such low concentrations. The problem of dimer formation can also be overcome by regenerating NADH with enzymes acting as electron shuttles between electrode and cofactor. Reactions with methyl viologen (MV2+) in combination with diaphorase, lipoamide dehydrogenase (LipDH), ferredoxin reductase (FNR), and AMAPOR (Fig. 3) are summarized in Table 2. In electroenzymatic reductions of some ketones into cyclohexanol,

Fig. 2 Electrochemical regeneration of NAD(P)H is a two-step reaction. First, one electron is transferred to NAD(P)+ generating a NAD(P) radical. Subsequent dimerization of the radicals leads to enzymatically inactive cofactors

Enzyme-Based Electrobiotechnological Synthesis Fig. 3 Indirect electrochemical regeneration of NAD(P)H from NAD(P)+ using methylviologen as mediator and AMAPOR for electron transfer

105 O

O NH2

NH2

+

N R

N R NAD(P)+

NAD(P)H AMAPOR

+N

N

N+

+N oxidized methylviologen

reduced methylviologen

ecathode

(1S,2S)-(+)-2-methylcyclohexanol, and (1S,3S)-(
)-2-methylcyclohexanol with alcohol dehydrogenase (ADH), nearly 100% ee and 50% yield from a 50 mM solution were achieved [38]. For these syntheses, a graphite felt electrode was coated with a cation-exchange polymer and a poly(acrylic acid) layer on which MV2+, NAD+, diaphorase, and ADH were immobilized. With this arrangement, a current efficiency of approximately 100% for the reduction of cyclohexanone, ()-2methylcyclohexanone, and ()-3-methylcyclohexanone was obtained. The optimal concentrations of each component were determined for the mediated electrocatalytic reduction of NAD+ in conversion of pyruvate to D-lactate by LDH [39]. The ascertained composition comprised 1.5 U mL
1 diaphorase, 0.2 mM MV2+, and 0.3 mM NAD+, resulting in 80% conversion of 50 mM pyruvate within 96 h. An increase in NAD+ concentration up to 4.8 mM did not improve the reaction rate. The comparison between MV2+- and FAD-mediated NADH regeneration for (R)-mandelate synthesis from benzoylformate using benzoylformate reductase (BFR) revealed that FAD-mediated regeneration gave much better results [40]. With cofactor regeneration using FAD and lipoamide dehydrogenase (LipDH), a conversion of 95% of 50 mM substrate in 18 h was obtained. Regeneration with MV2+ and diaphorase enabled a conversion of 80% in 30 h. The decreased productivity was thought to result from toxicity effects causing inhibition of BFR activity. The natural electron carrier FAD was not considered as an alternative because of cost. MV2+ and LipDH have been combined together with an ADH for synthesis of cyclohexanol from cyclohexanone in a continuous process in a dialysis-membrane electrochemical reactor (D-MER) [41]. The mediator was identified as the ratelimiting step during this reaction as the conversion of 100 mM cyclohexanone could be increased from 26 to 45% and then to 65% with sequential addition of LipDH. MV2+

106

L.M. Schmitz et al.

was also used together with LipDH for synthesis of D-lactate by LDH [42]. Reduction of 150 mmol pyruvate was achieved within 9 days, yielding 81% D-lactate with an ee of 94%. Ferredoxin reductase (FNR) was applied for the same reaction to yield a conversion of 90% in 14 days. The regeneration system was also used for L-glutamate production from 200 mM α-ketoglutarate by GluDH, yielding 84% conversion in 7 days. The reactions in these systems are limited by the low activities of LipDH (~1 U mg
1) and FNR (~3 U mg
1), leading to a long conversion time. The combination of AMAPORs with MV2+ or cobalt(III) sepulchrate trichloride (CoSep) revealed that MV2+ is beneficial for NADPH regeneration in the conversion of oxoglutarate into (S)-glutamate by GluDH [31]. For comparison, 0.003 mM h
1 were achieved with MV2+ as mediator and 0.0001 mM h
1 with CoSep. The reaction rates of GluDH in combination with AMAPOR and CoSep were reduced because of the low AMAPOR activity, which was only 3.5% of the activity obtained with MV2+. The reduction of α-ketoglutarate to L-glutamate by GluDH was coupled with a second enzyme, a hydrogenase from Alcaligenes eutrophus, for NADH regeneration [43]. L-Glutamate production was successfully performed at analytical scale with a tof of 450 h
1 in a thin-layer reactor. The reaction was then transferred from analytical scale to preparative scale in a stirred cylindrical reactor with divided halfcells. The complete conversion of 300 mM α-ketoglutarate was achieved within 5 h with a tof of approximately 200 h
1. The hydrogenase seemed to be less stable in the preparative scale system, representing the limiting step in the reaction cascade. It was concluded that the impaired stability was caused by the larger possible distance between electrode surface and mediator, because previous studies reported increased stability and activity for hydrogenases in close proximity to platinum electrodes as a result of protein reduction [44]. The asymmetric electroreduction of ketones and aldehydes by ADH was achieved with MV2+ and FNR or diaphorase for regenerating NADPH or with acetophenone [45]. Acetophenone is a mediator that can be regenerated by ADH. Electroreduction with ADH and acetophenone does not require additional enzymes for NADPH regeneration. The reaction is nevertheless more efficient using MV2+ and FNR or diaphorase, leading to doubled conversion and ee values of >99%. In recent decades, (2,20 -bipyridyl)rhodium complexes [Rh(bpy)] have received great attention as mediators enabling indirect electrochemical reduction of NAD(P)+ to NAD(P)H. In contrast to the described electroenzymatic reductions, the application of Rh(bpy) for NAD(P)H regeneration does not need an additional enzyme for electron transfer. Rh(bpy) complexes were the first mediators that fulfilled all the requirements for selective NAD(P)+ reduction postulated by Steckhan [7]. These conditions include the following [5]: 1. The mediator must transfer two electrons in one step or a hydride ion. 2. Mediator activation must be possible at potentials less negative than
0.9 V versus SCE to prevent cofactor dimerization. 3. The mediator must transfer electrons to NAD(P)+ but not to the substrate. 4. Only the enzymatically active 1,4-isomer of NAD(P)H must be formed.

Enzyme-Based Electrobiotechnological Synthesis

107

Rh(bpy) is an interesting catalyst for cofactor recycling systems because it is highly versatile, being able to reduce NAD(P)+ as well as FAD using either formate or electrons [46]. Steckhan and coworkers used Rh(bpy) complexes to study the reduction of 4-phenyl-2-butanone to (S)-4-phenyl-2-butanol by NADH-dependent HLADH and Rhodococcus sp. [5]. The mediator Cp*Rh (4-ethoxy-methyl-2,20 -bpy)L (where Cp* refers to C5Me5, and L to ligand) and several types of water-soluble polymer-bound versions of the mediator were applied in an electrochemical enzyme membrane reactor, achieving roughly 70% conversion of 10 mM 4-phenyl-2-butanone in 5 h. Rh(bpy) complexcoupled NADH regenerations for electroenzymatic syntheses are summarized in Table 3. One of the first examples was the electroenzymatic combination of indirect NADH regeneration using [Rh(bpy)2]+ with enzymatic reduction of 1.12 mmol cyclohexanone by HLADH [47]. Conversion of 26% into cyclohexanol was achieved with a ttn of 2.9 for the cofactor and 1.2 for the mediator. The small amount of cofactor regeneration arose from cathode passivation caused by formation of [Rh(bpy)2(H2O)2]Cl or [Rh(bpy)2(OH)2]Cl, thus decreasing the current during synthesis. The current increased when electrodes were cleaned mechanically or with a short pulse. Enhanced cofactor regeneration efficiency was determined by testing different NAD+/mediator ratios and potentials [48]. Ratios of 10:1 and potentials of
0.6 V were suggested. Under these conditions, [Cp(Me)5Rh(bpy) Cl]Cl was regenerated at a turnover rate of 5 h
1. The regeneration of NADH by [Cp(Me)5Rh(bpy)Cl]Cl enabled the electroenzymatic-coupled synthesis of 14 mM D-lactate from 20 mM pyruvate in 3 h. New types of reactors were developed to overcome mass transport limitations by localizing the enzymes in close proximity to electrodes using a dialysis membrane [49]. Substrates were supplied and products removed by a solvent flow tangential to the membrane. The compounds were transported through the membrane by the concentration gradient. This reactor was applied for the synthesis of cyclohexanol from 100 mM cyclohexanone by HLADH with [Cp(Me)5Rh(bpy)Cl]Cl for NADH regeneration; a substrate conversion of 92% in 45 days was achieved [49]. A significant improvement was obtained by using ultrafiltration membranes and an orthogonal flow, whereby water was preferentially transported and other compounds were held back by the membrane. Products were directly extracted from the enzyme solution, preventing product inhibition. With this system, complete conversion was achieved within 3 days with a reaction rate of 1.38 mM h
1 (63 nmol h
1 U
1 cm
2). The reduction of acetophenone to (R)-phenylalcohol by ADH from Lactobacillus brevis (LbADH) with electrochemical regeneration of NADPH mediated by a rhodium complex showed low enzyme stability [50]. Interestingly, the addition of a second protein such as bovine serum albumin (BSA) led to high ttn of 75,000 for the enzyme and a reaction rate of 4.8 mM h
1 (STY of 14 g L
1 day
1). The authors assumed that the previous low stability was caused by enzyme adsorption on porous carbon electrodes and subsequent inactivation. The competing adsorption of BSA

108

L.M. Schmitz et al.

on the electrode reduced the inactivation of ADH. Alternatively, ADH was immobilized, resulting in a lower reaction rate of 3.1 mM h
1 (STY of 9 g L
1 day
1) but enhanced utilization because of its reusability. Another possibility for improving reaction aspects was the separation of catalyst and the product/substrate reservoir phase by adding an organic phase for product and substrate extraction. This approach resulted in a lower reaction rate of about 1.0 mM h
1 (STY of 3 g L
1 day
1), but demonstrated the potential of two-phase systems because the cofactor and mediator could be utilized several times. In situ product extraction was also carried out for the electroenzymatic reduction of rac-3-methylcyclohexanone to (1S,3S)-3-methylcyclohexanol by a TADH from Thermus sp. using [Cp*Rh(bpy)(H2O)]2+ for NADH regeneration in a flow-through electrode [51]. The prevention of TADH inhibition using octane in the cathodic compartment improved the productivity twofold. The final product concentration was 11.6 mM in 10 h, with a productivity of 1.1 mM h
1. Although there are considerable advantages in using Cp*Rh(bpy), a major drawbacks is its low stability in the presence of enzymes (amino acids). Hildebrand and coworkers solved this stability problem by spatial separation of mediator and enzymatic synthesis reaction by using a membrane that selectively retained polymer-bound Cp*Rh(bpy) [52]. The system was used for synthesizing pchloro-(R)-phenylethanol from p-chloroacetophenone with immobilized LbADH, achieving a reaction rate of 0.42 mM h
1 and conversion of 90%, giving the first example of a Cp*Rh(bpy) process with total recovery of enzyme activity.

2.3

Ene Reductions

Asymmetric hydrogenations are significant reactions for synthesizing chiral building blocks, amino acid derivatives, fragrances, terpenoids, etc. Enzymes catalyzing asymmetric hydrogenation and, hence, generating up to two stereogenic centers, are known as ene-reductases. They are subdivided into four enzyme classes; old yellow enzymes (OYEs), enoate reductases, medium-chain dehydrogenases/reductases (MDRs), and flavin-independent short-chain dehydrogenases/reductases (SDRs). Ene-reductases require NAD(P)H as cofactor for hydride donation [53]. For the enoate reductases and OYE enzyme family, electroenzymatic methods for cofactor recycling have been developed and combined with synthesis reactions (Table 4). The synthesis of (2R)-2-methyl-3-phenylpropionate by an enoate reductase from Clostridium tyrobutyricum was combined with MV2+-mediated NADPH regeneration at a mercury cathode [54]. A conversion of 95% of 80 mM substrate in 12 h was achieved. The same enoate reductase was immobilized, coupled with MV2+mediated NADPH regeneration, and used for stereospecific reduction of (E)-2methyl-3-phenyl-2-propenoate and (E)-2-methyl-2-butenoate to their (R)-enantiomeric products [55]. The enzymes were immobilized either on cellulose filters in a reactor outside the electrochemical cell or immobilized directly on the carbon felt

Enzyme-Based Electrobiotechnological Synthesis

109

electrode; the enzyme half-lives in these two systems were 150 h and 350 h, respectively. The ee of the obtained products was over 98% for both setups. More recently, pentaerythritol tetranitrate reductase (PETNR) and thermophilic ene-reductase (TOYE) were used for reduction of unsaturated substrates [56] such as ketones, nitroalkenes, aldehydes, and carboxylic acids. Both PETNR and TOYE in combination with the mediator MV2+ were used for the reduction of 2-cyclohexen-1-one, ketoisophorone, trans-cinnamaldehyde, N-phenyl-2methylmaleimide, and 2-methyl-pentenal in the presence and absence of NADPH. Reduction rates were 15–70% of those obtained when NADPH was present. For example, a reaction rate of 1.25 mM h
1 was achieved for the reduction of 2-cyclohexen-1-one into cyclohexanone using TOYE without NADPH supply. By comparison, a reaction rate of 2.27 mM h
1 was obtained for the same reaction catalyzed by TOYE using [Cp*Rh(bpy)(H2O)]2+ for NADPH regeneration [57]. The omission of NADPH for electroenzymatic reactions is beneficial for saving the cost of cofactors, but is not advantageous in terms of productivity.

3 Oxidases Oxidases are mainly flavin-dependent enzymes catalyzing oxidation reactions. Unlike the enzyme classes described in Sect. 2, they use FAD instead of NAD(P)+ as cofactor. FAD is strongly bound to the enzyme and does not allow diffusion to the electrode; it can be regenerated by reduction of oxygen to hydrogen peroxide [58]. The reaction is therefore typically accompanied by hydrogen peroxide formation as byproduct, resulting in oxidase stability problems. An approach to counteract the effects of high hydrogen peroxide concentration is the concurrent addition of a catalase [58, 59]. An alternative method for cofactor regeneration without catalase is electrochemical cofactor regeneration under anaerobic conditions. Due to the central location of the active site and the FAD binding site, direct electron transport is not possible. Therefore, electron transfer to FAD by mediators is necessary. It has been shown that FAD can be regenerated using ferrocene derivatives as mediators [60, 61]. Table 5 gives an overview of published processes. The first example describes the oxidation of p-cresol into phydroxybenzaldehyde with the p-cresol methylhydroxylase (PCMH) from Pseudomonas alcaligenes [62]. Ferroceneboronic acid combined with a gold electrode was used as mediator. The conversion was 85% from 0.125 mmol inserted p-cresol. The same reaction was carried out using PEG-20000 ferrocene as mediator and a graphite electrode, achieving 84% conversion of 0.152 mmol p-cresol [60]. A continuous regeneration system for FAD was applied in the next examples because the reaction could be controlled more precisely [63, 64]. To guarantee a continuous process, the enzyme and the polymer-modified mediator have to be held back in the reaction vessel, whereas the product and substrate pass through. These requirements were fulfilled by the use of an electrochemical enzyme membrane reactor (EEMR). In an EEMR, an ultrafiltration membrane is linked to an

110

L.M. Schmitz et al.

electrochemical cell with a carbon felt working electrode to guarantee retention of the enzyme and mediator in the flow cycle. As before, the oxidation of p-cresol to p-hydroxybenzaldehyde by PCMH was performed in the EEMR [63]. The enzyme reached a ttn of 400,000 and the system was still active after 10 days. Product formation in the continuous process could be controlled by selection of the residence time, and a 95% conversion of 10 mM pcresol to p-hydroxybenzaldehyde was obtained at a residence time of 3 h. At lower residence times, the main product of the reaction was the intermediate phydroxybenzylic alcohol. Using PCMH, the oxidation of 4-ethylphenol to the chiral intermediate (S)-1-(4-hydroxyphenyl)-ethanol was obtained, with a conversion of 70% and an ee of 88%. A similar enzyme, p-ethylphenol methylene hydroxylase (EPMH), catalyzing the oxidation of 4-ethylphenol to (R)-1-(4-hydroxyphenyl)ethanol [63], was used in a second experiment, giving 100% conversion of 10 mM of the substrate. The same continuous reaction setup was used for the oxidation of xylitol to Lxylose by the galactose oxidase (GOase) from Fusarium sp. [64]. The ferrocene mediator was modified by the addition of polyethyleneglycol (PEG) to avoid washing-out from the reactor. Using the continuous system, a conversion of 3.7% from 200 mM xylitol was reached within 520 h (~3 weeks). The ttn of the GOase was 208,720. Polymer-bound ferrocenes were also used as mediators in the oxidation of L-glycerin-3-phosphate to dihydroxyacetone phosphate (DHAP) in an electrocatalytic batch process. A 23% conversion from 390 mM of the substrate was obtained within ~17 h [5]. To further improve electroenzymatic syntheses, the use of ionic liquids as additives can be favorable because they enhance the solubility of substrates [65]. Additionally, ionic liquids increase the conductivity of the reaction medium and can stabilize the catalyst. One of the first examples of using ionic liquids in electroenzymatic synthesis was the resolution of DL-methionine catalyzed by a Damino acid oxidase (D-AAO). The addition of 10% (v/v) of the ionic liquid [MMIM] [Me2PO4] resulted in a 1.5-fold increase of methionine racemate resolution productivity, from a reaction rate of 0.005 mM h
1 (STY of 18 g L
1 day
1) to 0.008 mM h
1 (STY of 27 g L
1 day
1). L-Methionine was formed with an optical purity (ee) of over 99.9%. The reduction of keto acids to amino acids is the reverse reaction performed by the FAD-dependent enzyme AAO. Normally, this enzyme oxidizes amino acids to imino acids, which spontaneously hydrolyze to the corresponding alpha-keto acids. The reaction can be reversed by applying a redox potential more negative than that of FAD by using the mediator 1-aminopropyl-10 -methyl-4,40 -dipyridinium iodide (ADPy). The mediator was immobilized on a glassy carbon electrode. With this system, 0.89 mM h
1 of D-alanine and 0.85 mM h
1 of L-phenylalanine were synthesized from pyruvic acid and phenylpyruvic acid [66]. As an alternative, the direct electrochemical regeneration of FAD by a catalytic electrode without electron mediator is described in the following example [67]. This system was applied for the oxidation of xylitol to xylose and the oxidation of succinate to fumarate. For the oxidation, cell extracts from Escherichia

Enzyme-Based Electrobiotechnological Synthesis

111

coli and Saccharomyces cerevisiae were used as biocatalysts. The reaction was carried out in a two-compartment electrochemical reactor with a graphite-Cu(II) electrode as anode and a graphite-Fe(III) electrode as cathode. Using direct electrochemical regeneration, the conversion of 2 mM succinate to fumerate was 40% and for 10 mM xylitol to xylose 5.5%, within 6 h and 10 h, respectively. It is noteworthy that these systems only work in cases where the FAD can diffuse relatively freely between electrode and enzyme, as was the case in the presented study. One millimole of FAD was added to the basal reaction mixture and the reaction started by adding the crude cell extracts [67].

4 Monooxygenases Monooxygenases are versatile biocatalysts that catalyze a broad range of reactions, such as hydroxylations, epoxidations, halogenations, and Baeyer–Villiger oxidations [68–70]. All these reactions include the transfer of one atom of molecular oxygen to X–H bonds while the other atom is simultaneously reduced to water. Therefore, monooxygenases are also referred to as mixed function oxygenases. Monooxygenases use a wide range of cofactors, such as transition metals copper (Cu2+) (e.g., dopamine β-hydroxylase (DβM) [71]) and iron (Fe2+) (e.g. P450 monoxygenases [72]) or organic cofactors such as pterins (e.g., phenylalanine 4-monooxygenase (PheH) [73]) and flavins (e.g., 2-hydroxybiphenyl 3-monooxygenase (HbpA) [74]). The electrons needed for the reduction typically derive from the reducing agent NAD(P)H. As for the other NAD(P)H-dependent enzymes, several methods of cofactor regeneration have been developed, including chemical, enzymatic, photochemical, and electrochemical approaches. Some electrochemical cofactor regeneration approaches are described in Table 6. The first example of a monooxygenase-based electrobiotechnological synthesis is the electrocatalytically driven ω-hydroxylation of lauric acid using electrochemical reduction with the organometallic mediator CoSep. In this system, the purified fusion protein rFP450(mRat4Al/mRat-OR)L1, a P450 monooxygenase fused with a rat NADPH-P450 reductase, was able to acquire two electrons for reductive oxygen activation from CoSep instead of the natural cofactor NADPH. Applying this setup, the product formation rate of 0.039 mM h
1 was comparable with that obtained using NADPH (0.042 mM h
1) [75]. CoSep was also used for the hydroxylation of lauric acid with the cytochrome P450 monoxygenase BM3. This led to a hydroxylation rate of 110 nmol nmol(P450)
1 min
1 compared with a rate of 900 nmol nmol(P450)
1 min
1 when using CoSep instead of NADPH [76]. As aggregation and the formation of reactive oxygen species (also referred to as uncoupling) are the main disadvantages of using CoSep, 1,10 -dicarboxycobaltocene was used as an alternative mediator for cytochrome P450 monoxygenase BM3 [77]. 1,10 -Dicarboxycobaltocene was used to reduce FAD and FMN in the reductase domain or the heme iron directly, yielding hydroxylation rates of 16.5 nmol nmol(P450) min
1 (reaction rate of 0.99 mM h
1) and 1.8 nmol nmol(P450)
1

112

L.M. Schmitz et al.

min
1 (0.11 mM h
1), respectively [77]. The results revealed that with both 1,10 -dicarboxycobaltocene and CoSep [rates for CoSep were 37.8 nmol nmol(P450)
1 min
1 (2.27 mM h
1) and 2.2 nmol nmol(P450)
1 min
1 (0.132 mM h
1)], rates in the same order of magnitude can be achieved. CoSep and other mediators such as phenosafranine (PSF), safranine T (SAF), FAD, and FMN were equally adequate mediators for the artificial electron transfer to P450cin via the redox partner protein cindoxin (CinC) [78]. The highest product formation of ~6.50 nmol nmol(P450)
1 min
1 cm
2 (~0.78 mM h
1) 2β-hydroxy-1,8-cineole produced from the natural substrate 1,8-cineole was achieved using cobalt sepulchrate. Applying PSF and SAF as mediators, even direct electron transfer to the heme center of P450cin in the absence of CinC was possible. The 2-hydroxybiphenyl-3-monooxygenase (HbpA) catalyzes the ortho-hydroxylation of several α-substituted phenol derivatives. The enzyme was coupled to indirect electrochemical regeneration of NADH using a [Cp*Rh(bpy)Cl]Cl complex as mediator. The main issue with [Cp*Rh(bpy)Cl]Cl mediator usage was the formation of hydrogen peroxide. Depending on the rate of aeration, molecular oxygen reacted with the hydridorhodium complex or molecular oxygen was directly reduced at the cathode. Despite HbpA stability problems caused by electrode adsorption and locally high hydrogen peroxide concentrations, a reaction rate of 1.10 mM h
1 was achieved [79]. Direct unmediated electron transfer to FAD, bypassing the regeneration of NADPH, was performed in the (S)-epoxidation of styrenes by the flavin-dependent styrene monooxygenase (StyA) from Pseudomonas sp. VLB120 [80]. Applying direct electrochemical regeneration, the complex natural system, consisting of three enzymes and the two cofactors, could be reduced to only StyA and FAD [80]. However, the epoxidation rates of direct electrochemical regeneration (0.07–0.22 mM h
1 ) were lower than for the enzymatic system. The main reason was the low activity of the biocatalyst StyA, which depended on the degree of aeration. Reduced flavins can be directly reoxidized by oxygen, resulting in undesired reduction reactions that concurrently lower the enzymatic epoxidation rate. To increase the cofactor regeneration rates per reaction volume, a reactor design was proposed that separates the high concentration of molecular oxygen from the cathode compartment [81]. The developed continuous flow-through reactor provided electrodes with a maximal volumetric surface. Using this reactor setup for direct electrochemical regeneration of FAD in the StyA system, reaction rates of 0.3–1.3 mM h
1 were achieved. However, the oxygen dilemma of oxygen reduction by reduced mediators is still a problem when electrochemical synthesis is performed with enzymes dependent on molecular oxygen. Tosstorff et al. investigated the conversion of p-xylene to 2,5-dimethylphenpol by the P450 BM3 monooxygenase using both cobalt sepulchrate and the more oxygen stable mediator [Cp*Rh(I)(bpy)H]+ [82]. Oxygen depletion was observed in the presence of the mediator, indicating that oxygen is able to reoxidize the reduced mediator, leading to uncoupling and therefore a loss in energy efficiency. The effect of the oxygen dilemma was also shown by a drop in the product formation rate, with 0.01–0.45 mM h
1 for the scale-up reactor with an increased electrode surface area compared to the small-scale system. Stability

Enzyme-Based Electrobiotechnological Synthesis

113

against oxygen could not be increased by applying the rhodium mediator [Cp*Rh(I) (bpy)H]+, resulting in even lower mediator efficiencies than for cobalt sepulchrate. Many different mediators are reported to transfer electrons artificially for cofactor generation with monooxygenases. To identify suitable mediators for an efficient electrochemically driven enzymatic process, a computational docking algorithm was developed by Str€ohle et al. [83]. With this algorithm, they were able to predict suitable mediators for P450cin with respect to binding conformation and binding tightness. The received data were compared with experimentally obtained product formation rates and the values fitted 70% of the prediction rate [83]. Despite the prediction of mediator choice, many parameters, such as applied potential or concentrations of mediator and enzyme, must be optimized to yield higher product formation. To optimize electroenzymatic processes, a screening system in microtiter plate format (electrochemical microtiter plate, eMTP) was developed by Ley et al. [84]. Applying the eMTP, it was possible to elucidate the influence of certain reaction parameters such as P450 BM-3 concentration, mediator concentration, and applied potential on the productivity in one run. Furthermore, eMTP could be successfully applied for P450 BM-3 mutein screening.

5 Peroxidases In contrast to monooxygenases, peroxidases do not depend on cofactors such as NAD(P)H or FADH. Peroxidases contain a heme, vanadate, or manganese prosthetic group and generally use hydrogen peroxide as electron acceptor to catalyze the oxidation of various substrates. The main negative aspect of using hydrogen peroxide as electron acceptor is the reduced stability of the enzyme if exposed to high concentrations. However, several approaches circumvent high local concentrations of hydrogen peroxide, including sensor-controlled feeding of hydrogen peroxide [85], in situ synthesis of hydrogen peroxide in combination with a second enzyme [86], and electrochemical generation of hydrogen peroxide [87]. Because sensor-dosed hydrogen peroxide addition requires high sensitivity of the sensor, and coupling of a second enzymatic reaction always results in coproduct formation, electrochemical generation of hydrogen peroxide is the method of choice. This system provides the benefit of adding a controlled concentration of hydrogen peroxide to the enzymatic reaction by varying the electrochemical parameters [87]. Hydrogen peroxide is directly formed by the cathodic reduction of oxygen, making the system very simple because no mediator is needed. The reaction occurs at certain electrodes, such as mercury, gold, and carbon. Carbon is the favored material for use in electrobiotechnological syntheses because it is safe and economically attractive. Some examples of electrobiotechnological syntheses using peroxidases and their productivities are presented in Table 7. The first example is N-demethylation of N,N-dimethylaniline (DMA) with free horseradish peroxidase (HRP) [88]. By increasing the applied current, higher hydrogen peroxide generation could be achieved, also resulting in increased

114

L.M. Schmitz et al.

formaldehyde formation rates. Using this system, an initial formaldehyde formation rate of 0.08 mM h
1 with an applied current of 0.08 mA was obtained. In another approach, veratryl alcohol is oxidized to veratraldehyde using the lignin peroxidase [89]. The reaction is performed in a divided batch reactor, where cathode and anode are separated by a cation-exchange membrane. The rates of veratraldehyde formation with electrochemically driven hydrogen peroxide generation were compared with the rate of the natural hydrogen peroxide-driven method (biochemical method). The biochemical method resulted in a higher veratraldehyde formation rate compared with the electrochemical method. The initial oxidation rate of the biochemical method was 3.3 mM h
1 (2 mM H2O2, 6 mM substrate) compared with a rate of 3 mM h
1 (
0.4 V, 6 mM substrate) for the electrochemical method. The heme-dependent chloroperoxidase (CPO) from the filamentous fungus Caldariomyces fumago is a very versatile peroxidase, which, among other reactions, chlorinates barbituric acid to 5-chlorobarbituric acid [90]. The first attempt used a bioelectrolytic system consisting of three interconnected units: an electrolytic cell for hydrogen peroxide generation, a hollow-fiber membrane for biocatalysis, and an anion exchanger for product collection. After 24 h, 0.33 mM of 5-chlorobarbituric acid was produced in the bioelectrolytic system, with an initial reaction rate of 0.17 mM h
1. CPO is also known to mediate oxidations of dialkyl sulfides to form sulfoxides [91]. In situ electrogeneration was applied for the oxidation of thioanisole to (R)methylphenyl sulfoxide with high optical purity [92]. Using a batch reactor operation mode, a reaction rate of 8.92 mM h
1 (STY of 30 g L
1 day
1) with a ttnCPO of 95,000 could be reached. To enhance the productivity of (R)-methylphenyl sulfoxide formation, a three-compartment cell with a packed bed cathode (flowby electrode) and two anodes were used in follow-up studies using the same CPO [93]. With optimized reaction conditions, a reaction rate of 30.91 mM h
1 (STY of 104 g L
1 day
1) and a ttnCPO of 145,000 could be achieved, which is comparable to sensor-controlled dosing of hydrogen peroxide (35.66 mM h
1/120 g L
1 day
1). The presence of ionic liquids was shown to enhance the reaction process, as also shown for oxidases (see Sect. 3), by overcoming the main limiting factors [65]. For the oxidation of thioanisole to (R)-methylphenyl sulfoxide [65], the presence of 2% of the ionic liquid 1-ethyl-3-methylimidazolium ethylsulfate [EMIM][EtSO4] improved productivity by a factor of up to 4.2, from 5.35 mM h
1 (STY of 18 g L
1 day
1) to 22.29 mM h
1 (STY of 75 g L
1 day
1) in a nonoptimized beaker cell. Because electrochemical generation was successful for the formation of (R)methylphenylsulfoxide, it was also applied for the synthesis of other sulfoxides, such as (R)-methyl p-tolylsulfoxide, (R)-methoxyphenyl methyl sulfoxide, and NMOC(methoxycarbonyl)-L-methionine methyl ester sulfoxide [94]. The reactions were carried out in batch experiments resulting in conversions of 76% from 21 mM added methyl p-tolylsulfoxide, 83% conversion of (R)-methoxyphenyl methyl sulfoxide, and 60% conversion of N-MOC-L-methionine methyl ester sulfoxide, respectively.

Enzyme-Based Electrobiotechnological Synthesis

115

Despite the factors mentioned above, electrochemical hydrogen peroxide generation also suffers from a limitation in mass transport arising from the limited electrode surface. To improve the mass transport, a new reaction system based on a gas diffusion electrode (GDE) was developed. GDEs have a solid, liquid, and gaseous interface (where oxygen is reduced to hydrogen peroxide) that provides the advantage of a highly specific electrode surface. The reaction system of GDE was used for the chlorination of monochlorodimedone, the sulfoxidation of thioanisole, and the oxidation of indole catalyzed by CPO [95]. It was shown that both a larger electrode surface and lower current density, guaranteeing less enzyme inactivation, led to improved ttn and STYs. A reaction rate of 1.89 mM h
1 (STY of 9.5 g L
1 day
1) was achieved for the chlorination of monochlorodimedone, 6.84 mM h
1 (STY of 23 g L
1 day
1) for sulfoxidation of thioanisole, and 2.60 mM h
1 (8.3 g L
1 day
1) for the oxidation of indole. In another study, the GDE-based system was applied for the oxidative chlorination of monochlorodimedone (MCD) [96]. In this study, the applicability of the GDE-based reaction system for hydrogen peroxide-dependent enzymes was emphasized, leading to high STY and ttn. For the CPO, a ttn of up to 1,150,000 was achieved and a reaction rate of up to 10.36 mM h
1 (STY of 52 g L
1 day
1), which was dependent on the electrochemically produced hydrogen peroxide. The heme-thiolate peroxygenase AaeUPO from the fungus Agrocybe aegerita (also known as unspecific peroxygenase AaeUPO) catalyzes the hydroxylation of ethylbenzene into 1-phenethyl alcohol [97]. An electrochemical hydrogen peroxide supply based on a GDE working electrode led to a ttn of up to 400,000 mol1-phenethyl
1 and a reaction rate of up to 8.52 mM h
1 (STY of 25 g L
1 day
1), alcohol molAaeUPO which was high compared with the productivities of AaeUPO with nonelectrochemical hydrogen peroxide supply.

6 Reaction Engineering Considerations As described in the previous sections, many improvements in technical concepts have been achieved during the last few years and increased the applicability of oxidoreductases for organic syntheses. The improvements were mainly made in four technical aspects, namely medium modification, compartmentation, catalyst modification, and nonnatural cosubstrates. It is important to develop the right combination of approaches for a certain reaction to make it a useful synthetic tool. The main considerations for three basic reaction engineering strategies are outlined in this section. First, we discuss the use of compartmentation to reduce enzyme stability problems in the presence of mediators. The second part discusses reaction media modifications to target better overall productivity. Finally, reactor design concepts are described with regard to improvements in mass transfer between electrolyte and electrode surface area.

116

6.1

L.M. Schmitz et al.

Enzyme Stabilization by Compartmentation

A general problem in electroenzymatic processes is the interface between the enzyme and the mediator because the presence of the mediator often causes enzyme stability problems. The attempts to overcome low enzyme stability and the reasons for complete separation of enzyme and mediator via compartmentation are explained using the mediator Cp*Rh(bpy) as an example. Cp*Rh(bpy) is a versatile tool for reducing NAD+ and NADP+ as well as FAD [46]. A drawback of this mediator, when used in continuous regeneration of NADPH, is enzyme instability that results in a steady drop in conversion with longer operation times. One obvious side reaction between the rhodium mediator and enzymes is the interaction of sulfur-containing amino acids with the heavy metal. This is especially known for cysteine residues in proteins [98]. Additionally, nucleophilic side chains (especially of lysine, histidine, and arginine) are thought to contribute to enzyme deactivation by forming inactive adducts with the rhodium complex (Fig. 4) [99]. This was deduced from the fact that enzyme stability in the electroenzymatic process was somewhat higher when using an immobilized enzyme, where these side chains were chemically blocked as a result of immobilization. However, these findings could not be proven in a similar study [52]. The inactivation was quite severe. When incubating the enzymes LbADH and TADH in buffer in the presence of the mediator, the half-lives of both enzymes were reduced by roughly an order of magnitude (Fig. 5). The set of previously studied bipyridine ligands of the Cp*Rh(bpy) mediator system was not exhaustive [100]. Therefore, further rhodium complexes were synthesized and their potential studied with respect to catalytic efficiency in reducing NADP+ and their effect on enzyme stability. With the aim of establishing electroenzymatic systems with higher operational stability, it was especially interesting to find out whether the effect of modified bipyridine ligands on enzyme stability would be different from that of unmodified catalyst. Figure 6 shows the

Fig. 4 Hypothesis for enzyme inactivation by basic amino acid side chains

Enzyme-Based Electrobiotechnological Synthesis 80

117

buffer 0.2 mM Rh(bpy)

70

half-life time [h]

60 50 40 30 20 10 0

LbADH

TADH

Fig. 5 Residual alcohol dehydrogenase (ADH) activity after incubation with and without Cp*Rh (bpy) mediator (Hildebrandt and Lütz, unpublished results). Reaction conditions: V ¼ 1 mL, T ¼ 20 C, pH ¼ 7, 50 mM phosphate buffer, 3 mM MgCl2, 0.2 mM substituted Rh(bpy), 1 U mL
1 enzyme. Samples of 10 μL were periodically withdrawn from the incubation and analyzed for residual enzyme activity. Assay conditions: 11 mM acetophenone, 0.2 mM NADPH, detection of cofactor consumption at λ ¼ 340 nm. One unit is defined as the amount of enzyme converting 1 μmol of substrate per min. TADH ADH from Thermoanaerobium brockii, LbADH ADH from Lactobacillus brevis

selection of bipyridine ligands investigated. None of the synthesized complexes (lower two rows) were described in literature previously. From this screening using a robotic system for electrochemical measurements, four promising new mediators (4,40 -dimethyl-bpy, 4,40 -dimethoxy-bpy, 4,40 -dicarboxy-bpy, and 5,50 -dimethylbpy) were identified. The 4,40 -dimethoxy- and 5,50 -dimethyl-bipyridine complexes showed much higher catalytic activity and thus much higher reaction rates (Table 8). Unfortunately, incubation of the new mediators with LbADH under conditions similar to the previous study (see Fig. 5) showed that all of the mediators were even more detrimental to enzyme stability (Fig. 7). Several approaches for improving enzyme stability are described in literature. These include the use of immobilized LbADH to block nucleophilic side chains from the immobilization and crosslinking procedures [99], the addition of nonreactive protein BSA, and the use of coordinating buffers such as Bis-Tris [101]. However, none of these approaches could prevent deactivation of LbADH [52] and none finally resulted in an electroenzymatic process. This highlights the fact that the mechanism of inactivation is not fully understood. Recorded cyclic voltammograms [102, 103] of the rhodium mediator in the presence and absence of all proteinogenic amino acids led to the conclusion that both the enzyme and mediator are deactivated by the cross-reaction, especially in the presence of cysteine, histidine, and tryptophan [52]. For detailed information on mediator stability, the reader is referred to a study by Hildebrand and Lütz [52].

118

L.M. Schmitz et al.

No complex formation

Ph

Ph

O

O HOOC

C9H19

C9H19

COOH N

N

N

2,2’-bipyridine-5,5’-dicarboxylic acid

Br

O N

N

N

4,4’-bis(bromomethyl)-2,2’-bipyridine

HO

OH

N

4,4’-dinonyl-2,2’-bipyridine

OH OH

N

N

Cl

N

4,4’-diphenyl-2,2’-bipyridine

Br

N

2,2’-bipyridine-4-carbaldehyde

N

N

2,2’-bipyridine-4,4’-dicarbaldehyde

N

N

4-methyl-2,2’-bipyridine

N

2,2’-bipyridine-3,3’-diol

Br

Br

OH N

N

N

N

N

N

4-chloro-2,2’-bipyridine

2,2’-bipyridine-4,4’-diyldimethanol

N

N

4,4’-dibromo-2,2’-bipyridine

N

2,2’-bipyridin-5-ol

N

2,2’-bipyridine

Complex formation

H2N

NH2

Cl

N

N

HOOC

O N

N

2,2’-bipyridine-3,3’-dicarboxylic acid

N

N

4,4’-dimethyl-2,2’-bipyridine

N

N

5,5’-dimethyl-2,2’-bipyridine

COOH N

N

2,2’-bipyridine-4,4’-dicarboxylic acid

N

N

2,2’-bipyridine-4,4’-diol

HO3S N

OH

HO

N

4,4’-dinitro-2,2’-bipyridine

HOOC

4,4’-dimethoxy-2,2’-bipyridine

N

N

N

2,2’-bipyridin-4-amine

O

NO2

O2 N N

4,4’-dichloro-2,2’-bipyridine

COOH

N

N

N

2,2’-bipyridine-4,4’-diamine

H2N

Cl

N

N

2,2’-bipyridine-5-sulfonic acid

N

4,4’-di-tert-butyl-2,2’-bipyridine

Fig. 6 Library of investigated bipyridine ligands: light grey commercially available bpy ligands, dark grey synthesized ligands [121]. Upper two rows do not form the mediator complex with the rhodium precomplex, lower two rows form the desired rhodium mediator complex

Eight deactivating amino acids (two Cys, four His, and two Trp) are present in LbADH; therefore, inactivation is inevitable. Inactivation can only be prevented if mediator and enzyme are fully separated, allowing the cofactor only to shuttle the redox equivalents between the electrochemical regeneration reaction and the biotransformation (Fig. 8). Based on the inactivation mechanism elucidated above, full spatial separation by a membrane should enable a stable electroenzymatic process [104]. However, the mediator must be chemically modified to make it retainable by a membrane. This can be achieved by increasing the molecular weight through coupling the mediator to a water-soluble polymer or polycondensation between mediator and an additional bifunctional monomer. Several methods for synthesis of such polymer-enlarged homogeneous catalysts (Chemzymes) have been described [105–112]. Polycondensation of 2,20 -bipyridine-4,40 -di-aldehyde and α,ω-functionalized amino-PEG (molecular weight 6,000 g mol
1) was chosen to produce a polymeric mediator [52] with 26% electrochemical activity per rhodium center compared with the monomeric mediator. Using this polymer-enlarged mediator in the reactor setup [52] together with an immobilized enzyme preparation led to an electroenzymatic ketone reduction process with unprecedented enzyme and mediator stability. No

Enzyme-Based Electrobiotechnological Synthesis

119

Fig. 7 Stability of LbADH in the presence of different rhodium complexes (Hildebrandt and Lütz, unpublished results). Reaction conditions: V ¼ 2 mL, T ¼ 20 C, pH ¼ 7, 50 mM phosphate buffer, 3 mM MgCl2, 0.2 mM substituted Rh(bpy), 1 U mL
1 LbADH. Samples of 10 μL were periodically withdrawn from the incubation and analyzed for residual enzyme activity. Assay conditions: 11 mM acetophenone, 0.2 mM NADPH, detection of cofactor consumption at λ ¼ 340 nm. One unit is defined as the amount of enzyme converting 1 μmol of substrate per min

additional dosing of enzyme or mediator during the reaction was necessary. After the reaction, 100% of the enzyme activity and 86% of the mediator activity could be recovered. This corresponds to a ttn of 214 for the Cp*Rh(bpy)-based mediator, which is the highest ttn reported so far for an electrochemical activation of this mediator. This proves that spatial separation can successfully circumvent the mutual inactivation between enzyme and mediator and is therefore a generally applicable strategy when using the Cp*Rh(bpy) complex together with enzymes.

6.2

Reaction Media

Major issues in electroenzymatic syntheses are the overall productivity and the final product concentration [18]. The goal is to increase the productivity (which means increasing the applied current) and increase the substrate loading (which is limited by the substrate solubility in the medium). As the applied currents used in electroenzymatic systems are usually low, within the range of milliamperes, the ohmic losses are normally low. However, to reach technically relevant

120

L.M. Schmitz et al.

Fig. 8 (a) Spatial separation of mediator and enzyme by membranes. The cofactor is used to transfer redox equivalents between electrochemical and enzyme compartments. (b) Reactor setup using polycondensed 2,20 -bipyridine-4,40 -di-aldehyde with α,ω-functionalized aminopolyethylene glycol as mediator

productivities of around 100 g L
1 day
1 (STY), currents in the ampere range are usually applied. At high currents, the conductivity of the media might become relevant. A study of CPO-catalyzed sulfoxidation investigated the conductivities of suitable electrolyte solutions for cathodic reduction of oxygen. Sodium acetate and sodium citrate buffers showed good buffering capacities, and conductivities could be enhanced by adding the supporting electrolyte sodium sulfate (Table 9) [93]. In addition to conductivities for increased productivity, current efficiencies were also taken into account. Sodium acetate showed the best current efficiency for high cell voltages (Fig. 9). For application of electroenzymatic systems at the production scale, it might become relevant to further increase reaction media conductivities while simultaneously increasing substrate loading; however, these requirements often contradict each other. Although it is standard practice in biocatalysis to increase the solubility of hydrophobic substrates by adding organic solvents, either as water-miscible cosolvents or in two-phase systems, this is detrimental to electroenzymatic syntheses [65] because organic solvents typically have lower dielectric constants than

Enzyme-Based Electrobiotechnological Synthesis

121

1,0

current efficiency (-)

0,9 0,8 0,7 0,6 0,5 0,4 0,3 0,2 0,1 0,0 1,4

1,5

1,6

1,7

1,8

1,9

cell voltage (V) sodium sulfate

sodium citrate

sodium acetate

Fig. 9 Current efficiency as a function of cell voltage for different buffers. Reaction conditions: 3D cell, vfl 100 mL min
1, 50 mM sodium sulfate, 100 mM sodium citrate, 100 mM sodium acetate with 50 mM Na2SO4, graphite beads (diameter 1.0–1.4 mm, 900 cm2 geometric surface area); sodium sulfate buffer serves as a reference electrolyte

Fig. 10 Dependence of the conductivity of the buffer medium and solubility of thioanisole on the cosolvent concentration. Reaction conditions: 100 mM sodium acetate buffer, pH 5, room temperature (adapted from [93])

water and, thus, the addition of organic solvents usually leads to lower conductivities. This is demonstrated using the maximum solubility of thioanisole, a substrate in CPO-catalyzed sulfoxidation (Fig. 10). As mentioned, ionic liquids have recently emerged as additives for electroenzymatic syntheses. In addition to increased enzyme and (in some cases) cofactor stability, both substrate solubility and conductivity were increased through addition of ionic liquids, resulting in higher STYs. This is especially evident in the

122

L.M. Schmitz et al.

case of sulfoxidation. Results from various syntheses are summarized in Table 10. Compared to previous systems, where STYs in the range of 5–30 g L
1 day
1 were achieved, the addition of only 2% (v/v) of an ionic liquid can increase productivity more than twofold. STYs higher than 100 g L
1 day
1 can be obtained when using a three-dimensional (3D) cell [93, 113] or hydrogen peroxide dosing [85, 114]. As a second system, the deracemization of methionine was carried out using AAO from Trigonopsis variabilis with electrochemical regeneration of the FAD cofactor using a ferrocene mediator [65]. In this case, the STY could also be increased from 17.6 to 27.1 g L
1 day
1 by adding 10% (v/v) of MMIM Me2PO4. In a third study, the electrochemical reduction of NADP+ to NADPH was carried out using Cp*Rh(bpy) [100, 113, 115, 116] as mediator. By adding EMPY EtSO4, the STY was increased from 20 to 55 g L
1 day
1, which is about 2.7-fold. Interestingly, small amounts of an ionic liquid also stabilized the nicotinamide cofactors [65]. The half-life of NADPH upon incubation increased from 14.4 h in plain buffer to over 40 h in the presence of the ionic liquid. In these three examples, ionic liquids were applied in electroenzymatic syntheses for the first time [117]. They synergistically combine the function of conducting salt and cosolvent and can also stabilize the enzymes or cofactors. Thus, this is another option for converting poorly water-soluble substrates when using electrochemistry to supply the redox equivalents for an oxidoreductase.

6.3

Reactor Design

The main challenges in electroenzymatic syntheses beyond the laboratory scale are technical reactor limitations and low productivities. Reactor concepts with enhanced mass transfer between electrolyte and electrode and increased electrode surface area per reaction volume were established to improve cofactor regeneration rates and, thus, to accelerate the electroenzymatic reaction [118]. 3D electrodes provide large surface-to-volume ratios can be adapted providing short ionic pathways, which have to be overcome by diffusion. Enhanced mass transfer can be realized by using flow-through cells, where the electrolyte buffer flows vertically and the current flows horizontally through a packed electrode bed (Fig. 11). The electrode bed frequently consists of glassy carbon spheres, graphite beads, or reticulated vitreous carbon (RVC). Kochius and coworkers compared a 3D flow-through electrode system with a two-dimensional (2D) electrode in a stirred cell and revealed that the diffusion of mediators from 2D electrodes into the bulk solution was the rate-limiting step [33]. The reaction rate of GDH-catalyzed oxidation of glucose with ABTS as mediator for NAD+ regeneration was improved eightfold to a maximum of 7.1 mM h
1 using the 3D flowthrough electrode system. In other systems, reaction rates of up to 212.4 mM h
1 (STY of 1 kg L
1 day
1) of reduced cofactor NAD(P)H were achieved with a 3D cathode design using Cp*Rh(bpy), with a current efficiency close to 100% [113]. This high STY with ttn values of up to 400 was a result of mediator

Enzyme-Based Electrobiotechnological Synthesis

123

Fig. 11 Setup for a flowthrough cell. The electrolyte buffer flows horizontally through the packed electrode bed (anode), which is between an electrode net (cathode). Anode and cathode are separated by an ion-exchange membrane

adsorption on the carbon surface, which required low mediator concentrations for large electrode surfaces and very low mediator consumption. Immobilization of enzymes on the electrode also resulted in increased stabilities. The reaction rate of 0.2 mM h
1 for the electroenzymatic conversion of DOPA quinone to L-DOPA with tyrosine was, however, only half of the reaction rate when the free enzyme was applied [119]. The disadvantage of the lower reaction rate as a result of enzyme immobilization was compensated by an at least eightfold longer stability and reusability of the enzyme for up to 10 cycles. The ionic pathway can hence be reduced by positioning two electrodes on each side of a packed bed electrode. Using a flow-by electrode arrangement with graphite particles as cathode material and ion-exchange membranes to separate anolyte and catholyte compartments, hydrogen peroxide was produced via reduction of dissolved oxygen [93]. The electrogeneration of hydrogen peroxide was combined with the oxidation of thioanisole to (R)-methylphenylsulfoxide, which was catalyzed by CPO from C. fumago. After optimizing the composition and conductivity of the biotransformation buffer, reactions rates of up to 30.9 mM h
1

124

L.M. Schmitz et al.

Fig. 12 Setup for a flowthrough cell. Electrodes consisting of reticulated vitreous carbon are separated by an ion-exchange membrane. For in situ product extraction, a second organic phase is used in the cathodic compartment

(R)-methylphenylsulfoxide and STY of 100 g L
1 day
1 were achieved. These values are close to the STYs of other reported CPO-catalyzed reactions using different hydrogen peroxide dosing methods [85] and more than three times higher than using a batch electrolysis cell using a graphite felt electrode [90]. In situ product extraction was applied using octane as organic phase in the cathodic compartment (Fig. 12). Under these conditions, a reaction rate of 1.1 mM h
1 for (1S,3S)-3-methylcyclohexanol by a TADH from Thermus sp. was obtained using [Cp*Rh(bpy)(H2O)]2+ as mediator for NADH regeneration in a flow-through electrode [51]. The reaction rate was twice as high as that without in situ product extraction with octane, indicating the potential of the two-phase reactor concept. The reaction was, however, limited by adsorption of the mediator to the Nafion® ion exchange membrane, which reduced the cofactor regeneration capacity. NADH regeneration was not maintained throughout the entire electroenzymatic reaction because the mediator was deactivated earlier. The membrane for separating cathodic and anodic compartments was omitted by Varnicic and coworkers for the oxidation of glucose to gluconic acid [120]. In their system, all catalysts and mediator were immobilized on the electrodes to enable a membraneless reactor concept with only a single electrolyte (Fig. 13). On the

Enzyme-Based Electrobiotechnological Synthesis

125

Fig. 13 Setup for a membraneless flow-through cell. Catalysts and mediator are immobilized on the electrodes. Glucose oxidase and the mediator tetrathiafulvalene are immobilized for glucose oxidase regeneration on the anode. Glucose oxidase and horseradish peroxidase are immobilized on the cathode

anode, glucose oxidase was immobilized with tetrathiafulvalene as mediator for glucose oxidase regeneration. On the cathode, an enzyme cascade consisting of glucose oxidase and HRP was immobilized. This cascade enabled glucose oxidation by oxygen at the cathode, forming hydrogen peroxide as by-product. Hydrogen peroxide was consumed by HRP at the cathode side and, thus preventing inhibition by hydrogen peroxide. Although the measured conversion of 47% of 20 mM glucose was minor compared to other electroenzymatic processes with different reactor concepts, the STYs in relation to the geometric area of the electrode (18 mg h
1 cm
2) were by far the highest reported values. In addition to the omission of membranes, the process also has the benefit of running spontaneously without any external electrical energy input. Because oxygen is the natural electron acceptor for glucose oxidase, it was supplied from the gas phase through a Toray

126

L.M. Schmitz et al.

Fig. 14 Setup for a gas diffusion electrode (GDE). Oxygen is reduced to hydrogen peroxide at the gas–liquid–solid phase of the cathode (GDE). H2O2 diffuses into the liquid phase, where the enzymatic reaction takes place. Water is oxidized to form H+ ions and oxygen gas at the anode

paper double layer that was in direct contact with the cathodic graphite particles. In this way, the presence of solute oxygen on the cathode side was ensured and minimized on the anode side. Oxygen supply for electroenzymatic processes from the gaseous phase can also be realized using GDEs [95]. With this system, hydrogen peroxide production rates can be controlled by the applied cell voltage and current, in contrast to other electrolytic cell setups (Fig. 14). The controllability enables avoiding excess hydrogen peroxide and the consequent inactivation of biocatalysts in electroenzymatic processes. With this system, the electroenzymatic chlorination of monochlorodimedone to dichlorodimedone with CPO from C. fumago was demonstrated [96]. The maximal achieved reaction rate was 12.5 mM h
1 (STY of 52 g L
1 day
1) and the ttn was up to 1,150,000. Higher STYs could be probably achieved by increasing the electrode surface. In addition to the abovementioned optimized mass transfer between electrode surface and bulk solution, an increase in accessible electrode surface also enhances the electrochemical cofactor regeneration. Surface areas up to 19,685 m2 m
3 have been established for a RVC electrode [81]. Applying this continuous flow-through electrode reactor concept, FAD regeneration rates up to 95 mM h
1 (73 g L
1 h
1) were achieved. The reduction of FAD was coupled to the epoxidation of styrene to

Enzyme-Based Electrobiotechnological Synthesis

127

styrene oxide with StyA, yielding a reaction rate of 1.3 mM h
1. The reported electroenzymatic styrene epoxidation activity of 43.3 U g
1 was far below the potential of the enzyme with its natural cofactor regeneration system (2,100 U g
1). This might be the result of moderately effective binding of FADH2 to isolated StyA. The performance of electroenzymatic reaction systems benefits from optimized mass transfer between electrolyte and electrode and increased electrode surface-tovolume ratios. In general, the presented developments demonstrate the importance of reactor design considerations for further improvement of electroenzymatic processes.

7 Conclusion and Outlook The current state of the art of electroenzymatic processes demonstrates their potential for synthetic purposes. However, most of the processes are inferior in terms of productivity and final product concentration compared with enzymatic cofactor regeneration systems. Even though some electroenzymatic processes achieve initial reaction rates of the same order of magnitude as processes with enzymatic cofactor regeneration systems, they are still limited in stability and final product concentration (Table 11). Several approaches have been demonstrated for optimizing electroenzymatic processes. For the majority of electroenzymatic syntheses, mediators are used for indirect electron transfer between electrode and cofactor to improve electron transfer kinetics between electrode and enzyme. Various mediator types have been developed for cofactor regeneration and have been applied for electroenzymatic catalysis with oxidoreductases. The mediator selection, cofactor/mediator ratio, and reaction conditions (e.g., current density, potential, and media conductivity) are interdependent and are usually coordinated individually with the enzymatic synthesis reaction. The smart combination of enzymatic synthesis with mediated electrochemical cofactor regeneration offers a chance for improved reaction stability. The application of mediators nevertheless generates new challenges such as inactivation that need to be addressed. The inactivation mechanism has been extensively investigated for the rhodium mediator and proteins. A membrane reactor was developed for spatial separation of enzyme and mediator to prevent inactivation [52]. Productivity limitations can be minimized by using advanced cell designs such as 3D electrodes. The performance of such systems benefits from enhanced mass transfer rates. Increased cofactor regeneration rates and productivities of electroenzymatic syntheses were obtained with high surface-to-volume ratios for electrodes [81]. High productivities entail that the medium composition has to be adjusted in terms of conductivity, stabilization of enzymes, and substrate availability in terms of solubility by using, for example, ionic liquids. Solubility problems or product inhibition can also be overcome by using aqueous/organic two-phase systems.

128

L.M. Schmitz et al.

There are obviously various interesting targets for process optimization of electroenzymatic synthesis. However, there is a lack of systematic approaches in order to combine promising reaction engineering concepts with the demands of the enzymes on reaction conditions. Electroenzymatic syntheses are still not competitive with enzyme-based cofactor regeneration systems. Synthetic contribution at preparative scale might nevertheless be obtained by establishing processes for highvalue chemicals such as pharmaceuticals, flavors, and fragrances or highthroughput syntheses with relevant productivities.

References 1. Das P, Das M, Chinnadayyala SR, Singha IM, Goswami P (2016) Recent advances on developing 3rd generation enzyme electrode for biosensor applications. Biosens Bioelectron 79:386–397 2. Yu EH, Scott K (2010) Enzymatic biofuel cells-fabrication of enzyme electrodes. Energies 3 (1):23–42 3. Pankratov D, Conzuelo F, Pinyou P, Alsaoub S, Schuhmann W, Shleev S (2016) A nernstian biosupercapacitor. Angew Chem 55(49):15434–15438 4. Sode K, Yamazaki T, Lee I, Hanashi T, Tsugawa W (2016) BioCapacitor: a novel principle for biosensors. Biosens Bioelectron 76:20–28 5. Steckhan E (1994) Electroenzymatic synthesis. Top Curr Chem 170:83–111 6. Steckhan E, Arns T, Heineman WR, Hilt G, Hoormann D, Jorissen J, Kroner L, Lewall B, Putter H (2001) Environmental protection and economization of resources by electroorganic and electroenzymatic syntheses. Chemosphere 43(1):63–73 7. Kohlmann C, Lütz S (2015) Electroenzymatic synthesis. In: Hammerich O (ed) Organic electrochemistry5th edn. CRC Press, New York, pp 1511–1542 8. Paul CE, Hollmann F (2016) A survey of synthetic nicotinamide cofactors in enzymatic processes. Appl Microbiol Biotechnol 100(11):4773–4778 9. Wandrey C, Liese A, Kihumbu D (2000) Industrial biocatalysis: past, present, and future. Org Process Res Dev 4(4):286–290 10. Hummel W, Kula MR (1989) Dehydrogenases for the synthesis of chiral compounds. Eur J Biochem 184(1):1–13 11. Wichmann R, Wandrey C, Buckmann AF, Kula MR (1981) Continuous enzymatic transformation in an enzyme membrane reactor with simultaneous NAD(H) regeneration. Biotechnol Bioeng 23(12):2789–2802 12. van der Donk WA, Zhao HM (2003) Recent developments in pyridine nucleotide regeneration. Curr Opin Biotechnol 14(4):421–426 13. Hollmann F, Schmid A (2004) Electrochemical regeneration of oxidoreductases for cell-free biocatalytic redox reactions. Biocatal Biotransform 22(2):63–88 14. Uppada V, Bhaduri S, Noronha SB (2014) Cofactor regeneration – an important aspect of biocatalysis. Curr Sci 106(7):946–957 15. Tishkov VI, Popov VO (2004) Catalytic mechanism and application of formate dehydrogenase. Biochem Mosc 69(11):1252–1267 16. Chenault HK, Simon ES, Whitesides GM (1988) Cofactor regeneration for enzyme-catalysed synthesis. Biotechnol Genet Eng Rev 6(1):221–270 17. Nagesetti A, Rodzinski A, Stimphil E, Khanal TSC, Wang P, Guduru R, Liang P, Agoulnik I, Horstmyer J, Khizroev S (2017) Multiferroic coreshell magnetoelectric nanoparticles as NMR sensitive nanoprobes for cancer cell detection. Sci Rep 7(1):1610

Enzyme-Based Electrobiotechnological Synthesis

129

18. Ruinatscha R, H€ollrigl V, Otto K, Schmid A (2006) Productivity of selective electroenzymatic reduction and oxidation reactions: theoretical and practical considerations. Adv Synth Catal 348(15):2015–2026 19. Wandrey C (2004) Biochemical reaction engineering for redox reactions. Chem Rec 4 (4):254–265 20. Grieshaber D, MacKenzie R, Voros J, Reimhult E (2008) Electrochemical biosensors – sensor principles and architectures. Sensors 8(3):1400–1458 21. Steckhan E (1986) Indirect electroorganic syntheses – a modern chapter of organic electrochemistry. Angew Chem 25(8):683–701 22. Goldberg K, Schroer K, Lütz S, Liese A (2007) Biocatalytic ketone reduction – a powerful tool for the production of chiral alcohols, part I: processes with isolated enzymes. Appl Microbiol Biotechnol 76(2):237–248 23. Bonnefoy J, Moiroux J, Laval JM, Bourdillon C (1988) Electrochemical regeneration of NAD+ – a new evaluation of its actual yield. J Chem Soc Faraday Trans 84:941–950 24. Manjon A, Obon JM, Casanova P, Fernandez VM, Ilborra JL (2002) Increased activity of glucose dehydrogenase co-immobilized with a redox mediator in a bioreactor with electrochemical NAD+ regeneration. Biotechnol Lett 24(15):1227–1232 25. Obon JM, Casanova P, Manjon A, Fernandez VM, Iborra JL (1997) Stabilization of glucose dehydrogenase with polyethyleneimine in an electrochemical reactor with NAD(P)+ regeneration. Biotechnol Prog 13(5):557–561 26. Hilt G, Steckhan E (1993) Transition-metal complexes of 1,10-phenanthroline-5,6-dione as efficient mediators for the regeneration of NAD+ in enzymatic-synthesis. J Chem Soc Chem Commun 22:1706–1707 27. Hilt G, Lewall B, Montero G, Utley JHP, Steckhan E (1997) Efficient in situ redox catalytic NAD(P)+ regeneration in enzymatic synthesis using transition-metal complexes of 1,10phenanthroline-5,6-dione and its N-monomethylated derivative as catalysts. Liebigs Ann Recl 11:2289–2296 28. Komoschinski J, Steckhan E (1988) Efficient indirect electrochemical in situ regeneration of NAD+ and NADP+ for enzymatic oxidations using iron bipyridine and phenanthroline complexes as redox catalysts. Tetrahedron Lett 29(27):3299–3300 29. Itoh S, Fukushima H, Komatsu M, Ohshiro Y (1992) Heterocyclic ortho-quinones – mediator for electrochemical oxidation of NADH. Chem Lett 8:1583–1586 30. Kashiwagi Y, Osa T (1993) Electrocatalytic oxidation of NADH on thin poly(acrylic acid) film coated graphite felt electrode coimmobilizing ferrocene and diaphorase. Chem Lett 4:677–680 31. Schulz M, Leichmann H, Gunther H, Simon H (1995) Electromicrobial regeneration of pyridine-nucleotides and other preparative redox transformations with Clostridium thermoaceticum. Appl Microbiol Biotechnol 42(6):916–922 32. Schroeder I, Steckhan E, Liese A (2003) In situ NAD+ regeneration using 2,20 -azinobis (3-ethylbenzothiazoline-6-sulfonate) as an electron transfer mediator. J Electroanal Chem 541:109–115 33. Kochius S, Park JB, Ley C, Konst P, Hollmann F, Schrader J, Holtmann D (2014) Electrochemical regeneration of oxidised nicotinamide cofactors in a scalable reactor. J Mol Catal B Enzym 103:94–99 34. Degenring D, Schroder I, Wandrey C, Liese A, Greiner L (2004) Resolution of 1,2-diols by enzyme-catalyzed oxidation with anodic, mediated cofactor regeneration in the extractive membrane reactor: gaining insight by adaptive simulation. Org Process Res Dev 8 (2):213–218 35. Wulf H, Perzborn M, Sievers G, Scholz F, Bornscheuer UT (2012) Kinetic resolution of glyceraldehyde using an aldehyde dehydrogenase from Deinococcus geothermalis DSM 11300 combined with electrochemical cofactor recycling. J Mol Catal B Enzym 74 (1-2):144–150

130

L.M. Schmitz et al.

36. Baik SH, Kang C, Jeon IC, Yun SE (1999) Direct electrochemical regeneration of NADH from NAD+ using cholesterol-modified gold amalgam electrode. Biotechnol Tech 13(1):1–5 37. Siu E, Won K, Park CB (2007) Electrochemical regeneration of NADH using conductive vanadia-silica xerogels. Biotechnol Prog 23(1):293–296 38. Kashiwagi Y, Yanagisawa Y, Shibayama N, Nakahara K, Kurashima F, Anzai J, Osa T (1997) Preparative, electroenzymatic reduction of ketones on an all components-immobilized graphite felt electrode. Electrochim Acta 42(13–14):2267–2270 39. Kang YW, Kang C, Hong JS, Yun SE (2001) Optimization of the mediated electrocatalytic reduction of NAD+ by cyclic voltammetry and construction of electrochemically driven enzyme bioreactor. Biotechnol Lett 23(8):599–604 40. Kim MH, Yun SE (2004) Construction of an electro-enzymatic bioreactor for the production of (R)-mandelate from benzoylformate. Biotechnol Lett 26(1):21–26 41. Delecouls-Servat K, Basseguy R, Bergel A (2002) Designing membrane electrochemical reactors for oxidoreductase-catalysed synthesis. Bioelectrochemistry 55(1–2):93–95 42. DiCosimo R, Wong CH, Daniels L, Whitesides GM (1981) Enzyme-catalyzed organicsynthesis – electrochemical regeneration of NAD(P)H from NAD(P) using methyl viologen and flavoenzymes. J Org Chem 46(22):4622–4623 43. Cantet J, Bergel A, Comtat M (1996) Coupling of the electroenzymatic reduction of NAD+ with a synthesis reaction. Enzyme Microb Technol 18(1):72–79 44. Cantet J, Bergel A, Comtat M (1992) Bioelectrocatalysis of NAD+ reduction. Bioelectrochem Bioenergy 27(3):475–486 45. Yuan R, Watanabe S, Kuwabata S, Yoneyama H (1997) Asymmetric electroreduction of ketone and aldehyde derivatives to the corresponding alcohols using alcohol dehydrogenase as an electrocatalyst. J Org Chem 62(8):2494–2499 46. Hollmann F, Lin PC, Witholt B, Schmid A (2003) Stereospecific biocatalytic epoxidation: the first example of direct regeneration of a FAD-dependent monooxygenase for catalysis. J Am Chem Soc 125(27):8209–8217 47. Wienkamp R, Steckhan E (1982) Indirect electrochemical regeneration of NADH by a bipyridinerhodium(I) complex as electron-transfer agent. Angew Chem 21(10):782–783 48. Ruppert R, Herrmann S, Steckhan E (1987) Efficient indirect electrochemical in situ regeneration of NADH – electrochemically driven enzymatic reduction of pyruvate catalyzed by D-LDH. Tetrahedron Lett 28(52):6583–6586 49. Delecouls-Servat K, Basseguy R, Bergel A (2002) Membrane electrochemical reactor (MER): application to NADH regeneration for ADH-catalysed synthesis. Chem Eng Sci 57 (21):4633–4642 50. Hildebrand F, Lütz S (2007) Electroenzymatic synthesis of chiral alcohols in an aqueousorganic two-phase system. Tetrahedron-Asymmetry 18(10):1187–1193 51. Hoellrigl V, Otto K, Schmid A (2007) Electroenzymatic asymmetric reduction of rac-3methylcyclohexanone to (1S,3S)-3-methylcyclohexanol in organic/aqueous media catalyzed by a thermophilic alcohol dehydrogenase. Adv Synth Catal 349(8-9):1337–1340 52. Hildebrand F, Lütz S (2009) Stable electroenzymatic processes by catalyst separation. Chem Eur J 15(20):4998–5001 53. Toogood HS, Knaus T, Scrutton NS (2014) Alternative hydride sources for ene-reductases: current trends. ChemCatChem 6(4):951–954 54. Simon H, Gunther H, Bader J, Tischer W (1981) Electro-enzymatic and electro-microbial stereospecific reductions. Angew Chem 20(10):861–863 55. Thanos ICG, Simon H (1987) Electro-enzymatic viologen-mediated stereospecific reduction of 2-enoates with free and immobilized enoate reductase on cellulose filters or modified carbon electrodes. J Biotechnol 6(1):13–29 56. Fisher K, Mohr S, Mansell D, Goddard NJ, Fielden PR, Scrutton NS (2013) Electroenzymatic viologen-mediated substrate reduction using pentaerythritol tetranitrate reductase and a parallel, segmented fluid flow system. Catal Sci Technol 3(6):1505–1511

Enzyme-Based Electrobiotechnological Synthesis

131

57. Tosstorff A, Kroner C, Opperman DJ, Hollmann F, Holtmann D (2017) Towards electroenzymatic processes involving old yellow enzymes and mediated cofactor regeneration. Eng Life Sci 1:71–76 58. Trost E-M, Fischer L (2002) Minimization of by-product formation during amino acid oxidase catalyzed racemate resolution of amino acids. J Mol Catal B Enzym 19-20:189–195 59. Buto´ S, Pollegioni L, D’Angiuro L, Pilone MS (1994) Evaluation of D-amino-acid oxidase from Rhodotorula gracilis for the production of α-keto acids: a reactor system. Biotechnol Bioeng 44:1288–1294 60. Frede M, Steckhan E (1991) Continuous electrochemical activation of flavoenzymes using polyethyleneglycol-bound ferrocenes as mediators – a model for the application of oxidoreductases as oxidation catalysts in organic synthesis. Tetrahedron Lett 32(38):5063–5066 61. Cass AE, Davis G, Francis GD, Hill HA, Aston WJ, Higgins IJ, Plotkin EV, Scott LD, Turner AP (1984) Ferrocene-mediated enzyme electrode for amperometric determination of glucose. Anal Chem 56(4):667–671 62. Hill HAO, Oliver BN, Page DJ, Hopper DJ (1985) The enzyme-catalyzed electrochemical conversion of para-cresol into para-hydroxybenzaldehyde. J Chem Soc Chem Commun 21:1469–1471 63. Brielbeck B, Frede M, Steckhan E (1994) Continuous electroenzymatic synthesis employing the electrochemical enzyme membrane reactor. Biocatalysis 10(1-4):49–64 64. Petersen A, Steckhan E (1999) Continuous indirect electrochemical regeneration of galactose oxidase. Bioorg Med Chem 7(10):2203–2208 65. Kohlmann C, Greiner L, Leitner W, Wandrey C, Lütz S (2009) Ionic liquids as performance additives for electroenzymatic syntheses. Chem Eur J 15(43):11692–11700 66. Kawabata S, Iwata N, Yoneyama H (2000) Asymmetric electrosynthesis of amino acid using an electrode modified with amino acid oxidase and electron mediator. Chem Lett 2:110–111 67. Jeon JS, Shin IH, Sang BI, Park DH (2005) Electrochemical regeneration of FAD by catalytic electrode without electron mediator and biochemical reducing power. J Microbiol Biotechnol 15(2):281–286 68. Sono M, Roach MP, Coulter ED, Dawson JH (1996) Heme-containing oxygenases. Chem Rev 96(7):2841–2888 69. Cirino PC, Arnold FH (2002) Protein engineering of oxygenases for biocatalysis. Curr Opin Chem Biol 6(2):130–135 70. Faber K (2000) Biotransformations in organic chemistry, 4th edn. Springer, New York (5th edn available) 71. Creveling CR, Daly JW, Witkop B, Undenfriend S (1962) Substrates and inhibitors of dopamine-beta-oxidase. Biochim Biophys Acta 64:125–134 72. Girhard M, Bakkes PJ, Mahmoud O, Urlacher VB (2005) P450 biotechnology. In: de Montellano O (ed) Cytochrome P450: structure, mechanism, and biochemistry, 3rd edn. Kluwer Academic/Plenum, New York, pp 451–520 73. Kaufman S (1963) The structure of the phenylalanine-hydroxylation cofactor. PNAS 50:1085–1093 74. Suske WA, Held M, Schmid A, Fleischmann T, Wubbolts MG, Kohler HPE (1997) Purification and characterization of 2-hydroxybiphenyl 3-monooxygenase, a novel NADHdependent, FAD-containing aromatic hydroxylase from Pseudomonas azelaica HBP1. J Biol Chem 272(39):24257–24265 75. Faulkner KM, Shet MS, Fisher CW, Estabrook RW (1995) Electrocatalytically driven omega-hydroxylation of fatty-acids using cytochrome-P450 4A1. PNAS 92(17):7705–7709 76. Estabrook RW, Faulkner KM, Shet MS, Fisher CW (1996) Application of electrochemistry for P450-catalyzed reactions. Methods Enzymol 272:44–51 77. Udit AK, Arnold FH, Gray HB (2004) Cobaltocene-mediated catalytic monooxygenation using holo and heme domain cytochrome P450BM3. J Inorg Biochem 98(9):1547–1550 78. C¸ekic¸ SZ, Holtmann D, Güven G, Mangold K-M, Schwaneberg U, Schrader J (2010) Mediated electron transfer with P450cin. Electrochem Commun 12(11):1547–1550

132

L.M. Schmitz et al.

79. Hollmann F, Schmid A, Steckhan E (2001) The first synthetic application of a monooxygenase employing indirect electrochemical NADH regeneration. Angew Chem Int Ed 40(1):169–171 80. Hollmann F, Hofstetter K, Habicher T, Hauer B, Schmid A (2005) Direct electrochemical regeneration of monooxygenase subunits for biocatalytic asymmetric epoxidation. J Am Chem Soc 127(18):6540–6541 81. Ruinatscha R, Buehler K, Schmid A (2014) Development of a high performance electrochemical cofactor regeneration module and its application to the continuous reduction of FAD. J Mol Catal B Enzym 103:100–105 82. Tosstorff A, Dennig A, Ruff AJ, Schwaneberg U, Sieber V, Mangold K-M, Schrader J, Holtmann D (2014) Mediated electron transfer with monooxygenases—insight in interactions between reduced mediators and the co-substrate oxygen. J Mol Catal B Enzym 108:51–58 83. Str€ ohle FW, Cekic SZ, Magnusson AO, Schwaneberg U, Roccatano D, Schrader J, Holtmann D (2013) A computational protocol to predict suitable redox mediators for substitution of NAD(P)H in P450 monooxygenases. J Mol Catal B Enzym 88:47–51 84. Ley C, Schewe H, Str€ohle FW, Ruff AJ, Schwaneberg U, Schrader J, Holtmann D (2013) Coupling of electrochemical and optical measurements in a microtiter plate for the fast development of electro enzymatic processes with P450s. J Mol Catal B Enzym 92:71–78 85. Seelbach K, van Deurzen MPJ, van Rantwijk F, Sheldon RA, Kragl U (1997) Improvement of the total turnover number and space-time yield for chloroperoxidase catalyzed oxidation. Biotechnol Bioeng 55(2):283–288 86. van de Velde F, van Rantwijk F, Sheldon RA (1999) Selective oxidations with molecular oxygen, catalyzed by chloroperoxidase in the presence of a reductant. J Mol Catal B Enzym 6 (5):453–461 87. Pletcher D (1999) Indirect oxidations using electrogenerated hydrogen peroxide. Acta Chem Scand 53(10):745–750 88. Chen JK, Nobe K (1993) Oxidation of dimethylaniline by horseradish-peroxidase and electrogenerated peroxide. J Electrochem Soc 140(2):299–303 89. Lee K, Moon SH (2003) Electroenzymatic oxidation of veratryl alcohol by lignin peroxidase. J Biotechnol 102(3):261–268 90. Laane C, Weyland A, Franssen M (1986) Bioelectrosynthesis of halogenated compounds using chloroperoxidase. Enzyme Microb Technol 8(6):345–348 91. Dembitsky VM (2003) Oxidation, epoxidation and sulfoxidation reactions catalysed by haloperoxidases. Tetrahedron 59(26):4701–4720 92. Lütz S, Steckhan E, Liese A (2004) First asymmetric electroenzymatic oxidation catalyzed by a peroxidase. Electrochem Commun 6(6):583–587 93. Lütz S, Vuorilehto K, Liese A (2007) Process development for the electroenzymatic synthesis of (R)-methylphenylsulfoxide by use of a 3-dimensional electrode. Biotechnol Bioeng 98:525–534 94. Kohlmann C, Lütz S (2006) Electroenzymatic synthesis of chiral sulfoxides. Eng Life Sci 6 (2):170–174 95. Krieg T, Huttmann S, Mangold KM, Schrader J, Holtmann D (2011) Gas diffusion electrode as novel reaction system for an electro-enzymatic process with chloroperoxidase. Green Chem 13(10):2686–2689 96. Holtmann D, Krieg T, Getrey L, Schrader J (2014) Electroenzymatic process to overcome enzyme instabilities. Catal Commun 51:82–85 97. Horst AEW, Bormann S, Meyer J, Steinhagen M, Ludwig R, Drews A, AnsorgeSchumacher M, Holtmann D (2016) Electro-enzymatic hydroxylation of ethylbenzene by the evolved unspecific peroxygenase of Agrocybe aegerita. J Mol Catal B Enzym. https://doi. org/10.1016/j.molcatb.2016.12.008

Enzyme-Based Electrobiotechnological Synthesis

133

98. Slusarczyk H, Felber S, Kula MR, Pohl M (2000) Stabilization of NAD-dependent formate dehydrogenase from Candida boidinii by site-directed mutagenesis of cysteine residues. Eur J Biochem 267(5):1280–1289 99. Lutz J, Hollmann F, Ho TV, Schnyder A, Fish RH, Schmid A (2004) Bioorganometallic chemistry: biocatalytic oxidation reactions with biomimetic NAD+/NADH co-factors and [Cp*Rh(bpy)H]+ for selective organic synthesis. J Organomet Chem 689(25):4783–4790 100. Steckhan E, Herrmann S, Ruppert R, Dietz E, Frede M, Spika E (1991) Analytical study of a series of substituted (2,20 -bipyridyl)(pentamethylcyclopentadienyl)rhodium and iridium complexes with regard to their effectiveness as redox catalysts for the indirect electrochemical and chemical-reduction of NAD(P)+. Organometallics 10(5):1568–1577 101. H€ ollrigl V, Otto K, Schmid A (2007) Electroenzymatic asymmetric reduction of rac-3methylcyclohexanone to (1S,3S)-3-methylcyclohexanol in organic/aqueous media catalyzed by a thermophilic alcohol dehydrogenase. Adv Synth Catal 349(8–9):1337–1340 102. Speiser B (1981) Elektroanalytische methoden II: cyclische voltammetrie. Chemie Unserer Zeit 15(2):62–67 103. Hamann CH, Vielstich W (1998) Elektrochemie, 3rd edn. Wiley, Weinheim 104. Lütz S, Rao N, Wandrey C (2006) Membranes in Biotechnology. Chem Eng Technol 29:1404–1415 105. Dreisbach C, Wischnewski G, Kragl U, Wandrey C (1995) Changes of enantioselectivity with the substrate ratio for the addition of diethylzinc to aldehydes using a catalyst coupled to a soluble polymer. J Chem Soc 7:875–878 106. Greiner L, Laue S, Liese A, Wandrey C (2006) Continuous homogeneous asymmetric transfer hydrogenation of ketones: lessons from kinetics. Chem Eur J 12(6):1818–1823 107. Kragl U, Dreisbach C (1996) Kontinuierliche asymmetrische synthese in einem membranreaktor. Angew Chem 108(6):684–685 108. Kragl U, Dwars T (2002) The development of new methods for the recycling of chiral catalysts. Trends Biotechnol 20(1):45–45 109. Liese A (2003) Biological principles applied to technical asymmetric catalysis. Habilitation, Rheinische Friedrich-Wilhelms-Universita¨t, Bonn 110. Rissom S, Beliczey J, Giffels G, Kragl U, Wandrey C (1999) Asymmetric reduction of acetophenone in membrane reactors: comparison of oxazaborolidine and alcohol dehydrogenase catalysed processes. Tetrahedron Asymmetry 10(5):923–928 111. Woltinger J, Bommarius AS, Drauz K, Wandrey C (2001) The chemzyme membrane reactor in the fine chemicals industry. Org Process Res Dev 5(3):241–248 112. Woltinger J, Drauz K, Bommarius AS (2001) The membrane reactor in the fine chemicals industry. Appl Catal A 221(1-2):171–185 113. Vuorilehto K, Lütz S, Wandrey C (2004) Indirect electrochemical reduction of nicotinamide coenzymes. Bioelectrochemistry 65(1):1–7 114. van Deurzen MPJ, Seelbach K, van Rantwijk F, Kragl U, Sheldon RA (1997) Chloroperoxidase: use of a hydrogen peroxide-stat for controlling reactions and improving enzyme performance. Biocatal Biotransform 15(1):1–16 115. Hollmann F, Witholt B, Schmid A (2002) [Cp*Rh(bpy)(H2O)]2+: a versatile tool for efficient and non-enzymatic regeneration of nicotinamide and flavin coenzymes. J Mol Catal B Enzym 19:167–176 116. Steckhan E, Frede M, Herrmann S, Ruppert R, Spika E, Dietz E (1992) Enzymatische synthesen durch indirekte elektrochemische prozesse. DECHEMA-monographien, vol 125. VCH, Weinheim, pp 723–753 117. Greiner D, Kohlmann C, Leitner W, Lütz S, Wandrey C (2009) Procedure for the conversion of substrates e.g. organic sulfides, ketones and amino acids, to products by a combined electrochemical and catalytic process in a reaction medium containing ionic liquid. Patent DE102007044379 A1, 26 March 2009 118. Krieg T, Sydow A, Schroder U, Schrader J, Holtmann D (2014) Reactor concepts for bioelectrochemical syntheses and energy conversion. Trends Biotechnol 32(12):645–655 119. Min K, Park DH, Yoo YJ (2010) Electroenzymatic synthesis of L-DOPA. J Biotechnol 146 (1–2):40–44

134

L.M. Schmitz et al.

120. Varnicic M, Vidakovic-Koch T, Sundmacher K (2015) Gluconic acid synthesis in an electroenzymatic reactor. Electrochim Acta 176:1523–1523 121. Hildebrand F, Kohlmann C, Franz A, Lütz S (2008) Synthesis, characterization and application of new rhodium complexes for indirect electrochemical cofactor regeneration. Adv Synth Catal 350(6):909–918 122. Hofstetter K, Lutz J, Lang I, Witholt B, Schmid A (2004) Coupling of biocatalytic asymmetric epoxidation with NADH regeneration in organic-aqueous emulsions. Angew Chem 43 (16):2163–2166 123. Ruinatscha R, Dusny C, Buehler K, Schmid A (2009) Productive asymmetric styrene epoxidation based on a next generation electroenzymatic methodology. Adv Synth Catal 351(14–15):2505–2515 124. Schmid A, Vereyken I, Held M, Witholt B (2001) Preparative regio- and chemoselective functionalization of hydrocarbons catalyzed by cell free preparations of 2-hydroxybiphenyl 3-monooxygenase. J Mol Catal B Enzym 11(4-6):455–462

Adv Biochem Eng Biotechnol (2019) 167: 135–180 DOI: 10.1007/10_2017_16 © Springer International Publishing AG 2017 Published online: 2 September 2017

Engineering of Microbial Electrodes Sven Kerzenmacher

Abstract This chapter provides an overview of the current state-of-the-art in the engineering of microbial electrodes for application in microbial electrosynthesis. First, important functional aspects and requirements of basic materials for microbial electrodes are introduced, including the meaningful benchmarking of electrode performance, a comparison of electrode materials, and methods to improve microbe–electrode interaction. Suitable current collectors and composite materials that combine different functionalities are also discussed. Subsequently, the chapter focuses on the design of macroscopic electrode structures. Aspects such as mass transfer and electrode topology are touched upon, and a comparison of the performance of microbial electrodes relevant for practical application is provided. The chapter closes with an overall conclusion and outlook, highlighting the future prospects and challenges for the engineering of microbial electrodes toward practical application in the field of microbial electrosynthesis.

S. Kerzenmacher (*) IMTEK - Department of Microsystems Engineering, University of Freiburg, Freiburg im Breisgau, Germany e-mail: [email protected]

136

S. Kerzenmacher

Graphical Abstract

Keywords Anode, Bioelectrochemical systems, Cathode, Materials, Microbial electrosynthesis, Microbial fuel cell

Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Basic Materials for Microbial Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Reporting and Benchmarking of Electrode Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Functional Aspects and Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Basic Electrode Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Methods to Improve Microbe–Electrode Interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Current Collector Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Composite Materials Combining Functionalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 From Functional Materials to Macroscopic Electrode Structures . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Mass Transfer in Microbial Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 A Comparison of Typical Electrode Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Bioelectrochemical Performance of Selected Microbial Electrodes of Practical Relevance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

137 139 139 141 141 145 148 150 154 154 157 161 171 173

Abbreviations Ag/ AgCl AQDS CB CNTs CP DET

The silver/silver chloride reference electrode, approx. + 199 mV vs SHE Anthraquinone-2,6-disulfonic disodium salt, a redox mediator Carbon black Carbon nanotubes Carbon paper Direct electron transfer

Engineering of Microbial Electrodes

ECSA GNR ITO MEC MET MFC PANI PPy PTFE Ra RΩ SCE SHE

137

Electrochemical accessible surface area Graphene nanoribbons Indium tin oxide, a transparent electronically conductive material Microbial electrolysis cell Mediated electron transfer Microbial fuel cell Polyaniline Polypyrrole Polytetrafluoroethylene Roughness average, arithmetic average of the absolute values of height deviations from the mean line Ohmic resistance, for example, of an electrode material The saturated calomel reference electrode, approx. + 244 mV vs SHE The standard hydrogen reference electrode

1 Introduction Electrodes play a key role in any bioelectrochemical system – be it in microbial fuel cells for generating electricity or microbial electrosynthesis reactors for producing commodities from electricity and, for example, CO2. Typically, these systems rely on microbial electrodes, in which electrons are exchanged between an abiotic electrode material and the metabolism of microorganisms. In pioneering works with electroactive microorganisms (microorganisms that exchange electrons across their cell membrane without the addition of artificial electron shuttles or mediators; see also [1]), simple two-dimensional electrodes made of glassy carbon [2, 3] or unpolished graphite [4, 5] have been used. However, soon afterward, electrodes with a three-dimensional structure and thus a larger specific surface area have been used, such as porous carbon paper [6]. The prime question in the development of microbial electrodes is how to design an electrode structure that forms an artificial functional habitat for electroactive microorganisms and enables the transfer of electrons to or from their metabolism (see Fig. 1). In this context, the different electron transfer mechanisms have to be considered [7, 8] (see also [1]). In the pioneering work of Potter [9], a catalytically active platinum electrode was used, on which microbial fermentation products or intermediates were oxidized. Later, artificial or natural mediator molecules were used to shuttle electrons from a microbial metabolism to an electrode. Only since the 1990s have electroactive microorganisms been known that are able to exchange electrons directly with an electrode via membrane cytochromes (e.g., Geobacter, Shewanella) [2–4, 10, 11] or self-secreted mediators (e.g., Shewanella) [12]. Using these bacteria, it is no longer necessary to add artificial electron mediators (e.g., methylene blue) to the system, which can become a significant cost factor [13].

138

S. Kerzenmacher

Fig. 1 Hierarchical levels and important parameters of microbial electrodes

This chapter is intended to give an overview of the state-of-the-art in the engineering of microbial electrodes for application in microbial electrosynthesis, with a strong focus on electrodes relying on direct electron transfer. These types of microbial electrodes are considered of particular relevance for practical application, as they require neither costly nor toxic mediators (which are eventually lost in continuously operated systems), nor expensive catalytically active materials such as platinum. Not considered are electrodes for analytical applications such as spectroscopy [14, 15] and studies with electrochemical quartz crystal microbalance [16]. These electrodes have other requirements (for instance, flat surface topology and optical transparency) and are often not useful for technical applications where high electron transfer rates and thus high productivities are required. Section 2 of this chapter covers the basic materials used for the construction of microbial electrodes. First, important aspects of benchmarking electrode performance and functional aspects and requirements are highlighted. This is followed by a comparison of electrode materials and methods to improve microbe–electrode interaction. Finally, suitable current collectors and composite materials that combine different functionalities are discussed. Section 3 focuses on the design of macroscopic electrode structures. It touches on aspects such as mass transfer and electrode topology, and provides a comparison of the performance of microbial electrodes relevant for practical application. The chapter ends with an overall conclusion and outlook highlighting the future prospects and challenges for the engineering of microbial electrodes.

Engineering of Microbial Electrodes

139

2 Basic Materials for Microbial Electrodes The engineering challenges in the development of microbial electrodes are manifold (for a recent review of the field see [17]). For practical application, microbial electrodes must enable sufficient electron transfer rates to or from the electroactive microorganisms at low overpotential. Furthermore, they must exhibit operational stability and must be amenable to scale-up and fabrication at low cost (see also [18]). In this context, the application of an electrode material as anode or cathode and its exact operational regime must be differentiated. For instance, in a certain potential range a material may be subject to corrosion and thus be unsuitable, whereas under different operational parameters it could be perfectly stable. Furthermore, depending on the type of microorganism (charge, shape, etc.) and the underlying electron transfer mechanism [17], the interaction between microbe and electrode may differ substantially (see also [1]. Depending on the application, it can also be advantageous to find a compromise between good performance and low cost. Thus it may be that the best performing electrodes are not always the prime choice; sometimes a cheaper electrode material could be of commercial advantage in the overall context. In the following sections the importance of meaningful reporting and benchmarking of electrode performance is discussed first, followed by important functional aspects and requirements regarding the technical application of microbial electrodes, an overview of basic electrode materials, methods to improve their interaction with microbes, and current collector materials. Finally, the evolving field of composite electrode structures, in which several materials are combined to optimize electrode functionality, is highlighted.

2.1

Reporting and Benchmarking of Electrode Performance

It is often difficult to compare experimental results from different work (see, for instance, the literature results listed in Tables 2 and 3). In many cases this is simply because the experimental procedures are not fully disclosed and results are not clearly reported. For instance, different materials are frequently compared only on the basis of the overall performance (e.g., power or power density) of the electrochemical cell. This is certainly useful in comparing the performance of different devices, but with respect to developing and benchmarking individual cell components (electrodes, separators, etc.,) such a comparison is of limited use. Instead, the main point of interest in electrode development is certainly the achievable current density at a given electrode polarization. It is thus of utmost importance to use reference electrodes to measure the individual potentials of anode and cathode, and report the polarization of anode and cathode separately. Furthermore, because electrode sizes typically vary greatly between studies, suitable normalization is required to allow for comparability (see Fig. 2). Current densities are therefore

140

S. Kerzenmacher

Fig. 2 Meaningful electrode comparison mandates suitable normalization

often reported on the basis of a geometric (projected) electrode area, which is useful for membrane-like electrodes of limited thickness. However, in the case of voluminous three-dimensional electrodes, this way of normalization can be very misleading. Instead, it would be more useful to normalize the current values with respect to the volume of the (porous) electrode structure or the reactor volume surrounding the electrode, particularly if planktonic cells contribute to the electrode reaction. In this context, it is equally important to report the overall size and shape of the electrode under investigation. With electrode materials of comparably low conductivity, the actual size of the electrode and the corresponding conduction losses can have a non-negligible influence on polarization data (see also Sect. 2.5). Similarly, mass transport inside a voluminous porous structure can be different in the case of small cubes, where edge effects can lead to significantly improved mass transfer, as compared to larger sheets where mass transport perpendicular to the plane dominates. Ideally, all relevant dimensions and values describing the electrode and experimental section are reported, so that other researchers can perform further analyses and draw their own conclusions based on the experimental results. The same is true for the microbial procedures (type and source of inoculum, inoculation procedure, time course of the experiment, etc.), and operational parameters of the reactor (e.g., temperature, reactant concentrations, media compositions, ionic conductivities, buffer capacities, flow rates, hydraulic retention times). As great variation between nominally identical experiments is often observed, the experimental setup and all relevant procedures should be well-controlled. Likewise, a sufficient number of replicate and/or parallel experiments should be performed, depending on the variability of the system under investigation. Another important aspect in the benchmarking of electrodes is the question of whether the electrode is operated under transient or steady-state conditions. If an electrode polarization curve is recorded too quickly (e.g., too fast potential/current sweep rate, too little time between different load resistors), its performance can be greatly overestimated. For instance, concentration profiles within the electrode may not fully develop or non-faradaic currents originating from charge stored in the biofilm can dominate. As discussed in detail elsewhere [52], electrode double-layer

Engineering of Microbial Electrodes

141

capacity, uncompensated resistance related to the ionic conductivity of the electrolyte, and the junction potential of the reference electrode can significantly falsify experimental results, in particular if dynamic electrochemical characterization methods are used. To assess reliably the performance of any microbial electrode, great care should be taken in designing reactors, planning experimental procedures, and analyzing and interpreting the experimental results. Further information on this topic is available from a tutorial review on techniques for the study of microbial fuel cells, which is also highly relevant in the context of bioelectrosynthesis [53].

2.2

Functional Aspects and Requirements

The foremost requirement of any electrode material is electrical conductivity. In the field of microbial electrodes, commonly used carbon-based materials or metals usually provide sufficient intrinsic electric conductivity for lab-scale applications. However, when larger electrodes are used, considerable potential gradients can evolve across two-dimensional and three-dimensional materials (see also Sect. 2.5). These potential gradients directly translate into electrical losses and reduced efficiency. Furthermore, the potential drop across an electrode results in locally different electrode potentials and thus current densities. To counter this, current collectors are usually included in electrodes of technical relevance. Current collectors can, for instance, be wires or meshes of stainless steel or other highly conductive materials that are in electrical contact with the actual electrode material, as explained further in Sect. 2.5. A requirement that distinguishes microbial electrodes from other electrodes (e.g., in batteries or conventional fuel cells) is the fact that, at least in the case of direct electron transfer (DET), microbial electrodes also usually have to perform as a host structure for microorganisms forming a biofilm. This first translates into the requirement of biocompatibility (meaning non-toxicity toward the electrogenic microorganisms). Furthermore, the electrode structure has to enable sufficient supply and removal of reactants to and from the living microbial cells (see also Sect. 3.1 of this chapter). Another important aspect is the electrode’s resistance to degradation, be it to microorganism-induced biofouling, the deposition of inorganic substances (salts) known as scaling, the resistance to corrosion and chemical degradation, and mechanical stability. In applications where only a limited stability to degradation can be reached, the amenability of the electrode material toward cleaning and re-use gains particular importance.

2.3

Basic Electrode Materials

An overview of the characteristics of some basic electrode materials is given in Table 1. Because carbon materials generally fulfill the basic electrode requirements

++

+ +/


+/


+

+

++ ++

Activated carbon

Graphene

Carbon nanotubes (CNTs) Metals Stainless steel Copper ? ? ++ ++ ++ +

+ +





++ ++

+/
+/


+ +

+ +/


+/


+/


++ ++

++

++

+

Biofilm formation

+ +

+ +

– +

+ +

++

++

+

+

Electron transfer

+/
+/






++ ++

++ ++

?

?

+/


+/


Mechanical stability

Rating scale: ++ ¼ very good, + ¼ good, +/
¼ reasonable,
¼ poor,

¼ very poor,? ¼ questionable/unknown

Titanium Nickel Conductive polymers Polyaniline PPy Conductive ceramics Ti4O7 Indium doped tin oxide (ITO)

++

+

++

Corrosion/degradation resistance

Electrical conductivity

Material Carbon materials Graphite

Table 1 Characteristics of some basic electrode materials (see text for explanations)

Results not necessarily transferable to other materials

Susceptible to degradation Susceptible to degradation

Susceptible to corrosion Susceptible to corrosion, toxic corrosion products Insulating oxide layer formed

Hydrophobic surface may hinder biofilm attachment Hydrophilic surface favors biofilm formation Hydrophobic surface may hinder biofilm attachment Hydrophobic surface may hinder biofilm attachment

Remarks

142 S. Kerzenmacher

Engineering of Microbial Electrodes

143

of sufficient electrical conductivity on the small scale (for larger electrodes current collectors are required; see also Sect. 2.5), biocompatibility, and relatively low cost, they are very prominent in the field of microbial electrodes. Independent of their topology (which is discussed in Sect. 3.2), carbon materials come in a wide chemical variety, as briefly discussed in the following. For detailed overviews on carbon materials see the review articles available elsewhere [54–56]. With its comparably high electrical conductivity, graphite is particularly popular. It is commonly used in the form of felts [34, 57, 58], fibers [59], carbon papers [60], or sheets [61]. A further widely used material is activated carbon. In contrast to graphite, these highly disordered materials have a large specific surface area and exhibit surface oxides such as, for example, carboxylic, phenolic, and lactonic groups [62]. Activated carbon is applied in the form of cloths and fibrous materials [26, 34, 63] or as activated carbon granules [64]. Carbon nanomaterials such as graphene [48, 65–67] (from reduced graphene oxide), carbon nanotubes [34, 49, 54, 68, 69], and carbon nanofibers [54, 70, 71] have found successful application in microbial anodes. However, they are usually applied as a coating of a base material or in composite materials [67, 68], as described further in Sect. 2.6. Furthermore, there is some controversy in the literature regarding the biocompatibility and possible toxicity of graphene [65] and carbon nanotubes [54] toward bacterial cells, and it is suggested that their modification with conductive polymers can serve as protective layers [54, 65]. Besides those mentioned above, various carbon materials derived from the carbonization of different precursors are used. These comprise, for instance, wood-based biochars [72], polyaniline-modified natural loofah sponge [73], as well as reticulated carbon foam derived from sponge-like organic structures [74]. Synthetic polymers such a polyacrylonitrile [24] are also commonly carbonized. Depending on precursors and carbonization procedure, the carbonized materials can differ greatly regarding their degree of graphitization, hydrophilicity, surface functional groups, and surface morphology. Among metal-based electrode materials, stainless steel is particularly prominent [19, 75]. In a comparative study with an acetate-fed mixed consortium, stainless steel plate anodes delivered current densities in the region of 20 A m
2, which is twice the values obtained with plain graphite, but approximately only 60% of the current densities reached with three-dimensional carbon cloth (34 A m
2) [19]. In the case of stainless steel, it has been observed that the current density decreases at too high electrode potentials, which has been attributed to the formation of an n-type semiconducting oxide layer at high potentials [76]. It has been postulated that at, potentials below – 120 mV vs SCE (below the flat-band value of superaustenitic stainless steel), the formation of the detrimental n-type semi conductive oxide layer can be avoided [19]. In a different study with an acetate-fed secondary biofilm, flat anodes made from silver, copper, nickel, cobalt, titanium, and stainless steel were systematically compared at – 0.2 V vs Ag/AgCl [22]. Whereas current densities on cobalt and titanium were negligible, flat electrodes made from silver and copper delivered current densities of 11 A m
2 and 15 A m
2, respectively.

144

S. Kerzenmacher

For comparison, flat graphite delivered similar current densities in the region of 10 A m
2, whereas stainless steel and nickel plate anodes showed lower values of 7 A m
2 and 4 A m
2, respectively [22]. A concern with non-noble metal electrodes is the risk of corrosion. Whereas steel and nickel are passivating metals whose oxide layer can protect them from being oxidized at more positive potentials, copper is prone to oxidative dissolution. However, Baudler et al. demonstrated that acetate-oxidizing anodes are typically used in a potential range where no copper oxidation occurs [22]. Amongst the metals, copper thus appears to be a particularly attractive electrode material that enables comparably high current densities and exhibits high electrical conductivity, which is of relevance for larger scale electrodes. This is remarkable because so far copper has been commonly considered to exhibit antimicrobial properties, and in an earlier study it was evaluated as an unsuitable anode material [77]. The difference in findings can be explained by the operation of the copper anode in an unsuitable potential range. During startup of the anode at too positive potentials, corrosion occurs and presumably toxic copper ions are released which prevent biofilm formation [22]. Thus, with electrodes made from copper or similar materials special care has to be taken to operate them always in a suitable potential range where no corrosion occurs. This is generally of no concern in the typical potential range of cathodes used for bioelectrosynthesis. A third class of electrode materials consists of electrically conductive polymers that can increase the specific electrode surface area, improve bacterial adhesion, and facilitate electron transfer. In most cases, electrically conductive polymers are deposited onto a carbon-based support material or used in a composite material (see Sect. 2.6). For a more in-depth summary on the use of aromatic conducting polymers as catalyst supporting matrices, see [78]. In the field of microbial anodes, the aromatic conducting polymers such as polyaniline and polypyrrole (PPy) have mainly been employed. Both materials exhibit better bacterial adhesion than carbon materials, which explains their improved performance, for instance, with a mixed electroactive consortium [79] or a pure culture of Shewanella oneidensis [80]. Improving the interaction of bacterial cells further with conductive polyaniline and its hydroxylated derivative PAAP (poly(aniline-co-m-aminophenol)) as well as its hydroxylated/aminated derivative PADAP (poly(aniline-co-2,4diaminophenol)) has been studied using Shewanella loihica, yielding anodic current densities of up to 2.5 A m
2 [36]. A major drawback of electrically conductive polymers is that they are susceptible to microbial degradation, in particular in complex microbial communities found in wastewater. Although polyaniline is prone to microbial attack in sewage sludge, the fluorinated polytetrafluoroaniline was found to be fully resistant in both sewage sludge and a culture of Clostridia, and able to perform reductive biotransformations [81]. Finally, some electrically conductive ceramics have also been successfully employed as material for microbial electrodes. For instance, electrically conducting Ti4O7 (Magne´li phase) structures were produced from TiO2 scaffolds by reaction with elemental Zr under vacuum at 1,000 C [28]. Using Geobacter as anode bacterium, the highly porous three-dimensional structures yielded volumetric

Engineering of Microbial Electrodes

145

current densities of up to 9.5 mA cm
3 at 0.2 V vs Ag/AgCl. As the porous TiO2 scaffolds were approx. 2 cm  1 cm  1 cm in size, culture medium had to be actively pumped through the porous structure to counter mass transfer limitations. For comparison, without energy-intensive pumping a similar volumetric current density of 8.2 mA cm
3 (at
0.4 V vs SCE) was reached with a 500-μm thin activated carbon cloth [26]. Ti2AlC materials have also been prepared as porous three-dimensional foams and suggested as microbial anode, but not tested so far [82]. For spectroelectrochemical studies on Geobacter sulfurreducens, flat indium tin oxide (ITO) has been successfully employed. Here limiting current densities of approx. 2 A m
2 have been reached [83]. Similarly, S. oneidensis was cultivated on ITO electrodes. At 0.04 A m
2 the current densities of 0.2 V vs Ag/AgCl were very low [84]. These experiments also revealed that, at the ITO electrode, direct electron transfer prevailed, although at the graphite electrode, electron transfer via secreted redox mediators, such as flavins and quinones, also occurred [84]. From an engineering perspective this means that results achieved with ITO electrodes are not necessarily comparable to other materials of relevance for application.

2.4

Methods to Improve Microbe–Electrode Interaction

To improve the interaction of carbon and steel electrode materials with electroactive microorganisms, a number of surface modification methods have been successfully employed. The rationale behind these methods is to facilitate biofilm formation and/or improve electron transfer kinetics [85]. This can be achieved by surface functionalization with positively charged or hydrophilic groups, or by the introduction of a suitable micro- or nanostructure that improves biofilm formation and/or electron transfer. Furthermore, electron transfer can be improved by an increase of the electrochemical accessible surface area of the electrode (ECSA; see also Fig. 1), and the introduction of redox active species that function as surface-bound electron mediators. A review covering this field is available elsewhere [17]. In the following section, examples of these approaches are discussed in more detail. However, often the underlying mechanism of improved microbe– electrode interaction cannot clearly be discerned in light of the experimental findings. The rationale behind surface functionalization with positive surface charges is the typically negative charge of bacteria, which can lead to improved attachment of the cells [86]. In one of the pioneering studies, carbon cloth anodes were modified with positive surface charges by treatment with 5% ammonia gas in nitrogen at 700 C. This reduced the startup time of an acetate-fed microbial anode from approx. 150 h to 60 h and increased its power density by approx. 20% [86]. However, the method was considered too costly for larger-scale applications [87]. Alternatively, an electrochemical technique, based on the reduction of in situ generated aryl diazonium salts of para-aniline derivatives, has been employed to introduce

146

S. Kerzenmacher

µm

20 10

20 10

10

10 20

20

different functional groups and study the effect of hydrophilicity and electrode surface charge on biofilm formation [88]. In a comparative study with acetate-fed anode biofilms on flat glassy carbon electrodes, surface modification with positive charges resulted in a faster startup time and the highest current density of 2 A m
2. Whereas the kinetics of the electron-transfer reaction appeared not to be affected by the different surface modifications, the results suggest that the hydrophilicity appears to be more important for good bacterial adhesion than surface charge. Consequently, current output correlated with biomass quantity on the electrode and negatively charged but hydrophilic surface groups (SO3
) yielded higher current density than a CH3-functionalized hydrophobic surface [88]. The latter showed a distinctively different form of biofilm, as can be seen from Fig. 3 Similarly, the results of Santoro et al. indicate that residual hydrophobicity (along with a lower number of pores in the 5–10 μm range) slowed down biofilm formation [89]. Furthermore, they showed that the formation of a biofilm can increase the hydrophilicity over time. Remarkably, the difference between more hydrophobic and more hydrophilic anodes vanished after approx. 2 weeks of operation [89].

-N+(CH3)3 Hydrophilic

1020

-OH Hydrophilic

20 10

10 20 20 10

µm

µm – -SO3 Hydrophilic

µm -CH3 Hydrophobic

Fig. 3 Three-dimensional images (taken with a confocal laser scanning microscope) of an acetate-fed biofilm on a glassy carbon electrode, its surface modified with different functional groups, as indicated. Geobacter sp. have been stained to appear blue in the images. Reprinted with permission from Guo et al. (2013) Environ Sci Technol 7563–7570 [88]. Copyright (2013) American Chemical Society

Engineering of Microbial Electrodes

147

A simple and fast method to hydrophilize carbon surfaces and to introduce carboxyl and ammonium functional groups without changing the surface morphology is to use a nitrogen-based atmospheric plasma jet [90]. In conjunction with Shewanella sp. as anodic microorganisms, the plasma treatment of the carbon cloth anode resulted in a microbial fuel cell with doubled power density. Another method to introduce hydrophilic surface functionalities is the electrochemical oxidation of the carbon surface. With these methods, the introduction of surface functional groups usually goes together with a modification of the surface morphology and an increase in the electrochemical accessible surface area of the electrode (ECSA). One example is the modification of hydrophilic carbon cloth anodes by electrochemical oxidation in either a mixture of HNO3 and H2SO4 in ammonium sulfate or in ammonium nitrate. These treatments increased the ECSA by two orders of magnitude and introduced functional groups (increased amounts of nitrogen, sulfur, oxygen, and unsaturated and oxidized carbon) which enhanced the electrode’s wettability. This in turn enhanced biomass attachment and promoted biofilm formation, and generally led to increased anode current densities with a mixed microbial consortium. However, the performance advantage produced by surface modification over the plain carbon cloth decreased over time [91], similar to the findings of Santoro et al. [89]. The authors hypothesized that, after the biomass attached on the surface, other factors such as bacteria activity and speciation play a major role [91]. In a different study with dairy waste as substrate and a compost leachate inoculum, graphite felt electrodes were electrochemically oxidized in phosphate buffer, leading to microcavities in the sub-micrometer range on the fiber surface. This accelerated the formation of electroactive biofilms and increased current densities by a factor of three [57]. The observed positive effect of micro-structures on the electrode surface is in agreement with results from other studies. For instance, with a pure culture of S. oneidensis it was observed that a carbon electrode from nanofibers in the region of 200 nm yielded an approx. ten times higher current density and biofilm formation compared to a carbon electrode of similar specific surface area but with 10-μm fibers. The electrode with 10-μm fibers did not exhibit any biofilm and the authors concluded that the smaller features of the nanofiber mat are beneficial for biofilm formation and thus current production with S. oneidensis [35]. On stainless steel the effect of surface roughness on electrode performance was also investigated more closely, both with pure and mixed cultures. With a pure culture of G. sulfurreducens catalyzing the cathodic reduction of fumarate to succinate, an increase in average surface roughness (Ra) from 0.1 or 2.3 to 4.0 μm led to an approx. 60% increased current density in the region of 6.5 A m
2. A value of 4.0 μm appeared to be optimal, as higher values of 6.5 μm did not lead to higher current. The authors concluded that the increased roughness did not simply increase the available surface area, but promoted biofilm formation [51]. This result is in contrast to a study by the same group, but with an anodic biofilm derived from a complex inoculum fed with acetate. Under these conditions, no significant differences with current densities in the range of 17–25 A m
2 were obtained between smooth, micro-structured (Ra ¼ 5 μm), and macro-structured (300 μm wide and 500 μm deep grooves) stainless steel electrodes [19]. Based on these experiments,

148

S. Kerzenmacher

one may hypothesize that, with mixed microbial communities, the surface roughness of a steel electrode plays only a minor role, but clearly further research is necessary to elucidate the underlying mechanisms. Several further nanostructured surface modifications have been successfully employed with stainless steel. For instance, in a study with an acetate-fed biofilm rich in Geobacter sp., a stainless steel felt anode with and without prior flame oxidation by means of a Bunsen burner has been compared. The flame-oxidized anodes exhibited iron oxide nanoparticles on their surface, which in turn improved the biocompatibility and yielded a fivefold increased current density compared to the non-oxidized controls [29]. In a different study, carbon nanostructures were synthesized in situ on the surface of stainless steel anodes, using an ethylene flame. This treatment led to a 60-fold increase in power density of glucose-fed microbial fuel cells inoculated with a mixed consortium. The carbon nanostructures led to a significant increase in biomass attachment, although the exact mechanism behind the performance increase is unknown [92]. As well as surface modification to improve biofilm attachment, the introduction of redox active species that act as electron shuttles can improve electron transfer rates between the bacterial cells and the electrode. For instance, anthraquinone-2sulfonic acid was covalently grafted as an electron transfer mediator onto graphite felt anode via spontaneous reduction [93]. In an acetate-fed microbial fuel cell operating with a mixed consortium, this anode modification led to a performance increase of approx. 90%, as compared to the unmodified graphite felt. In both cases, the bacterial attachment to the electrode was similar; the improved performance was attributed to the facilitated electron transfer from bacteria to the electrode. Stability tests of the anode over a period of 2 months under operating conditions in a fuel cell indicated no degradation of the surface modification. For electrosynthesis application, the use of a viologen-based redox polymer to transfer electrons cathodically to living Escherichia coli cells has been demonstrated [94]. In summary, in the case of carbon electrode surfaces the introduction of positive charges as well as the hydrophilization has been shown to improve biofilm attachment and/or electron transfer. With steel electrodes, surface modification with iron oxide nanoparticles or carbon nanostructures can also result in improved biofilm formation and anode current densities. Nevertheless, care should be taken in the interpretation of surface-modification results, because several studies have demonstrated that the performance of modified and non-modified electrodes approaches similar levels after extended periods of operation. Furthermore, as has been demonstrated with stainless steel, the effect of surface modification methods can differ substantially between pure cultures and mixed consortia.

2.5

Current Collector Materials

Regarding current collector materials, metals such as stainless steel [95] and nickel [96] meshes as well as titanium [59] and nickel-chromium wires [97] are commonly used. In some cases, graphite foils [98] or rods [99] are also employed.

Engineering of Microbial Electrodes

149

Furthermore, carbon cloth anodes can serve not only as electrode material but also as a current collector because of their intrinsic conductivity. However, because of the relatively low specific conductivity of carbon compared to metals, scaling up usually requires metal current collectors. Copper has not yet been proposed as a current collector, which is probably related to concerns regarding biocompatibility and corrosion stability. However, in the light of its recent successful application as anode material [22] and as part of a composite material [23], it may be considered in the future. The importance of current collectors for the scaling-up of electrodes was analyzed by theoretical calculations in a study by Xie et al. [67]. The basis of their calculations is the three-dimensional anode made from a graphene-coated sponge, as shown in Fig. 4b. For the graphene-coated sponge they assumed a relatively low bulk conductance of 1 S m
1, whereas 106 S m
1 was taken as the conductance of the central current collector made from stainless steel. Furthermore, they assumed negligible contact resistances and a uniform volumetric electrode current density of 103 A m
3. The voltage drop arising from the resistance of the electrode is governed by the thickness of the graphene sponge, the thickness of the stainless steel current collector, and the conduction length (i.e., the planar dimension of the electrode, as illustrated in Fig. 4b). In the example with a 4 mm thick graphene sponge and without the current collector already at conduction lengths exceeding 1 cm, the voltage drop increases dramatically. This is also reflected in the experimental results, where the electrode without current collector gave an order of magnitude lower current densities [67]. In contrast, by including a 1 mm thick stainless steel current collector, the voltage drop is approx. only 6 mV for conduction lengths in the region of 1 m [67]. Obviously, these numbers are greatly dependent on the operational current density and the conductivity of the electrode materials. Nevertheless, these calculations and experiments underline the importance of current collectors, particularly if larger scale electrodes for practical applications are to be developed. The calculation of conduction losses is thus a valuable tool for designing such electrodes with regard to minimized conduction losses. This

Fig. 4 (a) Schematic of the plain polyurethane sponge (left) with a three-dimensional open pore structure, and the sponge after conformal coating with graphene (right). (b) Schematic of the composite electrode made from the graphene-coated sponge with (right) and without (left) the central stainless steel mesh as current collector. Adapted from [67] with permission from The Royal Society of Chemistry

150

S. Kerzenmacher

is also highlighted in a simulation study that more closely investigates the optimal position for the lead-out terminal of up-scaled electrodes [100].

2.6

Composite Materials Combining Functionalities

Composite materials seek to combine the characteristics of at least two different materials into a functional structure. The rationale behind this approach is the combination of favorable material properties, such as electronic conductivity and improved microbe–electrode interaction, into a composite material. In the following section, some examples of different composite electrode materials that have been explored in the field of microbial electrodes are presented. A more in-depth review that also covers the field of composite materials for air cathodes and separator membranes is available elsewhere [101]. One approach is the combination of a three-dimensional macrostructure with good mass transport properties (see Sect. 3.1) and an electronic conductor. For instance, carbon nanotubes (CNTs) were coated onto a polyurethane sponge as 3D– scaffold to form an approx. 200 nm thick layer of high electrical conductivity [69]. Compared to the previous CNT-coated textile anodes, this resulted in an approx. 50% improved performance with a mixed electroactive consortium. Using glucose as substrate, approx. 10 mA cm
3 at 0 V vs Ag/AgCl was observed, whereas operation with wastewater resulted in lower current density of approx. 2.5 mA cm
3. To decrease electrode cost, the same group later used graphene instead of CNTs to coat the polyurethane sponge. Because of the lower conductivity of graphene, some of the electrodes were equipped with an additional stainless steel current collector [67]. In this way, slightly higher current densities of approx. 25 mA cm
3 at 0 V vs Ag/AgCl were obtained using a glucose-fed natural consortium. Without the current collector, the values were approx. an order of magnitude lower, reaching only 2.5 mA cm
3. Given that the electrode was only approx. 1 cm  1 cm  0.4 cm in size, this also underlines the large impact of current collectors on the relatively small scale of laboratory experiments. Similarly, melamine foam was coated with electrically conductive copper [23] and used as anode together with an acetate-fed microbial consortium. Volumetric current densities in the region of 16 mA cm
3 have been reached for a cubic electrode of size approx. 1 cm3. In this case, no current collector was necessary because of the good intrinsic conductivity of the copper layer. Besides improving electrical conductivity, composites can be designed to improve the microbe–electrode interaction. The exact mechanism of the improvement is often unclear, and sometimes, for instance, synergistic effects between improved cell attachment and improved electron transfer exist. One approach to improve microbial adhesion and facilitate electron transfer is the coating of a stainless steel mesh (300 μm thick) with carbon black (CB) particles. With an acetate-fed electroactive biofilm this lead to a 50 times increased current density of about 15 A m
2 or 50 mA cm
3 [21]. In a similar study

Engineering of Microbial Electrodes

151

with an acetate-fed microbial consortium, the decoration of a carbon cloth anode with different types of carbon nanotubes was investigated. Compared to the 0.86 A m
2 achieved with plain carbon cloth, its decoration with regular CNTs and bamboo-like nitrogen-doped CNTs resulted in current densities of 2.3 A m
2 and 3.6 A m
2, respectively [102]. The improved performance of the nitrogendoped CNTs was attributed to their hydrophilic defects and presumably an improved biocompatibility as well as electron transfer activity, though the authors did not study these aspects further. In a different study with S. oneidensis, approx. 40% larger current densities of approx. 0.6 A m
2 were observed when the CNTs used to modify a carbon cloth anode were nitrogen-doped (evaluated at a cell voltage of 0.5 V). This was attributed to both an increased adsorption of the flavins secreted by S. oneidensis as redox-mediator on the nitrogen-doped CNTs, and an enhanced direct electron transfer via the outer membrane c-type cytochromes [103]. Similarly, graphene (from graphene oxide) was coated onto porous nickel foam to promote cell growth. With pure cultures of Shewanella putrafaciens (oneidensis) this resulted in an approx. 13-fold higher power density compared to a conventional carbon cloth anode [104]. In a different study, a composite material formed in situ was obtained by electrochemical oxidation of a graphite plate in ammonium sulfate solution, which yielded a nitrogen-doped graphene layer formed by exfoliation. Using this approach, anodic current densities of an acetate-fed microbial consortium could be significantly improved, with up to approx. 100 mV lower polarization losses at 2 A m
2 [105]. For application in cathodic electrosynthesis, carbon cloth anodes were modified with reduced graphene oxide that was itself functionalized with tetraethylene pentamine (TEPA) to introduce positive charges for improved bacterial adhesion and self-assembly [48]. In this way, the acetate production rate could be improved 3.6-fold and current densities of up to 0.7 A m
2 were reached with the wild-type of Sporomusa ovata as production organism. With a methanol-adapted strain of S. ovata, even a sixfold higher current density of 2.4 A m
2 was enabled by this modification. Different forms of metal oxides and ions can serve as mediator to facilitate charge transfer between microorganisms and an electrode. For instance, the incorporation of charge transfer mediators such as Mn2+, Ni2+, Fe3O4, and α-FeOOH in anodes made from graphite or activated carbon resulted in approx. doubled charge transfer kinetics [106, 107]. A corresponding composite anode was fabricated from polytetrafluoroethylene (PTFE), activated carbon, and Fe3O4 rolled onto a stainless steel current collector. The addition of Fe3O4 resulted in a 22% increased maximum power density with an acetate-fed microbial consortium. However, judging by Figure S2 in the supplementary material of the paper the anodes exhibit similar performance after 2 months of operation [108]. Comparable results with approx. 36% improved performance were obtained by adding nanosemiconductor goethite (α-FeOOH) to an activated carbon anode [107]. Similarly, a stainless steel wire mesh anode, operated with an acetate-fed biofilm, was modified with goethite and hematite [109], yielding a fivefold and threefold increase in power density, respectively. Furthermore, it has been demonstrated that the presence of iron oxide

152

S. Kerzenmacher

minerals can affect both the electron transfer pathways and biofilm formation in Geobacter spp. [110] and in Shewanella spp. [111]. In several studies the addition of TiO2 has also been shown to improve significantly the performance of microbial anodes, though the underlying mechanisms are largely unclear. To overcome its low electrical conductivity, TiO2 is usually used together with electrically conductive additives [112]. For instance, a composite anode was fabricated by coating a natural sponge with TiO2 and egg white protein. This structure was subsequently carbonized to yield carbon-coated core-shell nanoparticles of TiO2 decorated on a 3D carbon sponge. Evaluated at
0.4 V vs SCE with an acetate-fed microbial consortium, this material exhibited an approximately two times higher current density compared to the carbonized sponge base material. The maximum volumetric current density of the material amounted to approx. 3 mA cm
3. The performance improvement was attributed to a synergistic effect between the egg white-derived carbon and the TiO2 nanoparticles [113]. Similarly, a TiO2/graphene composite material was synthesized from graphene oxide and tetrabutyl titanate, and pasted on the surface of a nickel foam electrode. Using a pure culture of S. oneidensis, this material exhibited an approx. eight times higher performance than a regular carbon paper anode [112]. Furthermore, CNTs were decorated with TiO2 nanoparticles and used to coat a carbon cloth anode, operated with an acetate-fed microbial consortium. Whereas coating of the carbon cloth with either TiO2 or CNTs alone yielded similar performance, the CNT/TiO2 nanohybrid material exhibited an approx. four times higher current density of 4 A m
2 at –0.4 V vs Ag/AgCl [114]. Similar results were obtained with a multi-walled carbon nanotube/SnO2 nanocomposite deposited onto glassy carbon. With a power density of 1.4 W m
2 at 2.9 A m
2, the addition of SnO2 effectively doubled the performance of a mediator-less microbial fuel cell operated with a pure culture of glucose-fed E. coli at the anode [38]. Composite electrodes made from carbon materials and electrically conductive polymers are particularly popular, despite concerns with respect to their long-term stability in a complex microbial community (see Sect. 2.3). For instance, a carbon cloth electrode was modified using a mesoporous tartaric acid-doped polyaniline nanowire network, yielding four times higher power density with a pure culture of S. oneidensis [115]. Similarly, polyaniline (PANI) was electropolymerized on the surface of macroporous graphite felt, which was subsequently modified with CNTs by electrophoretic deposition. PANI deposition resulted in a hydrophilic surface of the graphite fibers, and the addition of CNTs further increased the specific surface area and electrical conductivity of the electrode. With S. oneidensis, the triplecomposite material with CNTs yielded 343% and 186% enhanced powered densities compared to the plain and only PANI-modified graphite felt. The CNT/PANI materials also showed increased bacterial adhesion [116]. For use with an acetatefed microbial consortium, a carbon cloth anode was first modified with electrochemically reduced graphene oxide, followed by a coating of PANI nanofibers [117], which resulted in a three times larger power density compared to the carbon cloth control electrode.

Engineering of Microbial Electrodes

153

In several studies, synergistic effects have been demonstrated. For instance, with a pure culture of S. oneidensis, investigations with graphene nanoribbons (GNR) and PANI on carbon paper (CP) showed that the current density and power density of the CP/GNRs/PANI anode were much higher than those of each individual component, indicating the synergistic effect between PANI and GNRs [118]. In a thionin-mediated microbial fuel cell with a pure culture of P. vulgaris, individual experiments with coatings of carbon black (CB), PPy, and a CB/PPy composite on a carbon paper anode were performed. Whereas the power density was already increased by coating the carbon paper with either CB or PPy, the best performance was reached with the PPy/CB composite, demonstrating the synergistic effect [119]. For the application with E. coli, a polyaniline (PANI)/mesoporous tungsten trioxide composite electrode was tested in an indirect microbial electrode, where the tungsten trioxide serves as electrocatalyst to oxidize E. coli fermentation products. [39]. Here the composite catalyst yielded approx. 30% and 100% higher power density than the tungsten trioxide or PANI alone, respectively. In a similar way, composite anodes based on nickel foams coated with a titanium carbide catalyst and polyaniline or chitosan were successfully used together with a mixed consortium of Acetobacter aceti and Gluconobacter roseus [120]. Redox mediators have been combined with polymers and/or carbon to form composite electrodes. For instance, a mixture of an osmium redox polymer together with CB and the proton-exchange polymer Nafion, deposited on a carbon cloth anode, increased the power density of an acetate-fed microbial fuel cell by approx. 60% compared to a non-modified control. In cyclic voltammetry experiments, a three times larger limiting current was observed with the osmium polymer modification. The authors attributed this performance increase to the osmium polymer functioning as a molecular wire, which connects the biofilm to the carbon cloth electrode [121]. Similarly, the redox mediator anthraquinone-2,6-disulfonic disodium salt (AQDS) was electropolymerized together with PPy on a carbon felt anode. Using a pure culture of Shewanella decolorationis, increased cell adhesion and a 13 times increased power density was observed, the current at
0.5 V vs. SCE amounting to approx. 1.3 A m
2 [37]. This value is similar to the current densities obtained in the systematic study by Kipf et al. [34], though at a more favorable approx. 500 mV lower potential. In a mediator-less MFC with E. coli an increase of cell performance with a CNT/PPy composite electrode was also observed. This was attributed to the formation of a biocompatible redox-active layer that takes over the function of a soluble mediator [122]. In summary, composite materials are a suitable and promising strategy to optimize the various functionalities of an electrode structure. The great advantage of this approach is the possibility to choose optimal materials for, for example, long-range (>>1 cm) electron conduction, biofilm formation, and efficient electron transfer and to combine them into one functional unit. A potential disadvantage is the more complex electrode fabrication, which in some cases is likely to results in increased costs, in particular for larger-scale electrode structures. To facilitate electron transfer and biofilm formation, the use of electrically conductive polymers or redox-active materials such as metal oxides and ions is very popular and has

154

S. Kerzenmacher

shown promising results. However, it remains to be demonstrated that these materials exhibit sufficient long-term stability and functionality under conditions relevant for practical application. Of particular concern is the limited stability of polymers in diverse microbial communities, and the possible leaching and loss of redox-active species such as metal ions in continuously operated systems.

3 From Functional Materials to Macroscopic Electrode Structures As electrochemical processes are essentially surface processes, the engineering of functional electrodes seeks to maximize the usable surface area. A straightforward way to increase a bulk materials specific surface area (normalized, e.g., to its weight) is its segmentation into smaller structures such as particles or fibers, or by introducing grooves or pores. An essential requirement of a functional electrode structure is the interconnection/continuity of the base material in terms of electron conduction. Isolated particles or fibers that are not electrically connected cannot contribute to current production. Furthermore, the pore space must also be interconnected. Isolated or too small pores can neither be colonized by the microorganisms, nor can substrates or reaction products be transported to them. The main parameters describing a three-dimensional porous electrode structure are porosity, pore size, specific surface area, and overall dimensions. Furthermore, of importance are the material’s electrical conductivity, feature (ligament) size, pore space, and tortuosity. There is obviously a delicate interplay between these parameters. For instance, when seeking to maximize the specific surface area of a material, as small as possible a pore size together with a high porosity is desirable. On the other hand, too small pores impede bacterial colonization and mass transport to the interior of the electrode. Furthermore, feature size and surface structure can influence biofilm formation. Similarly, too small ligaments of the electrically conductive base material result in increased conduction losses, depending on current density. In the following section the importance of mass transfer and electron conductivity in three-dimensional electrodes is first discussed, followed by a presentation and comparison of reported electrode topologies. Finally, this section gives an overview of the current densities achieved with different microbial electrodes of relevance for practical application and also discuss these in the light of electrode topology.

3.1

Mass Transfer in Microbial Electrodes

A major difference between abiotic and microbial electrodes is the greatly increased complexity of the travelling species in the latter case. The electrochemical reaction of hydrogen oxidation at the anode of a PEM fuel cell involves, apart

Engineering of Microbial Electrodes

155

from the transport of the substrate (in this case molecular hydrogen), only electrons and protons. In the case of a microbial anode, cations, buffering species, and reaction products also have to be transported, as illustrated in Fig. 5. The microorganisms that act as bioelectrocatalyst (such as, e.g., Geobacter sp.) frequently grow in the form of biofilms on the electrode surface, its thickness typically limited by the transport of species in and out of the biofilm. Mass transport within a biofilm is usually based on diffusion, and even with convection in the bulk medium, a purely diffusive boundary layer of 10–100 microns can be present on top. As pointed out in the detailed analysis of Popat and Torres, there is a need to match the transport rates of all species involved in the overall reaction at both anode and cathode, as any of these can be rate limiting [123]. As illustrated in Fig. 5, the transport of ions generated by the electrochemical electrode reaction (e.g., protons at the anode) is one important aspect that can limit overall electrode performance. This transport is related to the migration of charged species driven by a potential gradient between the anode and cathode. Depending on the electrolyte’s ionic conductivity, considerable voltage losses can occur, which are commonly referred to as ohmic losses or ohmic overpotential. It depends on the electrode current, the ionic conductivity of the electrolyte, and the distance between anode and cathode [124]. At this point it should be mentioned that the electron transport resistance of the electrode material itself as well as any electrical resistances in the electric circuit contribute to the ohmic losses of the electrochemical

Fig. 5 Exemplary schematic of species travelling to and from a microbial anode and the relationship between the individual fluxes, according to [123]. See text for explanations. In principle, the same transport processes occur at microbial cathodes used for bioelectrosynthesis, although some travelling species are different (e.g., electron acceptor instead of electron donor, OH
instead of H+)

156

S. Kerzenmacher

cell [122]. Assuming that sufficiently conductive electrode materials are used, this contribution is small compared to the ion transport resistance of the electrolyte. If not, current collectors should be implemented into the electrode structure, as discussed in Sect. 2.5. The second important aspect is the transport of H+ or OH
to or from the biofilm. Considering the example of an anodic biofilm illustrated in Fig. 5, a proton is generated for each electron transferred to the anode. To achieve charge balance, for each electron that travels though the electric circuit a positive charge also has to be transported from anode to cathode by migration in the electric field. In the context of microbial electrodes, the low proton concentration at neutral pH results in the preferred migration of other ions, such as Na+ or Cl
, which are present at much higher concentration. In a typical anodic biofilm, protons can account for less than 15% of the migrating ionic species [125]. As a consequence, protons accumulate at the anode, leading to a decrease in pH. This not only leads to electric losses in the form of concentration overpotentials but can also inhibit microbial growth [123]. In the case of Geobacter, inhibition already occurs at pH below 5.8. With neutral pH in the bulk, the small tolerable proton gradient of approx. 1 pH unit (difference in proton concentration approx. 1 μM) renders proton transport by diffusion insignificant and, as outlined above, migration only plays a minor role. Instead, buffering substances (typically added in the concentration range of 10–100 mM) are an important constituent of the electrolyte to prevent acidification of the biofilm and thus reduced current production [123]. Although the addition of buffers is often not economically feasible in the treatment of wastewaters, the culture media typically used in microbial electrosynthesis may offer a larger degree of freedom in optimizing their composition. In any case, the use of buffers and their recycling represents an additional cost factor. In summary, optimization strategies to minimize anodic or cathodic gradients that can be rate limiting aim at reducing the distance between anode and cathode, and at shifting from diluted media (such as municipal wastewater) to high strength media with sufficient ion strength and buffer capacity [123]. Furthermore, forced convection through the porous material can greatly improve mass transport into the pores and reduce diffusion layer thickness, as illustrated in Fig. 6. However, the forced convection comes at the expense of additional energy for pumping, which may well be acceptable in the context of bio-production. In the case of gaseous educts such as CO2, mass transport can be significantly improved by using gas diffusion electrodes, as applied, for instance, in alkaline hydrogen fuel cells [126]. The microstructure of such electrodes is composed of hydrophilic and hydrophobic domains that form the so-called triple-phase boundary between liquid electrolyte, solid electrode, and gas phase [127]. The concept is now wellestablished as the oxygen-reduction cathode in MFCs, but has only recently been implemented in the context of bioelectrosynthesis [127]. In the following section of this chapter, several macroscopic electrode topologies are compared in the light of the above-mentioned interdependencies.

Engineering of Microbial Electrodes

157

Fig. 6 Schematic comparison of diffusion and conduction lengths in flat and porous electrode topologies

3.2

A Comparison of Typical Electrode Topologies

As outlined above, the engineering of a microbial electrode structure seeks to maximize the number of electrically connected microorganisms per volume (see Fig. 1) and minimize the transport resistance of the travelling species involved in the electrode reactions. Given that any of the transport processes involved (including electron transfer in the biofilm and the electrode material) can become ratelimiting, careful balancing is required to yield optimal results. For instance, it is not useful to increase the number of electrically connected microorganisms when the accumulation of protons limits their metabolic rate. As illustrated in Fig. 6, a porous material is able to host a larger amount of biofilm as compared to a flat electrode, but at the same time the diffusion lengths in the electrolyte and biofilm increase. As a consequence, a larger gradient of the species involved in the electrochemical reaction builds up, which can be lowered by forced convection through the electrode structure, as illustrated in Fig. 6 (see also Sect. 3.1). Similarly, the electron transport resistance in the electrode material increases because of increased conduction lengths to reach the current collector, leading to voltage losses and the build-up of a potential gradient in the electrode material. Starting from flat electrodes made of, for example, glassy carbon [2, 3] or unpolished graphite [4, 5] in the early studies, researchers quickly used different three-dimensional materials with a larger specific surface area for microbial colonization. In Fig. 7, an overview of the most important electrode topologies is given. Their general characteristics are discussed in the following. The first examples are fibrous materials, such as felts, cloths, and papers. Here predominantly carbon materials such as graphite [26, 34, 41], activated carbon [26, 34], or carbonized polymer fibers fabricated by electrospinning [24, 128] are

158

S. Kerzenmacher

Fig. 7 Important topologies of microbial electrodes. See text for explanation

used. Stainless steel fiber felts have also been successfully employed [29, 129]. Because of the large aspect ratio of the fibers, these materials typically provide good electrical connection throughout their three-dimensional structure. Furthermore, with woven fibrous materials, pores on different scales can easily be achieved to realize hierarchically structured materials. For instance, individual filaments with nanopores can form microporous bundles which in turn are knitted into a three-dimensional cloth with macropores (see, e.g., the supplementary material of [34]). In this context, a positive correlation between biomass attachment and the amount of pores in the range of 5–10 μm has been found using a mixed microbial consortium [89]. A general disadvantage of fibrous materials is the risk of pore clogging, for example, by biofilm overgrowth, depending on the void space between individual fibers [23]. In the case of relatively thick electrodes, the large ion diffusion and electron conduction lengths can become an additional disadvantage. Under such circumstances, the interior of thicker electrode structures may not contribute to the bioelectrochemical reaction. A special electrode topology of fibrous material is the brush-like radial arrangement of individual fibers around a central current collector [27], as illustrated in Fig. 7. This provides good mass transport and reduces the risk of clogging. However, because of the large distances

Engineering of Microbial Electrodes

159

the ion transport resistance is worse than with planar electrodes, an effect that is particularly pronounced in wastewaters with low ionic conductivity [123]. Furthermore, a significant potential gradient along the fibers can develop, depending on their length and conductivity, as well as the current density at their surface. Other examples are electrode structures with interconnected porosity. These can be realized, for instance, by packing millimeter-sized granules made from, for example, graphite [130] or activated carbon [64] into a fixed-bed electrode [131]. These electrodes are usually voluminous, the particles filling tubes or rectangular chambers of typically more than 100 mL. Already at these smaller laboratory scales, mass transport by diffusion is limited, and convective flow is often applied by pumping the electrolyte through the porous structure [130– 132]. Thinner structures with interconnected porosity can be realized, for instance, in the form of reticulated carbon foams [74], which can also be derived from carbonization of natural products [74, 133]. Similar foam-like structures are available from stainless steel [20] and nickel [104, 120], or metal-coated polymer materials [23]. Furthermore, stainless steel particles sintered into thin porous plates have been used as microbial anodes [134]. Corrugated cardboard has also been carbonized and used as a porous electrode structure [25], the channel-like macrostructure being of advantage for mass transport. As with the fibrous materials, the increased risk of pore clogging and large diffusion and conduction lengths can be a disadvantage. For instance, with the copper-coated melamine-foam electrodes operated with the secondary acetate-fed biofilm mentioned above, a reproducible decline in the electrochemical performance over time was observed. This was attributed to the formation of a dense biofilm on the surface of the porous electrode, which reduced mass transport in the interior of the three-dimensional electrode [23]. The risk of pore clogging is particularly pronounced in the case of electrodes operated with forced convection throughout the porous structure [130, 131, 135] to counter mass transport limitations. One way to achieve convective transport into a porous structure without excessive clogging is operation of the electrode in crossflow filtration mode, as illustrated in Fig. 8. This concept has been successfully demonstrated with a microbial anode made from porous plates of sintered stainless steel that simultaneously function as filtration membranes [33, 134] for application in membrane bioreactors. A radically different electrode topology is the suspension electrode [136], in which charge-storing particles (such as activated carbon powder) are suspended in the electrolyte. In the case of microbial electrodes, an electroactive biofilm forms on the surface of these carbon particles [30, 137]. Only through percolation or intermittent contact is the electrical charge accumulated by these particles transferred to a current collector. The particles’ characteristics such as surface chemistry, morphology, double layer capacitance, and pore structure have a significant influence on the achievable current densities [31]. The advantage of suspension electrodes is the greatly improved mass transport through the active mixing (fluidization) of the carbon particles. In addition, suspension electrodes offer an attractive way to increase the reaction volume (and thus current density) without increasing the size of the current collector or, more importantly, the separator membrane,

160

S. Kerzenmacher

Fig. 8 Schematic illustration of a porous microbial anode that is simultaneously operated as filtration membrane in cross-flow configuration. The convective permeate flow through the anode increases mass transfer and thus helps to prevents proton accumulation. At the same time, the perpendicular flow across the anode surface prevents the excessive build-up of a filter cake, which would clog the porous anode structure. Adapted from [32, 33]

which is often one of the major cost factors. Furthermore, the technology can, in principle, relatively easily be scaled up. Thus, suspension electrodes appear particularly attractive for implementation in regular bioreactors, where they could provide reducing equivalents to a production organism. A disadvantage is the additional energy required for mixing and fluidization of the particles. Furthermore, it is as yet unclear to what degree the constant movement of the particles results in excessive wear upon longer operation periods. In summary, the choice and optimization of electrode topology should always be based on the specific application and the characteristics of the microorganisms used. Important factors to consider are the minimization of transport limitations and conduction losses and the risk of pore clogging. Although a highly porous electrode with large specific surface can improve both geometric and volumetric current densities, the above-mentioned study [26] with a pure culture of Geobacter demonstrates that this is not always the case. Convective flow through a porous electrode structure or the use of suspension electrodes can also improve current densities, but comes at the expense of additional pumping or mixing energy. In any case, it should also be considered that the often reported geometric current density is not the only important figure of merit. To reduce system size and thus cost of the overall systems, it is equally if not more important to maximize the volumetric current densities on the basis of the total reactor volume (see also Fig. 2 and Sect. 2.1).

Engineering of Microbial Electrodes

3.3

161

Bioelectrochemical Performance of Selected Microbial Electrodes of Practical Relevance

In the following section, an overview of the performance of selected microbial electrodes is given. With respect to the microbial anodes listed in Table 2, the focus is intentionally on MFC electrodes operated with synthetic media based on easily degradable carbon sources (for a detailed list on substrates used in microbial fuel cells see [96]). Characterizing engineered electrodes under such conditions is reasonable, as in a real wastewater environment and in bioelectrosynthesis further operational parameters such as hydraulic retention time, loading rate, and degree of carbon elimination are important, which renders the interpretation of results in the light of electrode engineering difficult. For instance, substrate availability or the degradation of a complex substrate may be the process limiting electrode current density – a limitation that must mainly be overcome by reactor engineering and operation. However, transferring the results obtained for engineered microbial anodes to microbial electrosynthesis at the cathode is certainly not straightforward, for example, because of differences in the electron transfer mechanisms or operational potentials of the electrodes. Regarding anode operation in acetate-based synthetic media, carbon cloth, stainless steel, and carbonized cardboard emerge as attractive electrodes, enabling high geometric current densities. With flame-oxidized felts (0.7 mm thick) or flat plates of stainless steel, current geometric densities of 15 A m
2 and 25 A m
2 are reported, respectively. For comparison, three-dimensional carbon cloth can deliver up to 36 A m
2, and carbonized cardboard yields 70 A m
2 for a single layer of approx. 3 mm thickness. Regarding volumetric current densities based on anode volume, one of the highest values is reported for stainless steel mesh coated with CB, delivering up to 50 mA cm
3. Here the above-mentioned flame-oxidized stainless steel felt and the carbonized cardboard reach values of up to 21 mA cm
3 and 25 mA cm
3, respectively. Significantly lower current densities are reached when anodes are operated in real wastewaters. The reported values range from maximum values of approx. 10 A m
2 for two-dimensional carbon cloths, to below 0.1 mA cm
3 for graphite felts and three-dimensional graphite granules. With lactate-fed cultures of Shewanella sp., lower current densities are reached, which is a characteristic of these microorganisms [26]. The values range from 2.5 A m
2 for a poly(aniline-co-2,4-diaminophenol) nanowire network (unfortunately its thickness is not reported) on a carbon plate, to only 0.04 A m
2 for an approx. 30 μm thick buckypaper electrode made from CNTs [34]. However, in terms of volumetric current density based on anode volume, the buckypaper electrode actually exhibits the second highest value of 1 mA cm
3, which is only superseded by the 2.4 mA cm
3 of the carbon nanofiber mat (although this value was recorded at a 200 mV more positive anode potential). Regular graphite felt, approx. 2 mm thick, exhibits the lowest volumetric current density of only 0.06 mA cm
3 [32]. Fewer studies are available with glucose-fed anodes operated with E. coli as

162

S. Kerzenmacher

Table 2 Current densities of selected microbial anodes of relevance for practical application Anode material and design Acetate-fed Carbon cloth

Stainless steel (smooth, macroand microstructured) Stainless steel foam

Stainless steel mesh coated with CB

Flat graphite

Flat graphite

Flat copper

Flat nickel

Flat stainless steel

Copper-coated melamine foam

Operation conditions

Current densities jgeo, jvol

Remarks

Ref.


0.2 V vs SCE, 40 C, optimized soil inoculum
0.2 V vs SCE, 40 C, optimized soil inoculum

32–36 A m
2


0.2 V vs SCE, 40 C, soil inoculum +0.2 V vs Ag/AgCl, 35 C, secondary biofilm

4–5 mA cm
3 63–82 A m
2


0.2 V vs SCE, 40 C, optimized soil inoculum +0.2 V vs Ag/AgCl, 35 C, secondary biofilm

8–11 A m
2


0.2 V vs Ag/AgCl, 35 C, secondary biofilm
0.2 V vs Ag/AgCl, 35 C, secondary biofilm
0.2 V vs Ag/AgCl, 35 C, secondary biofilm
0.2 V vs Ag/AgCl, 35 C, secondary biofilm

15 A m
2

[22]

4 A m
2

[22]

7 A m
2

[22]

17–25 A m
2

50 mA cm
3 15 A m
2

10 A m
2

16  5 mA cm
3 23  6 A m
2 (initially) ~ 6 mA cm
3 ~ 9 A m
2 (after ~ 30 days)

[19]

No significant difference between smooth and structured anodes

[19]

[20]

300 μm thick electrode. 50 times increased current density compared to the stainless steel mesh

[21]

[22]

Though the same study, tested at a different potential, as indicated

[22]

Geometric cur[23] rent density based on the total projected surface area of a cubic electrode (i.e., the six sides); decrease because of biofilm overgrowth of the porous structure (continued)

Engineering of Microbial Electrodes

163

Table 2 (continued) Anode material and design Electrospun carbon fiber mat Carbonized corrugated card-board structures

Operation conditions +0.2 V vs Ag/AgCl, 35 C, secondary biofilm +0.2 V vs Ag/AgCl, 35 C, secondary biofilm

Current densities jgeo, jvol 30 A m
2 1 layer: 70 A m
2; 25 mA cm
3 3 layer: 200 A m
2; 24 mA cm
3 6 layer: 390 A m
2; 22 mA cm
3

Graphite felt


0.4 V vs SCE, 30 C, G. sulfurreducens

2.0  0.2 mA cm
3 4.0  0.4 A m
2

Activated carbon cloth


0.4 V vs SCE, 30 C, G. sulfurreducens

8.2  0.4 mA cm
3 4.1  0.2 A m
2

Brush-like anode made from carbon fibers


0.45 V vs Ag/AgCl, 30 C, secondary biofilm

~ 0.25 mA cm
3 ~ 10 A m
2

Porous Ti4O7-foam prepared by ice-templating

+0.2 V vs Ag/AgCl, 32 C, G. sulfurreducens

9.5 mA cm
3 129 A m
2

Remarks Ref. Thickness of the [24] sample not reported Size of the elec[25] trodes ~ 1  1 cm2; thickness of 1 layer ~ 3 mm; Current densities after approx. 40–80 h Data extracted [26] from a polarization curve recorded in a step-wise galvanostatic technique; volumetric current density based on anode volume Data extracted [26] from a polarization curve recorded in a step-wise galvanostatic technique; volumetric current density based on anode volume Volumetric cur[27] rent density based on the volume of the anolyte chamber; geometric current density based on projected cathode area. Data extracted from a polarization curve Convective flow [28, through the three- 31, dimensional 132] anode (cylinder of approx. 1–2 cm3); current density after ~ 200 h (continued)

164

S. Kerzenmacher

Table 2 (continued) Anode material and design Stainless steel felt

Operation conditions
0.2 V vs Ag/AgCl, 30 C, secondary biofilm

Stainless steel felt (flame-oxidized)


0.2 V vs Ag/AgCl, 30 C, secondary biofilm

21.6  1.4 mA cm
3 15.1  1.0 A m
2

Fluidized activated carbon particles (100 g, ~ 1 mm, discharged on a 11 cm2 graphite plate) Fluidized activated carbon particles (200 g, ~ 1 mm, discharged on a 11 cm2 graphite plate) Single activated carbon particle placed in an anode chamber


0.3 V vs Ag/AgCl, secondary biofilm

0.003 mA cm
3 ~ 0.56 A m
2


0.3 V vs Ag/AgCl, secondary biofilm

0.004 mA cm
3 ~ 1.3 A m
2


0.3 V vs Ag/AgCl, 35 C, secondary biofilm

60 mA cm
3 (based on particle volume) 0.6 mA cm
3 (based on the volume of the anode chamber)

0 V vs NHE, 30 C, G. sulfurreducens

0.4 mA cm
3 16 A m
2 (with convection); 0.1 mA cm
3 4 A m
2 (without convec.)

Porous stainless steel filter membrane

Current densities jgeo, jvol 1.5  0.1 mA cm
3 1.1  0.1 A m
2

Remarks Ref. Current density [29] after 12 days. Volumetric current density based on anode volume Current density [29] after 7 days. Volumetric current density based on anode volume. Thickness of the felts: 0.7 mm Volumetric cur[30] rent density based on the bulk volume of the activated carbon particles Volumetric cur[30] rent density based on the bulk volume of the activated carbon particles Volumetric cur[31] rent density based on the volume of a single activated carbon particle or on the volume of the anode chamber, as indicated Electrodes of [32, 2.5 mm thickness 33, and 0.5 μm nom- 134] inal pore size, with/without convection though the porous structure (see also Fig. 8). Volumetric current density based on anode volume (continued)

Engineering of Microbial Electrodes

165

Table 2 (continued) Anode material and design Lactate-fed Graphite felt

Operation conditions

Current densities jgeo, jvol

Remarks

Ref.


0.2 V vs SCE, 30 C, S. oneidensis

0.061  0.001 mA cm
3 0.120  0.002 A m
2

[34]

Activated carbon cloth (C-Tex 13)


0.2 V vs SCE, 30 C, S. oneidensis

0.48  0.01 mA cm
3 0.240  0.003 A m
2

Buckypaper


0.164 V vs SCE, 30 C, S. oneidensis

1.05  0.01 mA cm
3 0.040  0.001 A m
2

Carbon microfiber paper

+0.043 V vs Ag/AgCl, S. oneidensis

0.11 mA cm
3 0.04 A m
2

Carbon nanofiber mat

+0.043 V vs Ag/AgCl, S. oneidensis

2.4 mA cm
3 0.36 A m
2

Poly(aniline-co2,4-diaminophenol) nanowire network on a carbon plate Polyaniline nanowire network a carbon plate

+0.2 V vs Ag/AgCl, 30 C Shewanella loihica +0.2 V vs Ag/AgCl,30 C Shewanella loihica

2.5 A m
2

Data extracted from a polarization curve recorded in a step-wise galvanostatic technique; volumetric current density based on anode volume Data extracted from a polarization curve recorded in a step-wise galvanostatic technique; volumetric current density based on anode volume Data extracted from a polarization curve recorded in a step-wise galvanostatic technique; volumetric current density based on anode volume Current density after 15 days. Volumetric current density based on anode volume Current density after 7 days. Volumetric current density based on anode volume Current density after 33 h

Current density after 33 h

[36]

0.9 A m
2

[34]

[34]

[35]

[35]

[36]

(continued)

166

S. Kerzenmacher

Table 2 (continued) Anode material and design Composite of PPy/anthraquinone2,6-disulfonic disodium salt on carbon felt Glucose-fed Composite of MWCNTs and SnO2 on glassy carbon with PTFE binder

Composite of polyaniline/mesoporous tungsten trioxide

Operation conditions
0.5 V vs SCEl,30 C S. decolorationis

Current densities jgeo, jvol 0.07 mA cm
3 1.3 A m
2

Remarks Data extracted from a polarization curve; volumetric current density based on anode volume

Ref. [37]

37 C, E. coli without the addition of a mediator

3.5 A m
2

[38]

37 C, E. coli without the addition of a mediator

3.7 A m
2

Maximum current density taken from the polarization curve of a complete cell, no individual electrode potentials reported Maximum current density taken from the polarization curve of a complete cell, no individual electrode potentials reported

0.95  0.44 A m
2

Lab-scale prototype

[40]

0.1 mA cm
3 7 A m
2

Maximum cur[41] rent density taken from the polarization curve of a complete cell, no individual electrode potentials reported Current density [42] based on total reactor volume of a microbial electrolysis cell, electrode potential not reported. (continued)

Operated with real wastewater Graphite rods +0.2 V vs Ag/AgCl, 35 C mixed culture inoculum, operated with the effluent after the primary settlement tank of a domestic waste water treatment plant Graphite felt 30 C, inoculum taken from the soil of a rice paddy field, operated with model organic wastewater based on starch, peptone, and fish extract Graphite fiber 30 C, winery wastewater rich brushes in volatile fatty acids, mixed culture inoculum

0.002–0.007 mA cm
3

[39]

Engineering of Microbial Electrodes

167

Table 2 (continued) Anode material and design

Operation conditions

Current densities jgeo, jvol

Remarks

Ref.

Data recorded between 66 and 91 days of operation, during which wastewater diluted with boiler water to keep the chemical oxygen demand between 0.7 and 2 g L
1 Current density after ~ 25 days. In the same setup, the use of synthetic medium with the same acetate content gave approx. Double current densities Current density after ~ 25 days

[43]

Carbon cloth

+0.15 V vs SCE, 27 C, municipal wastewater enriched with 10 mM acetate, secondary biofilm

~ 10 A m
2

Carbon cloth

+0.15 V vs SCE, 27 C, municipal wastewater enriched with food waste, secondary biofilm 30 C, mixed culture inoculum from an anaerobic reactor, molasses wastewater

~ 6–8 A m
2


0.4 V vs Ag/AgCl, mixed culture inoculum from an anaerobic reactor, buffered and diluted piggery wastewater

~ 0.003 mA cm
3

Graphite granules (1–5 mm)

Graphite granules (2–7 mm)

~ 0.02 mA cm
3

Maximum current density based on anode volume, extracted from a polarization curve of the complete cell. Molasses wastewater was probably fed through an anaerobic reactor prior to use in the MFC Current density based on anode chamber volume

[43]

[44]

[45]

168

S. Kerzenmacher

electroactive microorganisms, the reported current densities being in the region of up to 4 A m
2. The performance characteristics of selected microbial cathodes for bioelectrosynthesis are listed in Table 3. Given that the field of microbial electrosynthesis has emerged only in recent years, fewer studies are available as compared to microbial anodes. Comparably high current densities of approx. 12 A m
2 are reported for a stainless steel plate operated with G. sulfurreducens producing succinate from fumarate. This value is in the range of anodic current densities reached with this microorganism and underlines its ability to exchange electrons with an electrode at high current density. The current densities for methane production from CO2 are in the region of up to 9 A m
2 for a carbon felt with a mixed culture. With respect to acetate production, approx. 2.4 A m
2 are reported for a carbon cloth anode modified with graphene and tetra-ethylene pentamine using S. ovata as the microorganism. In terms of volumetric current density based on electrode volume, a maximum value of 16 mA cm
3 is reported for acetate production with a mixed microbial consortium and a carbon-nanotube-modified electrode made from reticulated vitreous carbon (RVC). Much lower values of less than 0.2 mA cm
3 have been reported for non-modified RVC. Because of the different experimental procedures and conditions, it is difficult to draw clear conclusions, particularly regarding the quantitative differences between electrode materials and topologies, or variations thereof. However, some works provide systematic comparisons, as summarized in the following. For instance, flat graphite foil was compared to carbon fiber veil and rods made from either graphite or polycrystalline carbon as anode with an acetate-fed secondary biofilm [61]. Whereas with the graphite foil, with very smooth surface, a current density of only approx. 0.07 A m
2 was reached, an almost two orders of magnitude higher current density of approx. 5 A m
2 was reached with rods made from graphite and polycrystalline carbon. Remarkably, the two flat graphite materials gave very different results, presumably because of a difference in their surface structure and/or chemistry. The highest current density of approx. 7 A m
2 was reached with a carbon fiber veil, underlining the advantage of the three-dimensional fibrous material under these experimental conditions. A similar comparison of carbon-based anode materials is available with a pure culture of S. oneidensis [34]. At a potential of
0.2 V vs SCE, the highest current density of 0.24 A m
2 was reached with activated carbon cloth of high specific surface area (C-Tex 13). Graphite felt and carbon-nanotube-based buckypaper electrodes showed lower values of 0.12 A m
2 and less than 0.04 A m
2, respectively. In terms of volumetric current density (based on the volume of the actual electrode structure), the buckypaper material delivered approx. 1 mA cm
3, which exceeded the performance of activated carbon cloth and graphite felt by factors of 2 and 17, respectively. The same graphite felt and activated carbon cloth were also evaluated as anode with the microorganism G. sulfurreducens under comparable experimental conditions [26]. In this case, both materials surprisingly exhibited the same geometric current density of approx. 4 A m
2. The similar performance of both materials was attributed to the fact that G. sulfurreducens performs only direct

Engineering of Microbial Electrodes

169

Table 3 Current densities of selected microbial cathodes of relevance for practical application Cathode material and Operation design conditions Methane production Stainless steel 30 C, winery mesh wastewater rich in volatile fatty acids, mixed culture inoculum converting approx. 90% of the H2 produced at the cathode to CH4


1.050 V vs Ag/AgCl, 30 C, mixed culture inoculum producing mainly CH4 at the cathode Graphite gran-
0.85 V vs SHE, ules 22 C, mixed cul(1.5–5 mm) ture inoculum producing mainly CH4 at the cathode Acetate production Carbon cloth
0.690 V vs SHE, modified with room temperature, reduced methanol-adapted graphene strain of oxide/tetraSporomusa ovata ethylene reducing CO2 to acetate pentamine Reticulated vit-
0.85 V vs SHE, reous carbon 35 C mixed microbial consortium reducing CO2 to acetate Reticulated vit-
0.85 V vs SHE, reous carbon 35 C mixed modified with microbial consorMWCNTs tium reducing CO2 to acetate Carbon felt

Current densities jgeo, jvol ~ 0.4 A m
2

~ 9 A m
2

Remarks

Ref.

Current density based on projected area of the cathode, electrode potential not reported. Data recorded between 66 and 91 days of operation, during which wastewater diluted with boiler water to keep the chemical oxygen demand between 0.7 and 2 g L
1 Current density based on projected area of the cathode, thickness not reported

[42]

[46]

~ 0.22 mA cm
3

Volumetric current density based on cathode chamber volume

[47]

2.4  1.1 A m
2

Approx. threefold lower current densities recorded when using the wild type of Sporomusa ovata

[48]

0.17  0.05 mA cm
3 3.7  1 A m
2

Current density based [49] on projected area or volume of the cathode. Electrode size: 1.6  1.25  0.46 cm3 Current density based [49] on projected area or volume of the cathode. Electrode size: 0.6  0.6  0.44 cm3. 70  11% of the electrons consumed where recovered in acetate (continued)

1.6  0.1 mA cm
3 37  3 A m
2

170

S. Kerzenmacher

Table 3 (continued) Cathode material and design Reticulated vitreous carbon modified with MWCNTs

Operation conditions
1.1 V vs SHE, 35 C mixed microbial consortium reducing CO2 to acetate

Succinate production Stainless steel
0.287 V vs SHE, plate 30 C, G. sulfurreducens (254SMO) reducing fumarate to succinate

Current densities jgeo, jvol ~ 16 mA cm
3 ~ 200 A m
2

~ 12 A m
2

Remarks Current density based on projected area or volume of the cathode. Cubic electrode of approx. 1.7 cm3. 99% of the electrons consumed where recovered in acetate

Ref. [50]

Current density based on total projected cathode area. Up to 280 A m
2 was achieved locally when calculated based on actual biofilm coverage of the electrode

[51]

electron transfer and forms relatively thick biofilms. This can diminish the importance of the underlying electrode structure, for example, by the overgrowing of too small pores (see also [23]). Obviously, the characteristics of the microorganisms (e.g., electron transfer mechanisms, biofilm formation capability) can be of significant importance in the choice and development of optimized electrode materials [26]. It is thus not surprising that the results of the pure culture study with G. sulfurreducens are contrary to the findings of Liu et al. [61] with a mixed microbial consortium (see above). Under these conditions, the carbon fiber veil exhibited an approx. 40% higher current density compared to a flat surface of polycrystalline carbon. In the case of stainless steel, an approx. 6 mm thick foam-structure electrode operated with an acetate-fed microbial consortium showed an approx. four times increased geometric current density of up to 80 A m
2, as compared to a smooth sheet [20]. The significant influence of convective transport through a porous electrode structure made from sintered stainless steel is apparent from different work with pure cultures of Geobacter sulfurreducens. The anode was operated in cross-flow filtration mode across its surface, which prevents the build-up of a filter cake. The convective mass transport through the porous electrode structure (perpendicular to the cross-flow; see also Fig. 8) gave fourfold higher current densities of up to 16 A m
2, as compared to operation of the electrode in a stirred electrolyte solution [33, 134]. With the suspension electrodes, first comparisons of the effect of fluidization/ mixing are available. When switching from a fixed-bed to a fluidized electrode, an increase in power density of the corresponding microbial fuel cell (operated with an acetate-fed microbial consortium) by 17% has been observed [137]. In a different

Engineering of Microbial Electrodes

171

work the amount of activated carbon particles in the reactor was doubled, leading to a twofold increase in the geometric current density at the current collector [30]. This result underlines the scalability of the suspension electrode technology, an aspect that is of particular importance for bioelectrosynthesis . However, with values in the region of 0.004 mA cm
3, the absolute current densities of fluidized bed anodes are still several order of magnitude lower, as compared, for instance, to the approx. 20 mA cm
3 reached with electrodes made from stainless steel felt or copper coated melamine foam (see Table 2). Nevertheless, in studies on individual particles, very high volumetric current densities of up to 60 mA cm
3 have been demonstrated, which supersede all the other electrodes listed in Table 2. Electrical contact and possible biofilm damage in the fluidized bed are considered to be key factors for future improvement [31]. Indeed, it would be a major advancement of this technology if the high current density achievable with a single particle could be realized on a larger scale fluidized bed. In summary, as can be seen from the data listed in Tables 2 and 3, the achievable current density depends not only on the electrode material and topology but also on the microorganism and choice of substrate. In general, acetate-fed biofilms enable the highest anode current densities, whereas with wastewaters much lower values are reached. Pure cultures of Shewanella, which are a relevant organism for bioproduction based on unbalanced (steered) fermentations [138], typically exhibit lower current densities. The achievable current densities at microbial cathodes for electrosynthesis also show a trend toward lower values, although some are wellcomparable to acetate-fed anodes. Among the flat two-dimensional materials, stainless steel enables remarkably high geometric current densities with both acetate-fed anodic consortium and a pure culture of Geobacter at the cathode. Going from flat plates to three-dimensional structures can significantly increase geometric current densities, but does not necessarily lead to higher volumetric current densities because of increasing mass transport limitations. In this context, a relatively thin composite electrode made from stainless steel and CB enables the highest volumetric current density. In comparison, anodes based on brush-like arranged carbon fibers enable values two orders of magnitude lower. With suspension electrodes, so far only comparably small volumetric current densities have been demonstrated. Here further research is necessary to tap the very promising potential of this technology.

4 Conclusion and Outlook If we compare the current densities reached in the first works with electroactive microorganisms to the latest results, it becomes clear that there has been a tremendous improvement in the engineering of microbial electrodes. Nowadays, volumetric current densities well above 1 mA cm
3 can be reached (see Tables 2 and 3). With respect to the cathodic electrosynthesis of methane, this would in theory correspond to space-time-yields of more than 3 m3 CH4 m
3 day
1 (based on the

172

S. Kerzenmacher

electrode volume). This value is already comparable to the 1–10 m3 CH4 m
3 day
1 typically reached by high-rate anaerobic digesters [137]. Likewise, regarding, for instance, ethanol production, 1 mA cm
3 would correspond to a volumetric CO2fixation rate of 6.6 g L
1 d
1 or an ethanol production rate of 3.4 g L
1 d
1. This value is at the lower end of the productivity range reported for syngas fermentations using Clostridium strains (between 3.4 g L
1 d
1 and 22.58 L
1 d
1 at maximum pressures of 1 bar [138–140], up to 369 g L
1 d
1 at 6 bar [141, 142]). However, it should not be forgotten that there is a fundamental difference between the simple design of an anaerobic digester and the elaborate construction needed to place a large number of electrodes into a reactor. Furthermore, as these current densities have so far only been demonstrated with electrodes of only a few cubic centimeters in size, future research must be focused on larger-scale electrodes. To achieve this, development of hierarchically structured electrodes with optimized microbial colonization and minimized transport resistances is a promising avenue to enable high productivity and cost-efficiency of microbial electrodes on a larger scale. Similarly, electrode concepts with improved mass transport, such as suspension electrodes or operation with forced convection, can help to increase electrode performance. In this context, it would be desirable to develop a complete model that connects biofilm formation and microbial electron transfer processes to the topology of a three-dimensional electrode, its specific surface area, and the various transport resistances. With such a model, the various interdependencies between these parameters could be studied, and electrode topologies could be optimized toward maximum volumetric current densities for a given application. For practical application, the development of cost-efficient materials and electrode structures is highly relevant, particularly with respect to the up-scaling of electrode designs and fabrication processes. Promising new materials are composite materials, electrically conductive ceramics, and possibly also electrically conductive redox-active polymers with improved long-term stability. As several examples mentioned in this chapter show, there is the risk that the advantage of a particular material, surface modification, or electrode topology vanishes over time. To assess the economic feasibility and usefulness of a certain design, experiments over sufficiently long time scales that are relevant for practical application are required. From a scientific perspective, a better understanding of the electron transfer mechanism at the cathode and the underlying mechanisms of improved microbe–electrode interaction is also required. For the meaningful communication and comparison of research advances, it is necessary to establish a good practice of performing experiments and reporting results, as outlined in Sect. 2.1 of this chapter. A particular concern with the development of microbial electrodes is the often great variation between nominally identical experiments. It is thus of the utmost importance to use well-controlled experimental setups and procedures and to perform a sufficient number of replicates. Three parallel experiments are a good start, but may not be sufficient, depending on the variability of the system under investigation. Finally, it must not be forgotten that the electrodes are always part of a larger system and their individual optimization is possible only to a certain degree. Not

Engineering of Microbial Electrodes

173

only are the reactions at anode and cathode coupled and need to occur at the same rate, but the interaction between the components, processes, and operational parameters of the overall system (e.g., type of membrane, dynamics of the microbial consortium, loading rate, reactant cross-over, etc.) can significantly influence the performance of microbial electrodes.

References 1. Guo W et al. (2017) Synergizing 13C metabolic flux analysis and metabolic engineering for biochemical production. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_2 2. Kim BH, Ikeda T, Park HS, Kim HJ, Hyun MS, Kano K, Takagi K, Tatsumi H (1999) Electrochemical activity of an Fe(III)-reducing bacterium, Shewanella putrefaciens IR-1, in the presence of alternative electron acceptors. Biotechnol Tech 13(7):475–478 3. Kim BH, Kim HJ, Hyun MS, Park DH (1999) Direct electrode reaction of Fe(III)-reducing bacterium, Shewanella putrefaciens. J Microbiol Biotechnol 9(2):127–131 4. Bond DR, Lovley DR (2003) Electricity production by Geobacter sulfurreducens attached to electrodes. Appl Environ Microbiol 69(3):1548–1555 5. Bond DR, Holmes DE, Tender LM, Lovley DR (2002) Electrode-reducing microorganisms that harvest energy from marine sediments. Science 295(5554):483–485 6. Min B, Logan BE (2004) Continuous electricity generation from domestic wastewater and organic substrates in a flat plate microbial fuel cell. Environ Sci Technol 38(21):5809–5814 7. Borole AP, Reguera G, Ringeisen B, Wang Z, Feng Y, Kim BH (2011) Electroactive biofilms: current status and future research needs. Energ Environ Sci 4(12):4813 8. Sydow A, Krieg T, Mayer F, Schrader J, Holtmann D (2014) Electroactive bacteriamolecular mechanisms and genetic tools. Appl Microbiol Biotechnol 98:8481–8495 9. Potter MC (1911) Electrical effects accompanying the decomposition of organic compounds. Proc R Soc Lond B 84(571):260–276 10. Kim HJ, Park HS, Hyun MS, Chang IS, Kim M, Kim BH (2002) A mediator-less microbial fuel cell using a metal reducing bacterium, Shewanella putrefaciens. Enzyme Microb Technol 30(2): 145–152 11. Logan BE (2008) Microbial fuel cells. Wiley, Hoboken, NJ 12. Marsili E, Baron DB, Shikhare ID, Coursolle D, Gralnick JA, Bond DR (2008) Shewanella secretes flavins that mediate extracellular electron transfer. Proc Natl Acad Sci U S A 105 (10):3968–3973 13. Chang IS, Moon HS, Bretschger O, Jang JK, Park HI, Nealson KH, Kim BH (2006) Electrochemically active bacteria (EAB) and mediator-less microbial fuel cells. J Microbiol Biotechnol 16(2):163–177 14. Liu Y, Kim H, Franklin RR, Bond DR (2011) Linking spectral and electrochemical analysis to monitor c-type cytochrome redox status in living Geobacter sulfurreducens biofilms. ChemPhysChem 12(12):2235–2241 15. Millo D, Harnisch F, Patil SA, Ly HK, Schr€oder U, Hildebrandt P (2011) In situ spectroelectrochemical investigation of electrocatalytic microbial biofilms by surface-enhanced resonance raman spectroscopy. Angew Chem Int Ed 50(11):2625–2627 16. Babauta JT, Beasley CA, Beyenal H (2014) Investigation of electron transfer by geobacter sulfurreducens biofilms by using an electrochemical quartz crystal microbalance. ChemElectroChem 1(11):2007–2016 17. Guo K, Pre´voteau A, Patil SA, Rabaey K (2015) Engineering electrodes for microbial electrocatalysis. Curr Opin Biotechnol 33:149–156 18. Rouleau S et al. (2017) RNA G-Quadruplexes as key motifs of the transcriptome. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_8

174

S. Kerzenmacher

19. Pocaznoi D, Calmet A, Etcheverry L, Erable B, Bergel A (2012) Stainless steel is a promising electrode material for anodes of microbial fuel cells. Energy Environ Sci 5(11):9645–9652 20. Ketep SF, Bergel A, Calmet A, Erable B (2014) Stainless steel foam increases the current produced by microbial bioanodes in bioelectrochemical systems. Energy Environ Sci 7(5): 1633–1637 21. Zheng S, Yang F, Chen S, Liu L, Xiong Q, Yu T, Zhao F, Schr€ oder U, Hou H (2015) Binderfree carbon black/stainless steel mesh composite electrode for high-performance anode in microbial fuel cells. J Power Sources 284:252–257 22. Baudler A, Schmidt I, Langner M, Greiner A, Schroder U (2015) Does it have to be carbon? Metal anodes in microbial fuel cells and related bioelectrochemical systems. Energy Environ Sci 8:2048–2055 23. Baudler A, Langner M, Rohr C, Greiner A, Schroder U (2017) Metal-polymer hybrid architectures as novel anode platform for microbial electrochemical technologies. ChemSusChem 10(1):253–257 24. Chen SL, Hou HQ, Harnisch F, Patil SA, Carmona-Martinez AA, Agarwal S, Zhang YY, Sinha-Ray S, Yarin AL, Greiner A, Schr€oder U (2011) Electrospun and solution blown threedimensional carbon fiber nonwovens for application as electrodes in microbial fuel cells. Energy Environ Sci 4(4):1417–1421 25. Chen S, He G, Liu Q, Harnisch F, Zhou Y, Chen Y, Hanif M, Wang S, Peng X, Hou H, Schr€ oder U (2012) Layered corrugated electrode macrostructures boost microbial bioelectrocatalysis. Energy Environ Sci 5(12):9769 26. Kipf E, Zengerle R, Gescher J, Kerzenmacher S (2014) How does the choice of anode material influence electrical performance? A comparison of two microbial fuel cell model organisms. ChemElectroChem 1(11):1849–1853 27. Lanas V, Logan BE (2013) Evaluation of multi-brush anode systems in microbial fuel cells. Bioresour Technol 148:379–385 28. Massazza D, Parra R, Busalmen JP, Romeo HE (2015) New ceramic electrodes allow reaching the target current density in bioelectrochemical systems. Energy Environ Sci 8(9): 2707–2712 29. Guo K, Donose BC, Soeriyadi AH, Pre´voteau A, Patil SA, Freguia S, Gooding JJ, Rabaey K (2014) Flame oxidation of stainless steel felt enhances anodic biofilm formation and current output in bioelectrochemical systems. Environ Sci Technol 48(12):7151–7156 30. Deeke A, Sleutels THJA, Donkers TFW, Hamelers HVM, Buisman CJN, ter Heijne A (2015) Fluidized capacitive bioanode as a novel reactor concept for the microbial fuel cell. Environ Sci Technol 49(3):1929–1935 31. Borsje C, Liu D, Sleutels TH, Buisman CJ, ter Heijne A (2016) Performance of single carbon granules as perspective for larger scale capacitive bioanodes. J Power Sources 325:690–696 32. Danzer J, Kerzenmacher S (2014) Brennstoffzelle zur Filtration von Flüssigkeiten sowie Verwendung (German Patent DE 10 2013 012 663 B3) 33. Danzer J, Kerzenmacher S (2014) Filtration-active fuel cell (US Patent Application 2016/ 0190627 A1) 34. Kipf E, Koch J, Geiger B, Erben J, Richter K, Gescher J, Zengerle R, Kerzenmacher S (2013) Systematic screening of carbon-based anode materials for microbial fuel cells with Shewanella oneidensis MR-1. Bioresour Technol 146(0):386–392 35. Sanchez D, Jacobs D, Gregory K, Huang J, Hu Y, Vidic R, Yun M (2015) Changes in carbon electrode morphology affect microbial fuel cell performance with Shewanella oneidensis MR-1. Energies 8(3):1817–1829 36. Zhao Y, Nakanishi S, Watanabe K, Hashimoto K (2011) Hydroxylated and aminated polyaniline nanowire networks for improving anode performance in microbial fuel cells. J Biosci Bioeng 112(1):63–66 37. Feng C, Le M, Li F, Mai H, Lang X, Fan S (2010) A polypyrrole/anthraquinone-2,6disulphonic disodium salt (PPy/AQDS)-modified anode to improve performance of microbial fuel cells. Biosens Bioelectron 25(6):1516–1520

Engineering of Microbial Electrodes

175

38. Mehdinia A, Ziaei E, Jabbari A (2014) Multi-walled carbon nanotube/SnO2 nanocomposite: a novel anode material for microbial fuel cells. Electrochim Acta 130(1):512–518 39. Wang Y, Li B, Zeng L, Cui D, Xiang X, Li W (2013) Polyaniline/mesoporous tungsten trioxide composite as anode electrocatalyst for high-performance microbial fuel cells. Biosens Bioelectron 41(0):582–588 40. Brown RK, Harnisch F, Wirth S, Wahlandt H, Dockhorn T, Dichtl N, Schr€ oder U (2014) Evaluating the effects of scaling up on the performance of bioelectrochemical systems using a technical scale microbial electrolysis cell. Bioresour Technol 163:206–213 41. Shimoyama T, Komukai S, Yamazawa A, Ueno Y, Logan BE, Watanabe K (2008) Electricity generation from model organic wastewater in a cassette-electrode microbial fuel cell. Appl Microbiol Biotechnol 80(2):325–330 42. Cusick RD, Bryan B, Parker DS, Merrill MD, Mehanna M, Kiely PD, Liu G, Logan BE (2011) Performance of a pilot-scale continuous flow microbial electrolysis cell fed winery wastewater. Appl Microbiol Biotechnol 89(6):2053–2063 43. Blanchet E, Desmond E, Erable B, Bridier A, Bouchez T, Bergel A (2015) Comparison of synthetic medium and wastewater used as dilution medium to design scalable microbial anodes: application to food waste treatment. Bioresour Technol 185:106–115 44. Zhang BG, Zhao HZ, Zhou SG, Shi CH, Wang C, Ni JR (2009) A novel UASB-MFC-BAF integrated system for high strength molasses wastewater treatment and bioelectricity generation. Bioresour Technol 100(23):5687–5693 45. Ryu JH, Lee HL, Lee YP, Kim TS, Kim MK, Anh D, Tran HT, Ahn DH (2013) Simultaneous carbon and nitrogen removal from piggery wastewater using loop configuration microbial fuel cell. Process Biochem 48(7):1080–1085 46. Jiang Y, Su M, Zhang Y, Zhan G, Tao Y, Li D (2013) Bioelectrochemical systems for simultaneously production of methane and acetate from carbon dioxide at relatively high rate. Int J Hydrogen Energy 38(8):3497–3502 47. Clauwaert P, Verstraete W (2009) Methanogenesis in membraneless microbial electrolysis cells. Appl Microbiol Biotechnol 82(5):829–836 48. Chen L, Tremblay P, Mohanty S, Xu K, Zhang T (2016) Electrosynthesis of acetate from CO2 by a highly structured biofilm assembled with reduced graphene oxide–tetraethylene pentamine. J Mater Chem A 4(21):8395–8401 49. Jourdin L, Freguia S, Donose BC, Chen J, Wallace GG, Keller J, Flexer V (2014) A novel carbon nanotube modified scaffold as an efficient biocathode material for improved microbial electrosynthesis. J Mater Chem A 2(32):13093 50. Jourdin L, Freguia S, Flexer V, Keller J (2016) Bringing high-rate, CO2-based microbial electrosynthesis closer to practical implementation through improved electrode design and operating conditions. Environ Sci Technol 50(4):1982–1989 51. Pons L, De´lia ML, Bergel A (2011) Effect of surface roughness, biofilm coverage and biofilm structure on the electrochemical efficiency of microbial cathodes. Bioresour Technol 102(3): 2678–2683 52. Madjarov J, Popat SC, Erben J, G€otze A, Zengerle R, Kerzenmacher S (2017) Revisiting methods to characterize bioelectrochemical systems: the influence of uncompensated resistance (iR u-drop), double layer capacitance, and junction potential. J Power Sources 356: 408–418 53. Zhao F, Slade RCT, Varcoe JR (2009) Techniques for the study and development of microbial fuel cells: an electrochemical perspective. Chem Soc Rev 38:1926–1939 54. Ghasemi M, Daud WRW, Hassan SH, Oh S, Ismail M, Rahimnejad M, Jahim JM (2013) Nano-structured carbon as electrode material in microbial fuel cells: a comprehensive review. J Alloys Compd 580:245–255 55. Kumar GG, Sarathi VGS, Nahm KS (2013) Recent advances and challenges in the anode architecture and their modifications for the applications of microbial fuel cells. Biosens Bioelectron 43:461–475

176

S. Kerzenmacher

56. Wei J, Liang P, Huang X (2011) Recent progress in electrodes for microbial fuel cells. Bioresour Technol 102(20):9335–9344 57. Cercado-Quezada B, Delia M, Bergel A (2011) Electrochemical micro-structuring of graphite felt electrodes for accelerated formation of electroactive biofilms on microbial anodes. Electrochem Commun 13(5):440–443 58. Ringeisen BR, Ray R, Little B (2007) A miniature microbial fuel cell operating with an aerobic anode chamber. J Power Sources 165:591–597 59. Logan B, Cheng S, Watson V, Estadt G (2007) Graphite fiber brush anodes for increased power production in air-cathode microbial fuel cells. Environ Sci Technol 41(9):3341–3346 60. Srikanth S, Marsili E, Flickinger MC, Bond DR (2008) Electrochemical characterization of Geobacter sulfurreducens cells immobilized on graphite paper electrodes. Biotechnol Bioeng 99(5):1065–1073 61. Liu Y, Harnisch F, Fricke K, Schr€oder U, Climent V, Feliu JM (2010) The study of electrochemically active microbial biofilms on different carbon-based anode materials in microbial fuel cells. Biosens Bioelectron 25(9):2167–2171 62. Auer E, Freund A, Pietsch J, Tacke T (1998) Carbons as supports for industrial precious metal catalysts. Appl Catal 173(2):259–271 63. Manickam SS, Karra U, Huang L, Bui N, Li B, McCutcheon JR (2013) Activated carbon nanofiber anodes for microbial fuel cells. Carbon 53:19–28 64. Jiang D, Li B (2009) Granular activated carbon single-chamber microbial fuel cells (GAC-SCMFCs): a design suitable for large-scale wastewater treatment processes. Biochem Eng J 47(1–3):31–37 65. ElMekawy A, Hegab HM, Losic D, Saint CP, Pant D (2017) Applications of graphene in microbial fuel cells: the gap between promise and reality. Renew Sustain Energy Rev 72: 1389–1403 66. Xiao L, Damien J, Luo J, Jang HD, Huang J, He Z (2012) Crumpled graphene particles for microbial fuel cell electrodes. J Power Sources 208(0):187–192 67. Xie X, Yu G, Liu N, Bao Z, Criddle CS, Cui Y (2012) Graphene–sponges as highperformance low-cost anodes for microbial fuel cells. Energ Environ Sci 5(5):6862 68. Xie X, Hu L, Pasta M, Wells GF, Kong D, Criddle CS, Cui Y (2011) Three-dimensional carbon nanotube-textile anode for high-performance microbial fuel cells. Nano Lett 11(1): 291–296 69. Xie X, Ye M, Hu L, Liu N, McDonough JR, Chen W, Alshareef HN, Criddle CS, Cui Y (2012) Carbon nanotube-coated macroporous sponge for microbial fuel cell electrodes. Energy Environ Sci 5:5265–5270 70. Shen Y, Zhou Y, Chen S, Yang F, Zheng S, Hou H (2014) Carbon nanofibers modified graphite felt for high performance anode in high substrate concentration microbial fuel cells. Sci World J 2014:130185 71. Wang J, Li M, Liu F, Chen S (2016) Stainless steel mesh supported carbon nanofibers for electrode in bioelectrochemical system. J Nanomater 2016:1–5 72. Huggins T, Wang H, Kearns J, Jenkins P, Ren ZJ (2014) Biochar as a sustainable electrode material for electricity production in microbial fuel cells. Bioresour Technol 157:114–119 73. Yuan Y, Zhou S, Liu Y, Tang J (2013) Nanostructured macroporous bioanode based on polyaniline-modified natural loofah sponge for high-performance microbial fuel cells. Environ Sci Technol 47(24):14525–14532 74. Chen S, Liu Q, He G, Zhou Y, Hanif M, Peng X, Wang S, Hou H (2012) Reticulated carbon foam derived from a sponge-like natural product as a high-performance anode in microbial fuel cells. J Mater Chem 22(35):18609 75. Dumas C, Basseguy R, Bergel A (2008) Electrochemical activity of Geobacter sulfurreducens biofilms on stainless steel anodes. Electrochim Acta 53(16):5235–5241 76. Dumas C, Mollica A, Fe´ron D, Basseguy R, Etcheverry L, Bergel A (2008) Checking graphite and stainless anodes with an experimental model of marine microbial fuel cell. Bioresour Technol 99(18):8887–8894

Engineering of Microbial Electrodes

177

77. Zhu X, Logan BE (2014) Copper anode corrosion affects power generation in microbial fuel cells. J Chem Technol Biotechnol 89(3):471–474 78. Dutta K, Kundu PP (2014) A review on aromatic conducting polymers-based catalyst supporting matrices for application in microbial fuel cells. Polym Rev 54(3):401–435 79. Lai B, Tang X, Li H, Du Z, Liu X, Zhang Q (2011) Power production enhancement with a polyaniline modified anode in microbial fuel cells. Biosens Bioelectron 28(1):373–377 80. Zhao C, Wu J, Kjelleberg S, Loo JSC, Zhang Q (2015) Employing a flexible and low-cost polypyrrole nanotube membrane as an anode to enhance current generation in microbial fuel cells. Small 11(28):3440–3443 81. Niessen J, Schroder U, Rosenbaum M, Scholz F (2004) Fluorinated polyanilines as superior materials for electrocatalytic anodes in bacterial fuel cells. Electrochem Commun 6(6): 571–575 82. Bowen CR, Thomas T (2015) Macro-porous Ti2AlC MAX-phase ceramics by the foam replication method. Ceram Int 41(9):12178–12185 83. Jain A, Gazzola G, Panzera A, Zanoni M, Marsili E (2011) Visible spectroelectrochemical characterization of Geobacter sulfurreducens biofilms on optically transparent indium tin oxide electrode. Electrochim Acta 56(28):10776–10785 84. Jain A, Connolly JO, Woolley R, Krishnamurthy S, Marsili E (2013) Extracellular electron transfer mechanism in Shewanella loihica PV-4 biofilms formed at indium tin oxide and graphite electrodes. Int J Electrochem Sci 8(2):1778–1793 85. Lu M, Qian Y, Huang L, Xie X, Huang W (2015) Improving the performance of microbial fuel cells through anode manipulation. ChemPlusChem 80(8):1216–1225 86. Cheng S, Logan BE (2007) Ammonia treatment of carbon cloth anodes to enhance power generation of microbial fuel cells. Electrochem Commun 9(3):492–496 87. Feng Y, Yang Q, Wang X, Logan BE (2010) Treatment of carbon fiber brush anodes for improving power generation in air–cathode microbial fuel cells. J Power Sources 195(7): 1841–1844 88. Guo K, Freguia S, Dennis PG, Chen X, Donose BC, Keller J, Gooding JJ, Rabaey K (2013) Effects of surface charge and hydrophobicity on anodic biofilm formation, community composition, and current generation in bioelectrochemical systems. Environ Sci Technol 47(13): 7563–7570 89. Santoro C, Guilizzoni M, Correa Baena JP, Pasaogullari U, Casalegno A, Li B, Babanova S, Artyushkova K, Atanassov P (2014) The effects of carbon electrode surface properties on bacteria attachment and start up time of microbial fuel cells. Carbon 67:128–139 90. Chang S, Liou J, Liu J, Chiu Y, Xu C, Chen B, Chen J (2016) Feasibility study of surfacemodified carbon cloth electrodes using atmospheric pressure plasma jets for microbial fuel cells. J Power Sources 336:99–106 91. Li B, Zhou J, Zhou X, Wang X, Li B, Santoro C, Grattieri M, Babanova S, Artyushkova K, Atanassov P, Schuler AJ (2014) Surface modification of microbial fuel cells anodes: approaches to practical design. Electrochim Acta 134:116–126 92. Lamp JL, Guest JS, Naha S, Radavich KA, Love NG, Ellis MW, Puri IK (2011) Flame synthesis of carbon nanostructures on stainless steel anodes for use in microbial fuel cells. J Power Sources 196(14):5829–5834 93. Tang X, Li H, Du Z, Ng HY (2014) Spontaneous modification of graphite anode by anthraquinone-2-sulfonic acid for microbial fuel cells. Bioresour Technol 164:184–188 94. Kaneko M, Ishikawa M, Song J, Kato S, Hashimoto K, Nakanishi S (2017) Cathodic supply of electrons to living microbial cells via cytocompatible redox-active polymers. Electrochem Commun 75:17–20 95. Escapa A, San-Martin MI, Mateos R, Moran A (2015) Scaling-up of membraneless microbial electrolysis cells (MECs) for domestic wastewater treatment: bottlenecks and limitations. Bioresour Technol 180:72–78

178

S. Kerzenmacher

96. Zhang F, Cheng S, Pant D, van Bogaert G, Logan BE (2009) Power generation using an activated carbon and metal mesh cathode in a microbial fuel cell. Electrochem Commun 11(11):2177–2179 97. Gajda I, Greenman J, Melhuish C, Ieropoulos IA (2016) Electricity and disinfectant production from wastewater: microbial fuel cell as a self-powered electrolyser. Sci Rep 6:25571 98. Chen S, He G, Carmona-Martinez AA, Agarwal S, Greiner A, Hou H, Schr€ oder U (2011) Electrospun carbon fiber mat with layered architecture for anode in microbial fuel cells. Electrochem Commun 13(10):1026–1029 99. Clauwaert P, Mulenga S, Aelterman P, Verstraete W (2009) Litre-scale microbial fuel cells operated in a complete loop. Appl Microbiol Biotechnol 83(2):241–247 100. Cheng S, Ye Y, Ding W, Pan B (2014) Enhancing power generation of scale-up microbial fuel cells by optimizing the leading-out terminal of anode. J Power Sources 248:931–938 101. Antolini E (2015) Composite materials for polymer electrolyte membrane microbial fuel cells. Biosens Bioelectron 69:54–70 102. Ci S, Wen Z, Chen J, He Z (2012) Decorating anode with bamboo-like nitrogen-doped carbon nanotubes for microbial fuel cells. Electrochem Commun 14(1):71–74 103. Yu Y, Guo CX, Yong Y, Li CM, Song H (2015) Nitrogen doped carbon nanoparticles enhanced extracellular electron transfer for high-performance microbial fuel cells anode. Chemosphere 140:26–33 104. Qiao Y, Wu X, Ma C, He H, Li CM (2014) A hierarchical porous graphene/nickel anode that simultaneously boosts the bio- and electro-catalysis for high-performance microbial fuel cells. RSC Adv 4(42):21788 105. Tang J, Chen S, Yuan Y, Cai X, Zhou S (2015) In situ formation of graphene layers on graphite surfaces for efficient anodes of microbial fuel cells. Biosens Bioelectron 71:387–395 106. Lowy DA, Tender LM, Zeikus JG, Park DH, Lovley DR (2006) Harvesting energy from the marine sediment-water interface II. Kinetic activity of anode materials. Biosens Bioelectron 21(11):2058–2063 107. Peng X, Yu H, Wang X, Gao N, Geng L, Ai L (2013) Enhanced anode performance of microbial fuel cells by adding nanosemiconductor goethite. J Power Sources 223(0):94–99 108. Peng X, Yu H, Wang X, Zhou Q, Zhang S, Geng L, Sun J, Cai Z (2012) Enhanced performance and capacitance behavior of anode by rolling Fe3O4 into activated carbon in microbial fuel cells. Bioresour Technol 121:450–453 109. Jadhav DA, Ghadge AN, Ghangrekar MM (2015) Enhancing the power generation in microbial fuel cells with effective utilization of goethite recovered from mining mud as anodic catalyst. Bioresour Technol 191:110–116 110. Kato S, Hashimoto K, Watanabe K (2013) Iron-oxide minerals affect extracellular electrontransfer paths of Geobacter spp. Microb Environ 28(1):141–148 111. Xu S, Liu H, Fan Y, Schaller R, Jiao J, Chaplen F (2012) Enhanced performance and mechanism study of microbial electrolysis cells using Fe nanoparticle-decorated anodes. Appl Microbiol Biotechnol 93(2):871–880 112. Zhao C, Wang W, Sun D, Wang X, Zhang J, Zhu J (2014) Nanostructured graphene/TiO2 hybrids as high-performance anodes for microbial fuel cells. Chemistry 20(23):7091–7097 113. Tang J, Yuan Y, Liu T, Zhou S (2015) High-capacity carbon-coated titanium dioxide core– shell nanoparticles modified three dimensional anodes for improved energy output in microbial fuel cells. J Power Sources 274:170–176 114. Wen Z, Ci S, Mao S, Cui S, Lu G, Yu K, Luo S, He Z, Chen J (2013) TiO2 nanoparticlesdecorated carbon nanotubes for significantly improved bioelectricity generation in microbial fuel cells. J Power Sources 234:100–106 115. Liao Z, Sun J, Sun D, Si R, Yong Y (2015) Enhancement of power production with tartaric acid doped polyaniline nanowire network modified anode in microbial fuel cells. Bioresour Technol 192:831–834

Engineering of Microbial Electrodes

179

116. Cui H, Du L, Guo P, Zhu B, Luong JHT (2015) Controlled modification of carbon nanotubes and polyaniline on macroporous graphite felt for high-performance microbial fuel cell anode. J Power Sources 283:46–53 117. Hou J, Liu Z, Zhang P (2013) A new method for fabrication of graphene/polyaniline nanocomplex modified microbial fuel cell anodes. J Power Sources 224:139–144 118. Zhao C, Gai P, Liu C, Wang X, Xu H, Zhang J, Zhu J (2013) Polyaniline networks grown on graphene nanoribbons-coated carbon paper with a synergistic effect for high-performance microbial fuel cells. J Mater Chem A 1(40):12587 119. Yuan Y, Kim S (2008) Improved performance of a microbial fuel cell with polypyrrole/ carbon black composite coated carbon paper anodes. Bull Kor Chem Soc 29(7):1344–1348 120. Karthikeyan R, Krishnaraj N, Selvam A, Wong JW, Lee PKH, Leung MKH, Berchmans S (2016) Effect of composites based nickel foam anode in microbial fuel cell using Acetobacter aceti and Gluconobacter roseus as a biocatalysts. Bioresour Technol 217:113–120 121. Yuan Y, Shin H, Kang C, Kim S (2016) Wiring microbial biofilms to the electrode by osmium redox polymer for the performance enhancement of microbial fuel cells. Bioelectrochemistry 108:8–12 122. Zou Y, Xiang C, Yang L, Sun L, Xu F, Cao Z (2008) A mediatorless microbial fuel cell using polypyrrole coated carbon nanotubes composite as anode material. Int J Hydrogen Energy 33(18):4856–4862 123. Popat SC, Torres CI (2016) Critical transport rates that limit the performance of microbial electrochemistry technologies. Bioresour Technol 215:265–273 124. Logan BE, Hamelers B, Rozendal R, Schr€oder U, Keller J, Freguia S, Aelterman P, Verstraete W, Rabaey K (2006) Microbial fuel cells: methodology and technology. Environ Sci Technol 40(17):5181–5192 125. Marcus AK, Torres CI, Rittmann BE (2010) Evaluating the impacts of migration in the biofilm anode using the model PCBIOFILM. Electrochim Acta 55(23):6964–6972 126. Bidault F, Brett D, Middleton PH, Brandon NP (2009) Review of gas diffusion cathodes for alkaline fuel cells. J Power Sources 187(1):39–48 127. Bajracharya S, Vanbroekhoven K, Buisman CJN, Pant D, Strik DPBTB (2016) Application of gas diffusion biocathode in microbial electrosynthesis from carbon dioxide. Environ Sci Pollut Res Int 23(22):22292–22308 128. Chen S, He G, Carmona-Martinez AA, Agarwal S, Greiner A, Hou H, Schr€ oder U (2011) Electrospun carbon fiber mat with layered architecture for anode in microbial fuel cells. Electrochem Commun 13(10):1026–1029 129. Hou J, Liu Z, Yang S, Zhou Y (2014) Three-dimensional macroporous anodes based on stainless steel fiber felt for high-performance microbial fuel cells. J Power Sources 258: 204–209 130. Rabaey K, Clauwaert P, Aelterman P, Verstraete W (2005) Tubular microbial fuel cells for efficient electricity generation. Environ Sci Technol 39(20):8077–8082 131. Sun Y, Wei J, Liang P, Huang X (2011) Electricity generation and microbial community changes in microbial fuel cells packed with different anodic materials. Bioresour Technol 102(23): 10886–10891 132. Sleutels THJA, Lodder R, Hamelers HVM, Buisman CJN (2009) Improved performance of porous bio-anodes in microbial electrolysis cells by enhancing mass and charge transport. Int J Hydrogen Energy 34(24):9655–9661 133. Karthikeyan R, Wang B, Xuan J, Wong JWC, Lee PKH, Leung MKH (2015) Interfacial electron transfer and bioelectrocatalysis of carbonized plant material as effective anode of microbial fuel cell. Electrochim Acta 157(0):314–323 134. Danzer J, G€ otze A, Abdu S, Kerzenmacher S (2016) A new concept for the integration of microbial fuel cells into membrane bioreactors. In: Proceedings of the 13th IWA leading edge conference on water and wastewater technologies, Jerez de la Frontera

180

S. Kerzenmacher

135. Sleutels THJA, Hamelers HVM, Buisman CJN (2011) Effect of mass and charge transport speed and direction in porous anodes on microbial electrolysis cell performance. Bioresour Technol 102(1):399–403 136. Hatzell KB, Boota M, Gogotsi Y (2015) Materials for suspension (semi-solid) electrodes for energy and water technologies. Chem Soc Rev 44(23):8664–8687 137. Liu J, Zhang F, He W, Zhang X, Feng Y, Logan BE (2014) Intermittent contact of fluidized anode particles containing exoelectrogenic biofilms for continuous power generation in microbial fuel cells. J Power Sources 261(0):278–284 138. Flynn JM, Ross DE, Hunt KA, Bond DR, Gralnick JA (2010) Enabling unbalanced fermentations by using engineered electrode-interfaced bacteria. MBio 1(5):e00190-10 139. Geppert F, Liu D, van Eerten-Jansen M, Weidner E, Buisman C, ter Heijne A (2016) Bioelectrochemical power-to-gas: state of the art and future perspectives. Trends Biotechnol 34(11):879–894 140. Martin ME, Richter H, Saha S, Angenent LT (2016) Traits of selected Clostridium strains for syngas fermentation to ethanol. Biotechnol Bioeng 113(3):531–539 141. Shen Y, Brown R, Wen Z (2014) Syngas fermentation of Clostridium carboxidivoran P7 in a hollow fiber membrane biofilm reactor: evaluating the mass transfer coefficient and ethanol production performance. Biochem Eng J 85:21–29 142. Wang S, Huang H, Kahnt J, Mueller AP, Kopke M, Thauer RK (2013) NADP-specific electron-bifurcating FeFe-hydrogenase in a functional complex with formate dehydrogenase in Clostridium autoethanogenum grown on CO. J Bacteriol 195(19):4373–4386 143. Gaddy JL, Arora DK, Ko C, Phillips JR, Basu R, Wikstrom CV, Clausen EC (2007) Methods for increasing the production of ethanol from microbial fermentation (US Patent US7285402) 144. Liew F, Martin ME, Tappel RC, Heijstra BD, Mihalcea C, Kopke M (2016) Gas fermentationa flexible platform for commercial scale production of low-carbon-fuels and chemicals from waste and renewable feedstocks. Front Microbiol 7:694

Adv Biochem Eng Biotechnol (2019) 167: 181–202 DOI: 10.1007/10_2017_17 © Springer International Publishing AG 2017 Published online: 26 October 2017

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering Miriam A. Rosenbaum, Carola Berger, Simone Schmitz, and Ronny Uhlig

Abstract In the past 6 years, microbial bioelectrochemistry has strongly increased in attraction and audience when expanding from mainly environmental technology applications to biotechnology. In particular, the promise to combine electrosynthesis with microbial catalysis opens attractive approaches for new sustainable redoxcofactor recycling, redox-balancing, or even biosynthesis processes. Much of this promise is still not fulfilled, but it has opened and fueled entirely new research areas in this discipline. Activities in designing, tailoring, and applying specific microbial catalysts as pure or defined co-cultures for defined target bioproductions are greatly accelerating. This chapter gives an overview of the current progress as well as the emerging trends in molecular and ecological engineering of defined microbial biocatalysts to prepare them for evolving microbial electrosynthesis processes. In addition, the multitude of microbial electrosynthetic processes with complex undefined mixed cultures is covered by ter Heijne et al. (Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_15, 2017). Graphical Abstract

M.A. Rosenbaum (*), C. Berger, S. Schmitz, and R. Uhlig Institute of Applied Microbiology – iAMB, Aachen Biology and Biotechnology – ABBt, RWTH Aachen University, Worringerweg 1, 52074 Aachen, Germany e-mail: [email protected]

182

M.A. Rosenbaum et al.

Keywords Anodic and cathodic bioproductions, Defined mixed cultures, Metabolic engineering, Microbial electrosynthesis, Pure cultures Contents 1 Introduction to Microbial Electrosynthesis with Pure and Defined Mixed Cultures . . . . . 2 Current State and Strategies for Microbial Electrosynthesis with Pure and Defined Mixed Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Engineering Pure Cultures for Anodic/Oxidative Bioproductions . . . . . . . . . . . . . . . . . . 2.2 Engineering Pure Cultures for Cathodic/Reductive Bioproductions . . . . . . . . . . . . . . . . 2.3 Engineering Defined Mixed Cultures for Electroproduction . . . . . . . . . . . . . . . . . . . . . . . . 3 Trends and Strategies for Advancing Defined Culture Microbial Electrosynthesis . . . . . . 3.1 Overall Challenges for Pure and Defined Culture Microbial Electrosynthesis . . . . . 3.2 Trends in Engineering Pure Cultures for Anodic and Cathodic Bioproductions . . . 3.3 Trends in Engineering Defined Mixed Cultures for Electroproductions . . . . . . . . . . . . 4 Conclusion and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

182 183 183 186 189 192 192 194 195 197 198

1 Introduction to Microbial Electrosynthesis with Pure and Defined Mixed Cultures The term microbial electrosynthesis was first introduced in 2010 by Nevin et al. [1], who for the first time combined bioelectrochemical reactions with the synthesis of multicarbon compounds from CO2. Thus, the initial understanding of this term was limited to constructive bioelectrochemical reactions with CO2 as the substrate and possibly renewable electrical energy as the reductant. This concept quickly attracted wide interest because of its potential to provide a direct route for CO2 fixation and storage of electrical energy in chemical products. In recent years, however, the usage of this term has been expanded to include a wide range of microbial bioelectrochemical productions starting from CO2 or other carbon compounds that involve either an anode as a required electron acceptor or a cathode as a required electron donor. These electrode interactions can thereby enable bioproduction or increase its efficiency. For all cases, the electrons might be transferred between the microbial biocatalysts and the electrodes through direct cellto-electrode contact, natural or synthetic chemical electron mediators, or electroncarrying chemical reductants, such as hydrogen or formate, which directly participate in the biochemical synthesis reaction (see [2]). On the side of the microbial biocatalyst, a great variety exists. Specific microbial pure cultures might be naturally set up to perform microbial electrosynthesis, such as Sporumosa ovata or Clostridium ljungdahlii for the cathodic synthesis of acetate from CO2 [1, 3]. The great chance and promise of microbial electrosynthesis with pure cultures, however, is only tapped into when exploring recently advanced and advancing methods of metabolic engineering to tailor and optimize microbial biocatalysts for non-natural microbial electrosynthetic reactions [4]. Another often complementary strategy to achieve advanced bioproductions in bioelectrochemical systems is the use of specific, defined microbial co-cultures to share the workload in biochemical production processes. Here, metabolic and ecological engineering efforts are set to pave the way for successful collaborative productions. This

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

183

chapter focuses on these advancing strategies for targeted microbial electrosynthesis. It provides a comprehensive insight into the current status of electrosynthetic bioproductions and the tools that are underway to tailor and advance these production systems with the hope of achieving economic viability in the mid- to long-term future. Additionally, ter Heijne et al. [5] covers bioelectrochemical productions with undefined microbial mixed cultures, which is another rapidly growing route for biochemical productions from mainly CO2 and renewable electricity. The success and economic viability of these undefined microbial systems to date seems even more advanced than the pure or defined culture processes addressed here. In this case, process engineering is the main tool to direct these undefined microbial mixed cultures to produce a new and maybe higher value chemical product. In contrast, the scope and promise with pure or defined cultures lies in the production of specific target molecules, requiring initial engineering of the microbial biocatalysts or their interaction before bioprocess engineering can be applied to improve the overall production performance.

2 Current State and Strategies for Microbial Electrosynthesis with Pure and Defined Mixed Cultures 2.1

Engineering Pure Cultures for Anodic/Oxidative Bioproductions

Molecular engineering, and specifically synthetic biology, has the potential to broaden widely the spectrum of industrial application of microorganisms for bioproductions at an anode. The implementation of synthetic biology approaches can be addressed in two different manners (Fig. 1). The most obvious way is to tailor native electroactive microorganisms, such as bacteria from the Geobacteraceae or Shewanellaceae families, which already interact with the anode, toward bioproduction. The other approach is to engineer electroactivity by implementing electron conduits in common industrial hosts such as Escherichia coli or Pseudomonas putida. In 2010 Flynn and co-workers [6] succeeded in engineering the electroactive model organism Shewanella oneidensis for the bioconversion of glycerol to the more reduced ethanol. The generation of redox-balancing end products such as formate or 1,2-propanediol was avoided by shifting the electron flux toward the anode, which functioned as an electron sink. Two non-native metabolic pathways were expressed in S. oneidensis to enable glycerol uptake and utilization as well as ethanol production. The three genes necessary for the glycerol module originated from E. coli, whereas ethanol production was enabled by cloning two genes derived from Zymomonas mobilis. Indeed, the engineered S. oneidensis strain was able to produce ethanol with a productivity of 1 mmol L 1 h 1 and minor amounts of acetate as sole products. This conversion relied fully on the presence of an external

184

M.A. Rosenbaum et al.

Fig. 1 Challenges of engineering pure cultures for anodic/oxidative bioproductions. The implementation of synthetic biology approaches can be addressed in two different manners. On the one hand, native electroactive microorganisms, which already interact with the anode, can be tailored toward bioproduction (yellow microorganisms). Here, the main challenge is the limited genetic accessibility of the exoelectrogens and the high demand for suitable tools for genetic engineering. On the other hand, electroactivity can be engineered by implementing electron conduits in common industrial hosts (green microorganisms). The main challenge is the complexity of the genetic pathways and the associated high metabolic burden for the strain

electron acceptor as electron sink. Thus, the necessity of the anode to control the electron fluxes was demonstrated and the proof of concept for S. oneidensis as a chassis for bioconversion processes was completed [6]. Until now, this remains the only successful attempt in tailoring native exoelectrogens for anodic bioconversions using synthetic biology. This is partially because of the limited number of known native electrogenic microorganisms and the still ongoing research trying to elucidate the underlying mechanisms of extracellular electron transfer. Furthermore, the native electrogens have significantly limited biosynthetic capabilities compared to industrially relevant hosts [4]. Only a few exoelectrogens are easily genetically accessible as is S. oneidensis and the tools for genetic engineering of these organisms are rare. Therefore, some attempts were made to implement the pathways necessary for extracellular electron transfer directly in industrially relevant hosts. Several studies focus on the expression of the electron transport pathway of S. oneidensis MR-1 in E. coli to create a new synthetic exoelectrogen [7–13]. External electron transfer of S. oneidensis is facilitated via the Mtr pathway, which consists of the c-type cytochromes CymA, MtrA, and MtrC, and MtrB, which structurally supports the outer membrane protein complex. CymA is anchored in the inner membrane and releases electrons by oxidizing (mena)quinol. The electrons are then transferred to and through the outer membrane via MtrA (with structural help of MtrB) to MtC. In 2003, Pitts et al. succeeded in expressing

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

185

MtrA and NapC (a constitutively expressed native c-type cytochrome involved in nitrate reduction) in E. coli [10], thereby, enabling Fe3+ reduction with cell extracts, indicating that the electrons could not be transferred over the outer membrane by the native E. coli enzyme machinery. Since then, the focus has been on reconstructing the Mtr pathway piece by piece in E. coli [7–13]. For the functionality of the Mtr pathway, the need to span the electron transport chain completely from the inner to the outer membrane was unraveled. Thereby, the importance of fine tuning the expression levels of multi-heme c-type cytochromes proved crucial. In 2014, TerAvest et al. could not only show extracellular electron transfer to an anode but also a shift toward a more oxidized product spectrum, indicating a connection of extracellular electron discharge with E. coli’s central metabolism and a change in metabolic fluxes [13]. Thus, albeit at very low efficiency, an exoelectrogenic E. coli strain was generated, which can now be further developed as a platform for anodic bioproductions. However, the most difficult step of extracellular electron transfer in E. coli lays in the successful reconstruction of the outer membrane c-type cytochrome complexes. Even before these challenges are resolved, bioproductions in E. coli might become feasible by combining heterologous c-type cytochrome electron transfer with diffusible redox mediators. By adding a periplasmic cytochrome c (STC) and the electron mediator methylene blue, Sturm-Richter et al. recently demonstrated an electrode-assisted fermentation of glycerol toward more oxidized end products in an electroactive E. coli strain [12] (Table 1). Table 1 Summary of anodic biocatalyst engineering strategies and results Microbe

S. oneidensis glpF, glpK, glpD, tpiA pdc, adh expression pta knockout E. coli CymA, MtrA STC expression P. putida phzA1-G1, phzS, phzM expression P. putida –

P. putida

a

Products

Quantitative measuresa

Glycerol

Ethanol

Productivity of 1 mmol etha- [6] nol/(l  h)

Glycerol + methylene blue

Electrical current more oxidized end products Electrical current

183% accelerated electron transfer into the periplasm

Engineering Educts

ubiC expression pobA knockout

Glucose

Reference

[12]

[17] Maximum current density 12 μA cm 2 Collected charge 977 C 2-Ketogluconic Over 90% overall carbon yield [16] acid in 2-ketogluconic acid

Glucose + different redox mediators Citrate + pHBA ferricyanide

Maximal current density 12,5 mA cm 2 Maximum pHBA yield 9.91 mmolCpHBA/molCcitr.

[15]

Because of the novelty of the topic, no uniform or regular information content has been established

186

M.A. Rosenbaum et al.

In the last decade, P. putida also gained increasing biotechnological interest as a microbial production host [14] and several studies focused on establishing mediated electron transfer for bioconversions with P. putida at the anode (Table 1) [15–17]. The successful application of an anode in a production process with P. putida potentially has an additional advantage. It permits operation of a production process with an obligate aerobic organism under oxygen-limited or even anaerobic conditions, which lowers the process energy demand drastically. In 2015, Schmitz and co-workers generated an exoelectrogenic P. putida strain by introducing a pathway for phenazine redox-mediator synthesis [17]. The nine genes necessary for the production of phenazines in P. putida originated from the close relative Pseudomonas aeruginosa, which, among other functions, uses these phenazines for anodic electron discharge under oxygen-limited conditions [18]. The mediator-producing P. putida strain showed metabolic and electrochemical activity under oxygen-limited condition. Furthermore, a switch in metabolic fluxes toward lower internal storage molecules (e.g., polyhydroxyalcanoates) and a higher production of gluconate and 2-ketogluconate was recognized. This work laid a foundation for future oxygen-limited biocatalysis with P. putida strains such as rhamnolipid production, which under full aeration often causes reactor foaming. In addition, attempts using exogenously-added mediators were recently reported for biocatalysis with P. putida. In 2016, Bin Lai et al. were able to demonstrate a fully anaerobic bioconversion with P. putida via mediator-based redox balancing with the anode. By the addition of different synthetic electron shuttles, P. putida was able to convert glucose to 2-ketogluconate in very high yields [16]. Hintermayer et al. succeeded in producing p-hydroxybenzoic acid, used for the production of poly-crystal polymers, from citric acid under anaerobic conditions with an engineered P. putida strain [15]. The connection to the anode was established by supplying ferricyanide as redox-mediator, although very high concentrations of the mediator were required in this study. In conclusion, the door for genetic engineering or synthetic biology approaches to realize anodic bioproductions was successfully opened (Fig. 1). This now needs to be further advanced to realize economically feasible and competitive bioproductions.

2.2

Engineering Pure Cultures for Cathodic/Reductive Bioproductions

After the initial reports of cathodic bioproductions from CO2 with homoacetogenic bacteria [1, 3] and methanogens [19], great interest arose in the further development of this new technology for target reductive bioproductions. As the molecular interaction of the homoacetogens with the cathode are still unknown, initial efforts to improve cathodic bioproductions further focused on improvements of the bioprocess. Optimization of the growth medium, for example, was shown to improve the bioproductivity of S. ovata and redirect the electron flow during the electrosynthesis to more reduced products [20]. With redesigning and simplifying the former reactors, an increase of the

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

187

Fig. 2 Possibilities of cathodic bioproductions. Electrogens such as S. ovata (grey) can use direct electron transfer for the reduction of CO2, whereas others rely on small molecules (hydrogen and/or formate) to shuttle the electrons to the organism. These mediators can be produced by extracellular enzymes (orange) as in the case of M. maripaludis (green) or electrochemically as shown for the engineered strains of C. necator (yellow) or Pyrococcus furiosus (blue)

productivity was also achieved [21]. However, one of the greatest advantages of this new membrane-less, one pot reactor system is the easier up-scaling. Going further, metabolic engineering of cathodic biocatalysts can be used to channel the electrical energy provided by a cathode to produce a desirable product, but also to improve the carbon assimilation or the current uptake of a microorganism. As the cathodic electroreduction with microorganisms is a relatively new technology, only a few examples of the engineered production of biochemicals can be found in the literature (Fig. 2). A first example of the possibilities of metabolically-engineered cathodic bioproductions was given with a genetically modified Cupriavidus necator (formerly Ralstonia eutropha) strain, which produced significant amounts of 3-methyl-1-butanol and isobutanol (over 140 mg L 1). Here, the genes for 3-methyl-1-butanol and isobutanol production were expressed from a plasmid in a strain without polyhydroxybutyrate synthesis. In bioreactor experiments, where CO2 was electrochemically converted into formate, this strain (LH74D) now produced the C4- and C5-alcohols [22]. In other groundbreaking work, C. necator was first engineered to produce isopropanol from fructose and was later applied in an electrochemical setup. Thereby, a plasmid (pEG12), harboring the genetic information for a thiolase (phaA) and a CoA-transferase (ctf) from C. necator as well as two genes for an acetoacetate

188

M.A. Rosenbaum et al.

decarboxylase (adc) and alcohol dehydrogenase (adh) from Clostridium sp., were expressed in a non-polyhydroxybutyrate-producing mutant strain [23]. The constructed strain C. necator pEG12 was then grown in a nutrient-limited electrochemical setup, where CO2 served as carbon source and in situ generated hydrogen as energy source. The concentration of 216 mg L 1 isopropanol achieved, is so far the highest reported titer for a cathodic bioproduction [24]. Hydrogenotrophic microorganisms or microorganisms that are modified to use molecular H2 as energy source might be able to grow in a microbial electrosynthesis cell. A P. furiosus strain, which already overexpressed the soluble native hydrogenase I [25], was further engineered to use carbon dioxide as sole carbon source for the synthesis of 3-hydroxypropionate (3-HP). Thereby, the first part of the 3-HP/4hydroxybutyrate (4-HB) pathway of Metallosphaera sedula was integrated into the genome [26]. This important work showed that a carbon fixation pathway can be installed in a non-autotrophic organism and might help in the future to create new microorganisms that can perform microbial electrosynthesis from CO2. In the last few years, metabolic engineering has not only been used to aim for the production of biochemicals from CO2, it has also served as a technique to gain a better understanding of electron transfer between the cathode and the microorganism. Some well-studied electroactive microorganisms, such as Geobacter sulfurreducens or S. oneidensis, can directly accept electrons from a cathodes [27, 28], whereas other bacteria such as C. necator or the engineered P. furiosus utilize electrochemically produced molecules (e.g., hydrogen or formate) as electron donors [22, 26]. However, for a wide range of recently identified electroactive acetogenic microorganisms (e.g., S. ovata or C. ljungdahlii) or the methanogen Methanococcus maripaludis, the electron transfer is still only poorly understood (Fig. 2) [1, 3, 19]. A good example for metabolic engineering approaches to help understanding the electron uptake mechanism from a cathode was shown for the last mentioned microorganism. A mutant strain of M. maripaludis, which has no measurable hydrogenase activity, was characterized with a cathode as electron donor. With this mutant strain it was shown that small (bio)electrochemically synthesized molecules such as hydrogen or formate play an important role in cathode interaction, but also that extracellular enzymes are involved in the direct electron uptake from a cathode [29]. Once we have a solid understanding of the mechanisms (and involved elements) of cathodic electron transfer, we can in principle use a large toolbox of molecular methods to engineer desired pathways in the electroactive biocatalysts. The most challenging part is to develop these new microbial catalysts beyond a proof-ofprinciple state to reach titers, yields, and productivities that can compete with other biotechnological or chemical processes.

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

2.3

189

Engineering Defined Mixed Cultures for Electroproduction

Although the use of pure cultures is well-suited to elucidate fundamental mechanisms and to perform simple chemical conversions, more complex transformation processes might not be economically feasible at a production scale level. In particular, when using metabolically engineered pure cultures, a negative effect on the growth rate and on the resilience of the culture can be pronounced after extensive metabolic engineering [30]. These undesirable side effects often limit the number of possible alterations in a single strain [31]. Additionally, not all microorganisms with interesting electrochemical or bioconversion abilities are accessible for pure culture metabolic engineering, or the genetic basis of interesting traits is not fully resolved. By using a defined microbial mixed culture, the target biological process – and its metabolic burden – can be divided among the members of the consortia (Fig. 3a). This can lead to a more robust system that is capable of addressing wider metabolic applications than would be feasible for one organism alone [30]. Thereby, each member of the community can be designed, that is, chosen and engineered, to function in the most efficient way [32]. Especially in the field of bioelectrochemical systems research, bioproduction studies so far have mainly focused on either pure cultures or undefined mixed cultures, with the inoculum of the latter commonly derived from some sort of sewage or anaerobic sludge. Defined co-cultures to address bioelectrochemical questions are reported only rarely in scientific publications. However, potentially these defined consortia have some great advantages over undefined mixed cultures. When all the members of a consortium are known, the interactions occurring between them are limited, traceable, and more easily adjustable because of the finite number of genetic traits. This makes these systems feasible targets for many tools used in synthetic biology, such as transcriptional, translational, and post-translational manipulations [33], and allow for new approaches in ecology engineering. In a cathodic microbial electrosynthesis cell, the advantages of using a defined co-culture to separate the electron uptake step from the biosynthesis step have recently been highlighted [34]. The authors used a defined consortium of the Fe(0)corroding strain IS4 and the hydrogenotrophic archaeon Methanococcus maripaludis for methane production or the Gram positive bacterium Acetobacterium woodii for acetate formation from CO2, respectively (Table 2). Achieving very good electromethanogenesis or acetate formation rates with these co-cultures, the authors were able to reduce overpotentials by >200 mV and the electron uptake rates of the acetate forming co-culture (0.2–0.5 A m 2 at 400 mV vs SHE) were slightly higher than the rates published with pure homoacetogenic cultures [3, 34]. A further advantage of the proposed system is the opportunity to optimize two fundamentally different metabolic processes without tradeoffs. The system shows an improved metabolic flux without the accumulation of undesirable intermediates (Table 2) [34]. The study mentioned is also an example of how to deploy the spatial separation of two species in one BES (Fig. 3b) defining clear ecological niches in the reactor

190

M.A. Rosenbaum et al.

Fig. 3 Approaches for engineering defined co-cultures for bioelectrosynthesis (further explanations and example studies are given in the text). (a) Dividing the metabolic burden. (b) Spatial separation (reduction of competition). (c) Widening metabolic capacities of exoelectrogenes. (d) Scavenging dissolved oxygen. (e) Electrifying non-exoelectrogenes Table 2 Co-cultures of IS4 strain and M. maripaludis or A. woodii [34]

IS4/M. maripaludis IS4/A. woodii

Final product Methane Acetate

Product formation rate 400 mV vs SHE 0.1–0.14 μmol cm 2 h 1 0.21–0.23 μmol cm 2 h 1

500 mV vs SHE 0.6–0.9 μmol cm 2 h 1 0.57–0.74 μmol cm 2 h

1

environment. In the case of Deutzmann and Spormann (2016), the electron uptaking strain has to be in direct contact with the cathode, that is, in a biofilm, and the end product is produced by planktonic cells in the bulk of the reactor Several strategies for spatial separation that can be adapted for biosynthesis have

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

191

been developed for bioelectrochemical current production in microbial fuel cells (MFCs). Thus, Thao et al. immobilized riboflavin-secreting E. coli cells in alginate beads and used them in a defined co-culture with S. oneidensis MR-1 [35]. Through this physical separation, the competition of S. oneidensis and E. coli for active electrode surface area was minimized and a current density increase of 9.6-fold compared to a control MFC was measured. Yang et al. spatially separated riboflavin-producing E. coli from S. oneidensis CP2-1-S1 by engineering the surface properties of the latter strain to increase its adhesion to the carbon electrode [32]. This approach increased the percentage of S. oneidensis cells in the anodic biofilm from 48% to 98%, leading to a 6.8-fold higher power density. It is likely that the concept of spatial separation not only increases current production but is also directly beneficial for bioelectrosynthesis, as it can facilitate downstream processes such as product separation. Besides acting as reversible dissolved redox mediator [36], riboflavin also enhances the direct electron transfer of S. oneidensis via binding the active sites of the outer membrane c-type cytochromes [37]. These effects were also observed in a co-culture of an engineered Bacillus subtilis RH33 strain with S. oneidensis MR-1. Here, the evaluated levels of riboflavin led to a 4.9-fold increase in power density compared to S. oneidensis MR-1 pure cultures [38]. In the work above, the engineered E. coli strains not only boost the electron transfer to the anode by S. oneidensis but also widens the metabolic capacity of the MFC (Fig. 3c) [32, 35]. Although S. oneidensis is one of the best studied exoelectrogens, it has only a narrow spectrum of carbon sources that it can utilize naturally [39]. However, through co-cultures of S. oneidensis as the only exoelectrogenic organism with other microbes, xylose, glucose, or glycerol were made accessible for power generation in MFCs [32, 40, 41]. By extending the metabolic portfolio of exoelectrogens via co-culture engineering, new ideas, pathways, and concepts for the bioelectrochemical production of valuable products can be investigated in the future. In another example of a beneficial BES co-culture, Wang et al. combined S. oneidensis with P. putida for azo dye removal with simultaneous bioelectricity production [42]. Thereby, the used azo dye Congo Red was only degraded in the co-culture and not by P. putida pure cultures, and the current density was higher in the co-culture than in S. oneidensis pure cultures. Besides S. oneidensis, the native exoelectrogen G. sulfurreducens has also been used in multiple defined co-culture studies. G. sulfurreducens is a strictly anaerobic organism [43]. Hence, to ensure stable high current generation, Qu et al. employed E. coli as a non-exoelectrogenic bacterium to scavenge dissolved oxygen, which potentially leaks into the system and limits functionality of G. sulfureducens (Fig. 3d) [44]. Besides this effect, G. sulfurreducens has mainly been used in co-cultures with fermentative bacteria, where fermentation products served as electron donors for current generation with G. sulfurreducens. This approach can be used to remove unwanted fermentation byproducts and was shown to lead to a purified ethanol yield of 90% by the glycerol fermenting Clostridium cellobioparum [45]. In addition to the conversion of the unwanted byproduct

192

M.A. Rosenbaum et al.

acetate to electric current by G. sulfurreducens, an increased energy recovery was achieved by using a microbial electrolysis cell setup, where H2 was abiotically co-generated at the cathode [45, 46]. Likewise, Clostridium cellulolyticum has been shown to be synergistically more active in cellulose degradation when in a defined co-culture with G. sulfurreducens [47]. However, for a synergistic utilization of more recalcitrant or complex substrates than purified cellulose, undefined or complex mixed microbial cultures currently provide a clear advantage over pure cultures or defined co-cultures (see [5]). In a different co-culture approach, non-exoelectrogenic bacteria were electrified by the use of natural electron mediators produced by an exoelectrogenic partner (Fig. 3e) [48]. In this example, phenazines produced by Pseudomonas aeruginosa were co-utilized by Enterobacter aerogenes for enhanced synergistic current generation of both partners. This co-culture was mutualistic, as E. aerogenes had a higher energy gain through respiration with the electrode and P. aeruginosa benefited from the fermentation product 2,3-butanediol as a carbon source. However, as Read et al. showed, the type of co-culture interaction when combining non-exoelectrogens and electrogens can be variable [49]. Positive, that is, enhanced current production such as for different anode-respiring bacteria and Enterococcus faecium, negative, that is, reduced current production such as for different anode-respiring bacteria and Clostridium acetobutylicum, or neutral, that is, in a mere food-chain relationship such as with G. sulfurreducens and C. cellulolyticum [47, 49]. This shows that it is not feasible to make predictions on how two organisms interact in a co-culture. So putative partners for bioelectrosynthesis need to be tested together and tailor-made engineering solutions have to be applied to optimize the systems and maximize product yields and rates.

3 Trends and Strategies for Advancing Defined Culture Microbial Electrosynthesis 3.1

Overall Challenges for Pure and Defined Culture Microbial Electrosynthesis

There is a strong and crucial need for improving the performance of microbial electrosynthesis for use on an industrial scale. In cathodic bioproductions, the productivities achieved so far are often still close to the limit of detectability and therefore orders of magnitude lower than what are required for an economically feasible breakthrough [50]. Hence, intelligent strain and culture designs can aid in closing this gap. To date, a lot of research has been performed with undefined mixed cultures [51–53]. However, this can only be an intermediate step toward truly tailored bioproductions. As well as having to compete for suitable substrates, an undefined culture is very likely to accumulate undesired side products, which would make downstream processes quite laborious and costly [54, 55]. Nevertheless,

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

193

undefined mixed cultures can be good starting points for selecting putative candidate species for defined mixed culture or even pure culture applications [56, 57]. Thereby, one crucial biotic factor that can greatly determine the developing community is the inoculum source [54]. However, cell-to-cell interactions are also strongly influenced by the experimental setup and other abiotic factors, which influence the extracellular environment and steer the development of undefined mix-cultures. These factors need to be well-understood to establish bioprocess engineering strategies that allow for an optimal performance of the biocatalysts – no matter whether they are complex or defined. Among these factors, the set electrode potential [58], the utilized substrate [54], reactor designs [59], accessible active electrode surface area [45], and used materials [60] have proven to be of vast importance and have already greatly improved the performance of BESs [50]. In all cases, a detailed understanding is required for the microbial, the electrochemical, and the thermodynamic process – current generation in a microbial culture broth might not always be the result of a specific microbial reaction but rather of thermodynamically driven processes in the microbial electrolyte [61]. The biggest common challenge of advancing microbial pure and defined cultures for microbial electrosynthesis is the lack of understanding of their detailed metabolic processes and genetic foundations. The “standard” microbial physiology knowledge as derived over the past 100 years from very few microbial model microorganism only partly, and often very crudely, applies to the biocatalysts studied here. Therefore, significant time and resources need to be spent on profiling and understanding the specific microbial biocatalysts of interest with the recently established highly advanced methods to study microbial physiology (e.g., omics techniques, metabolic modeling, high-throughput culturing, and analytics). Further, with defined microbial co-cultures, many basic analytical methods to decipher population behavior and interactions are still underdeveloped or just emerging [33, 55, 62]. However, because defined co-cultures show great promise for many biotechnological production processes, this is a quickly advancing field of research. To make electrosynthesis with pure or defined mixed cultures economically feasible, initial niche applications for this technology have to be identified. To achieve good productivity with a specialized pure or defined culture, a controlled environment is required, which raises production costs compared to open, undefined, and mixed culture productions. Therefore, this kind of production process seems only feasible for high value and low-to-medium quantity products, not for medium-to-low value or bulk products [30]. Strict control of pure cultures is necessary to prevent contamination with other species. Defined consortia already tend to be more stable against contamination and the need for working in extreme sterile conditions is decreased [63]. When applied in continuous production, pure and defined cultures can also be subject to evolution and natural selection. This can lead to a stabilization of the culture but can also make it less predictable [64]. Additionally, interactions between pure culture populations or co-culture members are not always stable and predictable, and might change the performance of the whole system over time. The following sections highlight specific challenges and strategies for realizing viable future bioproductions with BES pure or defined mixed cultures.

194

3.2

M.A. Rosenbaum et al.

Trends in Engineering Pure Cultures for Anodic and Cathodic Bioproductions

In the last decade, synthetic biology approaches have rapidly made their way into numerous fields of biotechnology and broadened the spectrum of metabolic engineering enormously. Therefore, it is to be expected that the ongoing development of synthetic biology tools also affects the engineering of exoelectrogens. New methods such as the CRISPR-Cas9 system hold great promise to speed up the expansion of genome engineering regarding gene knock-outs, editing, repression, or activation of genetic traits [65]. In particular, the research on native exoelectrogens could benefit tremendously from such universal tools for genome engineering. However, engineering electroactivity in well-established hosts such as E. coli and P. putida also requires cutting-edge approaches because of the complexity of the molecular and metabolic pathways and associated high metabolic burdens. Up to now, the engineering attempts have focused mainly on the establishment of the electroactivity itself in the industrial host. Linking this new electroactivity to the host’s own energy metabolism, however, is still a challenging task. Future studies should go even further and also focus on the additional synthesis of products, which benefit from the oxidative capacity of the anode or the reductive power of the cathode. Cathodic microbial electrosynthesis has so far focused mainly on CO2 as a substrate for the production of green chemicals. However, although quite some work has been performed to understand and optimize the processes, little effort (or success) in metabolic engineering for non-native target chemicals has been reported. One promising process for microbial electrosynthesis from CO2 is isopropanol production with a genetically modified C. necator strain [24]. However, there are some great challenges to this approach: on the one hand, the high electron demand for CO2 conversion and the other, the limited energy supply in this process for the pathways considered so far [66]. So instead of starting with CO2, microbial electro-fermentation or the modulation or steering of microbial fermentation reactions through cathodic influence might be more feasible. The main challenge seems to be the often unsustainable use of soluble exogenous redox mediators, which are employed to carry the electrons to the non-electroactive fermenting bacteria. However, in 2014 Choi et al. found that Clostridium pasteurianum can take up electrons from the cathode without the use of a mediator. In this work they realized an increased yield in the conversion of glycerol to 1,3-propanediol or glucose to butanol with additional energy supplied from the cathode [67]. Also, for both types of cathodic bioproductions – starting from CO2 or organic substrates – new strategies in synthetic biology, such as genome editing with the CRISPR/Cas9 system or the recently established triple cross mutagenesis system, not only help to increase the productivity but also help to target new products [65, 68]. Computational supervision and guidance for the development of such microbial production processes should also gain increasing importance. As a first example, Kracke and Kr€ omer simulated beneficial production processes for anodic and cathodic electrosynthesis via elementary mode analysis [66]. They identified

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

195

several advantageous products for anode bioproductions derived from glycolysis (isoprene, 2,3-butanediol), from the tricarboxylic acid cycle and its derivatives (malate, succinate, propanol, aspartate, γ-aminobutyric acid), or from the shikimate pathway ( p-aminobenzoic acid or p-hydroxybenzoic acid). Thus, these products could be the first promising targets for anodic electrosynthesis in the recently engineered electroactive industrial hosts. For the coupling to a cathode as input for reduction equivalents, the same study showed that, with glycerol as substrate, products such as 3-hydroxypropionic acid, 1,2-propanediol, and 1,3-propanediol are more favored whereas with glucose the production of propionic acid, adipic acid, and 1,4-butanediol seem more preferred [66]. In the near future, such systematic evaluations are expected to help in understanding the benefits and limitations of microbial electrosynthesis and electro-fermentation and draw a roadmap for the molecular and biotechnological realization of these processes.

3.3

Trends in Engineering Defined Mixed Cultures for Electroproductions

Ideally, synthetic consortia can be designed in such a way that community members fill ecological niches, which would otherwise be filled by invading species, thereby stabilizing the desired co-culture [30]. This goal is aided by an increasingly comprehensive understanding of the microbiological metabolic processes and intermicrobial communication, which are required to convert complex carbon sources efficiently to targeted products and electrical energy [54]. Putative interactions between the microbes can be studied via genetic or biochemical observations through classical experiments. A complementary approach to study co-cultural behavior and interaction is via computational modeling. Thereby, more complex interactions can be revealed through genome scale reconstructions and different interactions can be simulated in a short time period and can, to a certain extent, replace work-intensive and large-scale experiments [64, 69]. The basis for modeling complex consortia is laid by firstly broadening our understanding of cellular processes via integrating omics-datasets of individual community members. Through these multi-scale models, novel molecular interaction mechanisms can be revealed, which can subsequently be implemented in more complex consortia-based models to provide new insights into the design as well as construction and finally optimization of synthetic consortia [33]. Different computational approaches exist to simulate interactions between bacterial strains and, as a valuable summary, the most recent advances and most promising modeling approaches were recently highlighted by Johns et al. [70]. One step further, active engineering of cell-to-cell communications and interactions is the center of engineering synthetic consortia [33]. In addition to putative interactions, abiotic factors such as media composition can determine whether two organisms can be active together. Also here, models were proposed to help determine optimal media conditions for synthetic communities [71, 72].

196

M.A. Rosenbaum et al.

Pandit and Mahadevan used an in silico approach to characterize systematically the effect of microbial electrosynthesis on the metabolically engineered production of biochemicals [73]. Strains for which classical engineering leads to a decrease in growth rates can benefit greatly when they are provided with additional electrons from a cathode. In bioelectrochemical applications, cathodes offer a constant electron supply, whereas anodes can act as a constant electron sink. Thus, redox balances can be drastically altered in BES and, thereby, normally unbalanced fermentation processes are able to proceed [52]. This additional oxidation or reduction power can be used to produce desired metabolites. Regarding the underlying molecular mechanism, Pandit and Mahadevan hypothesize that an enhancement in proton motive force results in an increase of the ATP and biomass yield. This effect is more pronounced the more oxidized the metabolic substrate. For other strains, however, a lack of electrical current input leads to a maximized growth but a minimized product yield. Therefore, the authors propose to use the applied electrical current as a regulatory tool in dynamic processes to keep growth rate and production rate at a desired ratio [73]. Another aspect that helps us to combine species better in a synergistic manner is our increasing understanding of how electrons are shuttled between organisms and thereby form a syntrophic interaction. For this interspecies electron transfer, two different mechanisms have been described to date: first, the contact-independent transfer of reducing equivalents, such as hydrogen and formate, as well as the use of different shuttle molecules and mediators for mediated electron transfer between microbes and, second, contact-based direct interspecies electron transfer [33, 74, 75]. Geobacter species have been shown to be capable of both modes, which qualifies these bacteria for a wide co-cultural use with other partners, such as methanogenic microorganisms that can then produce methane at the cathode of a BES [76–78]. In future years, a tremendous increase in knowledge on interspecies electron transfer is expected, which can fuel strategic co-culture development. To create reliably stable or dynamic synergistic consortia, where not one organism with time eliminates the other, population ratios often have to be optimized [64]. A way to ensure this is via cross-feeding and the creation and use of artificial or natural auxotrophies, respectively, as well as other negative and positive feedback mechanisms [30, 79, 80]. Besides supplying metabolites required for growth of a partner organism, two species can interact via cell-to-cell communication. This interspecies quorum sensing occurs naturally but can also be engineered into strains to enable population-level control of gene expression and to coordinate the different populations [33, 81]. However, the limited number of described independent communication modules and crosstalk between the different communication signals are major challenges when engineering this type of cell-to-cell communication [55]. Rather than direct production of valuable products based on electrical current, co-cultures provide the opportunity for a two-stage biochemical production with an intermediary microorganism for the synthesis of initial building blocks (e.g., formate or acetate) from recalcitrant substrates such as CO2. Partner microbes subsequently use these building blocks for the production of larger, more valuable

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

197

molecules [82]. In silico investigation of CO2-based electrosynthesis revealed that for the transformation and reduction to succinate a fairly high electron uptake flux would be required [73]. Therefore, only very low product yield is to be expected, making this process unsuitable on an industrial scale. A solution could be the co-utilization of CO2 with hexose and pentose sugars. In general, it is important to identify economically feasible products for future electroproductions. At present, inexpensive C2 compounds are oversupplied by the chemical industry, so BES-based production of these compounds is most likely not competitive. On the other hand, the bioproduction of C3–C6 bodies might be more feasible, as a shortage of these compounds is expected because of a change in chemical feedstocks [50]. Recently, Diender et al. produced medium-chain fatty acids and higher alcohols with a customized consortium of Clostridium autoethanogenum and Clostridium kluyveri and CO or syngas as carbon source [83]. In this study, hydrogen and CO acted as electron donors for reduction of acetate to ethanol. However, this system likely could almost be directly implemented in a microbial electrosynthesis cell, where a cathode provides the required reduction power. Similar ideas with hydrogen as an electron donor are proposed for a co-culture of Pseudomonas stutzeri and S. ovata for denitrification [84]. The cathodic electron uptake of S. ovata in BES is well-documented [1]. Thus, also here this synthetic consortium could be directly implemented in a BES. In summary, defined microbial co-cultures for microbial electrosynthesis applications are expected to advance greatly and gain importance as our knowledge and handling of microbial pure cultures in BESs improves. Such a process can benefit from the synergies of successful pure culture engineering and co-culture regulation and engineering as it is also developed for other new biotechnological production systems.

4 Conclusion and Outlook Today, research activities on designing and tailoring microbial pure or defined co-cultures for targeted microbial electrosynthesis still only represent a small fraction of all the activities in the fast-growing research community of microbial electrochemistry and technology. However, with the chances and potential opportunities of this application of BES, it certainly is one of the fastest growing fields within the community. Thereby, researchers are contacting and integrating yet another scientific discipline with the already very interdisciplinary activities of bioelectrochemical systems research: synthetic and systems biology. Even less explored in general is the engineering of ecological interactions between microorganisms to combine their activities in a concerted way for bioproduction. With so many new ideas, visions, and open questions, this field is an exciting playground for researchers exploring new research approaches, novel methods, and breakthrough solutions. Where this new path leads and how successful microbial electrosynthesis with pure or defined co-cultures can eventually become cannot be credibly

198

M.A. Rosenbaum et al.

predicted at this early point. However, BES research has its roots and its core activities in environmental engineering. The quickly expanding molecular engineering knowledge discussed here not only fuels the future realization of microbial bioproductions with BES but can also benefit many of the traditional environmental applications of BES, such as biosensing and bioremediation.

References 1. Nevin KP, Woodard TL, Franks AE, Summers ZM, Lovley DR (2010) Microbial electrosynthesis: feeding microbes electricity to convert carbon dioxide and water to multicarbon extracellular organic compounds. MBio 1(2):e00103–e00110 2. Simonte F et al (2017) Extracellular electron transfer and biosensors. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_34 3. Nevin KP, Hensley SA, Franks AE, Summers ZM, Ou J, Woodard TL, Snoeyenbos-West OL, Lovley DR (2011) Electrosynthesis of organic compounds from carbon dioxide is catalyzed by a diversity of acetogenic microorganisms. Appl Environ Microbiol 77(9):2882–2886 4. Rosenbaum MA, Henrich AW (2014) Engineering microbial electrocatalysis for chemical and fuel production. Curr Opin Biotechnol 29:93–98 5. ter Heijne A et al (2017) Mixed culture biocathodes for production of hydrogen, methane and carboxylates. Adv Biochem Eng Biotechnol. https://doi.org/10.1007/10_2017_15 6. Flynn JM, Ross DE, Hunt KA, Bond DR, Gralnick JA (2010) Enabling unbalanced fermentations by using engineered electrode-interfaced bacteria. mBio 1(5):e00190–10 7. Gescher JS, Cordova CD, Spormann AM (2008) Dissimilatory iron reduction in Escherichia coli: identification of CymA of Shewanella oneidensis and NapC of E. coli as ferric reductases. Mol Microbiol 68(3):706–719 8. Goldbeck CP, Jensen HM, TerAvest MA, Beedle N, Appling Y, Hepler M, Cambray G, Mutalik V, Angenent LT, Ajo-Franklin CM (2013) Tuning promoter strengths for improved synthesis and function of electron conduits in Escherichia coli. ACS Synth Biol 2(3):150–159 9. Jensen HM, Albers AE, Malley KR, Londer YY, Cohen BE, Helms BA, Weigele P, Groves JT, Ajo-Franklin CM (2010) Engineering of a synthetic electron conduit in living cells. Proc Natl Acad Sci U S A 107(45):19213–19218 10. Pitts KE, Dobbin PS, Reyes-Ramirez F, Thomson AJ, Richardson DJ, Seward HE (2003) Characterization of the Shewanella oneidensis MR-1 decaheme cytochrome MtrA: expression in Escherichia coli confers the ability to reduce soluble Fe(III) chelates. J Biol Chem 278 (30):27758–27765 11. Schuetz B, Schicklberger M, Kuermann J, Spormann AM, Gescher J (2009) Periplasmic electron transfer via the c-type cytochromes MtrA and FccA of Shewanella oneidensis MR-1. Appl Environ Microbiol 75(24):7789–7796 12. Sturm-Richter K, Golitsch F, Sturm G, Kipf E, Dittrich A, Beblawy S, Kerzenmacher S, Gescher J (2015) Unbalanced fermentation of glycerol in Escherichia coli via heterologous production of an electron transport chain and electrode interaction in microbial electrochemical cells. Bioresour Technol 186:89–96 13. TerAvest MA, Zajdel TJ, Ajo-Franklin CM (2014) The Mtr pathway of Shewanella oneidensis MR-1 couples substrate utilization to current production in Escherichia coli. ChemElectroChem 1(11):1874–1879 14. Loeschcke A, Thies S (2015) Pseudomonas putida-a versatile host for the production of natural products. Appl Microbiol Biotechnol 99(15):6197–6214 15. Hintermayer S, Yu S, Kr€omer JO, Weuster-Botz D (2016) Anodic respiration of Pseudomonas putida KT2440 in a stirred-tank bioreactor. Biochem Eng J 115:1–13

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

199

16. Lai B, Yu S, Bernhardt PV, Rabaey K, Virdis B, Kr€omer JO (2016) Anoxic metabolism and biochemical production in Pseudomonas putida F1 driven by a bioelectrochemical system. Biotechnol Biofuels 9:39 17. Schmitz S, Nies S, Wierckx N, Blank LM, Rosenbaum MA (2015) Engineering mediatorbased electroactivity in the obligate aerobic bacterium Pseudomonas putida KT2440. Front Microbiol 6:284 18. Venkataraman A, Rosenbaum M, Arends JBA, Halitsche R, Angenent LT (2010) Quorum sensing regulates electric current generation of Pseudomonas aeruginosa PA14 in bioelectrochemical systems. Electrochem Commun 12(3):459–462 19. Lohner ST, Deutzmann JS, Logan BE, Leigh J, Spormann AM (2014) Hydrogenaseindependent uptake and metabolism of electrons by the archaeon Methanococcus maripaludis. ISME J 8(8):1673–1681 20. Ammam F, Tremblay PL, Lizak DM, Zhang T (2016) Effect of tungstate on acetate and ethanol production by the electrosynthetic bacterium Sporomusa ovata. Biotechnol Biofuels 9:163 21. Giddings CG, Nevin KP, Woodward T, Lovley DR, Butler CS (2015) Simplifying microbial electrosynthesis reactor design. Front Microbiol 6:468 22. Li H, Opgenorth PH, Wernick DG, Rogers S, TY W, Higashide W, Malati P, Huo YX, Cho KM, Liao JC (2012) Integrated electromicrobial conversion of CO2 to higher alcohols. Science 335:1596 23. Grousseau E, Lu J, Gorret N, Guillouet SE, Sinskey AJ (2014) Isopropanol production with engineered Cupriavidus necator as bioproduction platform. Appl Microbiol Biotechnol 98:4277–4290 24. Torella JP, Gagliardi CJ, Chen JS, Bediako DK, Colon B, Way JC, Silver PA, Nocera DG (2015) Efficient solar-to-fuels production from a hybrid microbial-water-splitting catalyst system. Proc Natl Acad Sci U S A 112(8):2337–2342 25. Chandrayan SK, McTernan PM, Hopkins RC, Sun JS, Jenney FE, Adams MWW (2012) Engineering hyperthermophilic archaeon Pyrococcus furiosus to overproduce its cytoplasmic [NiFe]-hydrogenase. J Biol Chem 287(5):3257–3264 26. Keller MW, Schut GJ, Lipscomb GL, Menon AL, Iwuchukwu IJ, Leuko TT, Thorgersen MP, Nixon WJ, Hawkins AS, Kelly RM, Adams MWW (2013) Exploiting microbial hyperthermophilicity to produce an industrial chemical, using hydrogen and carbon dioxide. Proc Natl Acad Sci U S A 110(15):5840–5845 27. Gorby YA, Yanina S, McLean JS, Rosso KM, Moyles D, Dohnalkova A, Beveridge TJ, Chang IS, Kim BH, Kim KS, Culley DE, Reed SB, Romine MF, Saffarini DA, Hill EA, Shi L, Elias DA, Kennedy DW, Pinchuk G, Watanabe K, Ishii S, Logan B, Nealson KH, Fredrickson JK (2006) Electrically conductive bacterial nanowires produced by Shewanella oneidensis strain MR-1 and other microorganisms. Proc Natl Acad Sci U S A 103(30):11358–11363 28. Gregory KB, Bond DR, Lovley DR (2004) Graphite electrodes as electron donors for anaerobic respiration. Environ Microbiol 6(6):596–604 29. Deutzmann JS, Sahin M, Spormann AM (2015) Extracellular enzymes facilitate electron uptake in biocorrosion and bioelectrosynthesis. mBio 6(2):e00496–15 30. H€ offner K, Barton PI (2014) Design of microbial consortia for industrial biotechnology. Comput Aided Chem Eng 34:65–74 31. Bernstein HC, Carlson RP (2012) Microbial consortia engineering for cellular factories: in vitro to in silico systems. Comput Struct Biotechnol J 3(4):1–8 32. Yang Y, Wu Y, Hu Y, Cao Y, Poh CL, Cao B, Song H (2015) Engineering electrode-attached microbial consortia for high-performance xylose-fed microbial fuel cell. ACS Catal 5 (11):6937–6945 33. Song H, Ding M-Z, Jia X-Q, Ma Q, Yuan Y-J (2014) Synthetic microbial consortia: from systematic analysis to construction and applications. Chem Soc Rev 43(20):6954–6981 34. Deutzmann JS, Spormann AM (2017) Enhanced microbial electrosynthesis by using defined co-cultures. ISME J 11(3):704–714

200

M.A. Rosenbaum et al.

35. Tao L, Wang H, Xie M, Thia L, Chen WN, Wang X (2015) Improving mediated electron transport in anodic bioelectrocatalysis. Chem Commun 51(61):12170–12173 36. Marsili E, Baron DB, Shikhare ID, Coursolle D, Gralnick JA, Bond DR (2008) Shewanella secretes flavins that mediate extracellular electron transfer. Proc Natl Acad Sci U S A 105 (10):3968–3973 37. Tokunou Y, Hashimoto K, Okamoto A (2016) Acceleration of extracellular electron transfer by alternative redox-active molecules to riboflavin for outer-membrane cytochrome c of Shewanella oneidensis MR-1. J Phys Chem C 120(29):16168–16173 38. Liu T, YY Y, Chen T, Chen WN (2016) A synthetic microbial consortium of Shewanella and Bacillus for enhanced generation of bioelectricity. Biotechnol Bioeng 114(3):526–532 39. Rodionov DA, Yang C, Li X, Rodionova IA, Wang Y, Obraztsova AY, Zagnitko OP, Overbeek R, Romine MF, Reed S (2010) Genomic encyclopedia of sugar utilization pathways in the Shewanella genus. BMC Genomics 11(1):1 40. Kim C, Song YE, Lee CR, Jeon B-H, Kim JR (2016) Glycerol-fed microbial fuel cell with a co-culture of Shewanella oneidensis MR-1 and Klebsiella pneumonae J2B. J Ind Microbiol Biotechnol 43(10):1397–1403 41. Rosenbaum M, Aulenta F, Villano M, Angenent LT (2011) Cathodes as electron donors for microbial metabolism: which extracellular electron transfer mechanisms are involved? Bioresour Technol 102(1):324–333 42. Wang VB, Chua S-L, Cai Z, Sivakumar K, Zhang Q, Kjelleberg S, Cao B, Loo SCJ, Yang L (2014) A stable synergistic microbial consortium for simultaneous azo dye removal and bioelectricity generation. Bioresour Technol 155:71–76 43. Caccavo F, Lonergan DJ, Lovley DR, Davis M, Stolz JF, Mcinerney MJ (1994) Geobacter sulfurreducens sp-nov, a hydrogen-oxidizing and acetate-oxidizing dissimilatory metalreducing microorganism. Appl Environ Microbiol 60(10):3752–3759 44. Qu Y, Feng Y, Wang X, Logan BE (2012) Use of a coculture to enable current production by Geobacter sulfurreducens. Appl Environ Microbiol 78(9):3484–3487 45. Speers AM, Young JM, Reguera G (2014) Fermentation of glycerol into ethanol in a microbial electrolysis cell driven by a customized consortium. Environ Sci Technol 48(11):6350–6358 46. Speers AM, Reguera G (2012) Consolidated bioprocessing of AFEX-pretreated corn stover to ethanol and hydrogen in a microbial electrolysis cell. Environ Sci Technol 46(14):7875–7881 47. Ren Z, Ward TE, Regan JM (2007) Electricity production from cellulose in a microbial fuel cell using a defined binary culture. Environ Sci Technol 41(13):4781–4786 48. Venkataraman A, Rosenbaum MA, Perkins SD, Werner JJ, Angenent LT (2011) Metabolitebased mutualism between Pseudomonas aeruginosa PA14 and Enterobacter aerogenes enhances current generation in bioelectrochemical systems. Energy Environ Sci 4(11):4550 49. Read ST, Dutta P, Bond PL, Keller J, Rabaey K (2010) Initial development and structure of biofilms on microbial fuel cell anodes. BMC Microbiol 10(1):1 50. Harnisch F, Rosa LF, Kracke F, Virdis B, Kr€omer JO (2015) Electrifying white biotechnology: engineering and economic potential of electricity-driven bio-production. ChemSusChem 8 (5):758–766 51. Escapa A, Mateos R, Martı´nez E, Blanes J (2016) Microbial electrolysis cells: an emerging technology for wastewater treatment and energy recovery. From laboratory to pilot plant and beyond. Renew Sust Energ Rev 55:942–956 52. Moscoviz R, Toledo-Alarco´n J, Trably E, Bernet N (2016) Electro-fermentation: how to drive fermentation using electrochemical systems. Trends Biotechnol 34(11):856–865 53. Tremblay P-L, Zhang T (2015) Electrifying microbes for the production of chemicals. Front Microbiol 6:201 54. Kiely PD, Regan JM, Logan BE (2011) The electric picnic: synergistic requirements for exoelectrogenic microbial communities. Curr Opin Biotechnol 22(3):378–385 55. Shong J, Diaz MRJ, Collins CH (2012) Towards synthetic microbial consortia for bioprocessing. Curr Opin Biotechnol 23(5):798–802

Microbial Electrosynthesis I: Pure and Defined Mixed Culture Engineering

201

56. Miceli JF, Garcia-Pe~na I, Parameswaran P, Torres CI, Krajmalnik-Brown R (2014) Combining microbial cultures for efficient production of electricity from butyrate in a microbial electrochemical cell. Bioresour Technol 169:169–174 57. Sun D, Call DF, Kiely PD, Wang A, Logan BE (2012) Syntrophic interactions improve power production in formic acid fed MFCs operated with set anode potentials or fixed resistances. Biotechnol Bioeng 109(2):405–414 58. Torres CI, Krajmalnik-Brown R, Parameswaran P, Marcus AK, Wanger G, Gorby YA, Rittmann BE (2009) Selecting anode-respiring bacteria based on anode potential: phylogenetic, electrochemical, and microscopic characterization. Environ Sci Technol 43 (24):9519–9524 59. Kadier A, Simayi Y, Abdeshahian P, Azman NF, Chandrasekhar K, Kalil MS (2016) A comprehensive review of microbial electrolysis cells (MEC) reactor designs and configurations for sustainable hydrogen gas production. Alex Eng J 55(1):427–443 60. Baudler A, Schmidt I, Langner M, Greiner A, Schr€oder U (2015) Does it have to be carbon? Metal anodes in microbial fuel cells and related bioelectrochemical systems. Energy Environ Sci 8(7):2048–2055 61. Gimkiewicz C, Hunger S, Harnisch F (2016) Evaluating the feasibility of microbial electrosynthesis based on Gluconobacter oxydans. ChemElectroChem 3(9):1337–1346 62. Minty JJ, Singer ME, Scholz SA, Bae CH, Ahn JH, Foster CE, Liao JC, Lin XN (2013) Design and characterization of synthetic fungal-bacterial consortia for direct production of isobutanol from cellulosic biomass. Proc Natl Acad Sci U S A 110(36):14592–14597 63. Schievano A, Sciarria TP, Vanbroekhoven K, De Wever H, Puig S, Andersen SJ, Rabaey K, Pant D (2016) Electro-fermentation–merging electrochemistry with fermentation in industrial applications. Trends Biotechnol 34(11):866–878 64. Goers L, Freemont P, Polizzi KM (2014) Co-culture systems and technologies: taking synthetic biology to the next level. J R Soc Interface 11(96):20140065 65. Jiang W, Bikard D, Cox D, Zhang F, Marraffini LA (2013) RNA-guided editing of bacterial genomes using CRISPR-Cas systems. Nat Biotechnol 31(3):233–239 66. Kracke F, Kr€ omer JO (2014) Identifying target processes for microbial electrosynthesis by elementary mode analysis. BMC Bioinf 15:410 67. Choi O, Kim T, Woo HM, Um Y (2014) Electricity-driven metabolic shift through direct electron uptake by electroactive heterotroph Clostridium pasteurianum. Sci Rep 4:6961 68. Liew F, Martin ME, Tappel RC, Heijstra BD, Mihalcea C, Kopke M (2016) Gas fermentationa flexible platform for commercial scale production of low-carbon-fuels and chemicals from waste and renewable feedstocks. Front Microbiol 7:694 69. Zengler K, Palsson BO (2012) A road map for the development of community systems (CoSy) biology. Nat Rev Microbiol 10(5):366–372 70. Johns NI, Blazejewski T, Gomes AL, Wang HH (2016) Principles for designing synthetic microbial communities. Curr Opin Microbiol 31:146–153 71. Klitgord N, Segre D (2010) Environments that induce synthetic microbial ecosystems. PLoS Comput Biol 6(11):e1001002 72. Zampieri M, Sauer U (2016) Model-based media selection to minimize the cost of metabolic cooperation in microbial ecosystems. Bioinformatics 32(11):1733–1739 73. Pandit AV, Mahadevan R (2011) In silico characterization of microbial electrosynthesis for metabolic engineering of biochemicals. Microb Cell Factories 10(1):1 74. Kouzuma A, Kato S, Watanabe K (2015) Microbial interspecies interactions: recent findings in syntrophic consortia. Front Microbiol 6:477 75. Shrestha PM, Rotaru AE (2014) Plugging in or going wireless: strategies for interspecies electron transfer. Front Microbiol 5:237 76. Kobayashi H, Saito N, Fu Q, Kawaguchi H, Vilcaez J, Wakayama T, Maeda H, Sato K (2013) Bio-electrochemical property and phylogenetic diversity of microbial communities associated with bioelectrodes of an electromethanogenic reactor. J Biosci Bioeng 116(1):114–117

202

M.A. Rosenbaum et al.

77. Rotaru A-E, Shrestha PM, Liu F, Ueki T, Nevin K, Summers ZM, Lovley DR (2012) Interspecies electron transfer via hydrogen and formate rather than direct electrical connections in cocultures of Pelobacter carbinolicus and Geobacter sulfurreducens. Appl Environ Microbiol 78(21):7645–7651 78. Summers ZM, Fogarty HE, Leang C, Franks AE, Malvankar NS, Lovley DR (2010) Direct exchange of electrons within aggregates of an evolved syntrophic coculture of anaerobic bacteria. Science 330(6009):1413–1415 79. Chuang JS, Rivoire O, Leibler S (2010) Cooperation and Hamilton’s rule in a simple synthetic microbial system. Mol Syst Biol 6(1):398 80. Wintermute EH, Silver PA (2010) Emergent cooperation in microbial metabolism. Mol Syst Biol 6(1):407 81. Clinton A, Rumbaugh KP (2015) Interspecies and interkingdom signaling via Quorum signals. Isr J Chem 56(5):265–272 82. Rabaey K, Rozendal RA (2010) Microbial electrosynthesis — revisiting the electrical route for microbial production. Nat Rev Microbiol 8(10):706–716 83. Diender M, Stams AJ, Sousa DZ (2016) Production of medium-chain fatty acids and higher alcohols by a synthetic co-culture grown on carbon monoxide or syngas. Biotechnol Biofuels 9 (1):1 84. Xiao Z, Awata T, Zhang D, Katayama A (2016) Denitrification by Pseudomonas stutzeri coupled with CO2 reduction by Sporomusa ovata with hydrogen as an electron donor assisted by solid-phase humin. J Biosci Bioeng 122(3):307–313

Adv Biochem Eng Biotechnol (2019) 167: 203–230 DOI: 10.1007/10_2017_15 © Springer International Publishing AG 2017 Published online: 26 October 2017

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates Annemiek ter Heijne, Florian Geppert, Tom H.J.A. Sleutels, Pau Batlle-Vilanova, Dandan Liu, and Sebasti
a Puig

Abstract Formation of hydrogen, methane, and organics at biocathodes is an attractive new application of bioelectrochemical systems (BESs). Using mixed cultures, these products can be formed at certain cathode potentials using specific operating conditions, of which pH is important. Thermodynamically, the reduction of CO2 to methane is the most favorable reaction, followed by reduction of CO2 to acetate and ethanol, and hydrogen. In practice, however, the cathode potential at which these reactions occur is more negative, meaning that more energy is required to drive the reaction. Therefore, hydrogen is often found as a second product or

A. ter Heijne (*) and D. Liu Sub-Department of Environmental Technology, Wageningen University, Bornse Weilanden 9, 6708 WG Wageningen, The Netherlands e-mail: [email protected] F. Geppert Fraunhofer Institute for Environmental, Safety, and Energy Technology UMSICHT, Osterfelder Str. 3, 46047 Oberhausen, Germany Department of Mechanical Engineering, Ruhr-University Bochum, Universita¨tsstr. 150, 44801 Bochum, Germany T.H.J.A. Sleutels Wetsus, European Centre of Excellence for Sustainable Water Technology, Oostergoweg 9, 8911 MA Leeuwarden, The Netherlands P. Batlle-Vilanova LEQUiA, Institute of the Environment, University of Girona, C/Maria Aure`lia Capmany, 69, Facultat de Cie`ncies, E-17071 Girona, Spain Department of Innovation and Technology, FCC Aqualia, Balmes Street, 36, 6th Floor, 08007 Barcelona, Spain S. Puig LEQUiA, Institute of the Environment, University of Girona, C/Maria Aure`lia Capmany, 69, Facultat de Cie`ncies, E-17071 Girona, Spain

204

A. ter Heijne et al.

intermediate in the conversion of CO2 to both methane and carboxylates. In this chapter we summarize the inocula used for biocathode processes and discuss the achieved conversion rates and cathode potentials for formation of hydrogen, methane, and carboxylates. Although this overview reveals that BESs offer new opportunities for the bioproduction of different compounds, there are still challenges that need to be overcome before these systems can be applied on a larger scale. Graphical Abstract

Keywords Biocathode, Bioelectrochemical systems, Carboxylate, Hydrogen, Methane Contents 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Production of Value-Added Products or Valuable Commodities at Biocathodes: Methods and Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Thermodynamics of Bioproduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Sources for Mixed Cultures Used in Bioproduction of Hydrogen, Methane, and Carboxylates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Hydrogen Production at Abiotic and Biotic Cathodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Products: Hydrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Typical Potentials and Production Rates for Hydrogen Production . . . . . . . . . . . . . . . . . 3.3 Microorganisms for Biotic Hydrogen Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Methane Production at Biocathodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 Products: Methane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Typical Potentials and Production Rates in Methane-Producing BESs . . . . . . . . . . . . . 4.3 Microorganisms for Methane Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Carboxylate Production at Biocathodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1 Products: Volatile Fatty Acids and Alcohols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Typical Potentials and Conversion Rates for Carboxylate Production . . . . . . . . . . . . . . 5.3 Microorganisms for Carboxylate Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

205 205 207 208 210 210 210 212 212 212 214 217 217 217 220 222

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates

205

6 Design Aspects for BESs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 7 Closing Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

1 Introduction Bioelectrochemical systems (BESs) use a combination of electrodes and microorganisms for different types of conversions. Research into BESs started with the discovery that microorganisms could exchange electrons with the anode without the use of mediators [1]. These so-called bioanodes formed the start of development of microbial fuel cells (MFCs) which convert chemical energy from organic material, for example, sugars and fatty acids in wastewater into electricity [2]. Soon afterward, it was found that bioanodes could drive the formation of hydrogen in microbial electrolysis cells (MECs) when using electrical energy as the input [3]. This provided a new, energy-efficient means to produce hydrogen from wastewater and electricity. The research area expanded further from bioanodes toward biocathodes, in which microorganisms and electrodes were used for different kinds of reduction reactions, for example, oxygen reduction [4, 5], denitrification [6], and formation of hydrogen [7], methane [8], and carboxylates: organic acids and alcohols [9, 10]. Production of hydrogen at cathodes is attractive, for hydrogen not only to be used as energy carrier, for example, for transport, but also to be used as electron donor in other processes. For example, it can act as a mediator for the conversion of carbon dioxide (CO2) into methane and other organic products. The production of methane, acetate, and carboxylates at biocathodes provides a means to produce organics from electricity and CO2 alone, without the need for agricultural land to grow crops. It offers a new way to produce organics with a smaller environmental footprint, for example, in terms of water, nutrients, and required area. In addition, organics produced at biocathodes may be used to store (excess) renewable electricity, for example, in the form of methane [11], or in the form of acetate, for example, forming a microbial rechargeable battery when coupled to an MFC [12]. The process of using CO2 as the carbon source in combination with renewable electricity to produce fuels and precursors for chemicals in BESs is nowadays known as microbial electrosynthesis (MES). It represents not only a promising approach that could reduce the emissions of CO2 in the future but also a technology that helps the transition toward a circular economy. This chapter reviews the recent developments of hydrogen, methane, and carboxylate production at mixed culture biocathodes.

1.1

Production of Value-Added Products or Valuable Commodities at Biocathodes: Methods and Mechanisms

Different approaches and routes can be used to produce components at biocathodes: (1) using enzymes as a catalyst [13], (2) using microorganisms in pure cultures [14],

206

A. ter Heijne et al.

and (3) using microorganisms in mixed cultures (this chapter). The characteristics of each approach are summarized in Table 1. Whereas enzymes and pure cultures are advantageous when high specificity is required, the use of mixed cultures offers advantages in terms of flexibility and resilience. Hydrogen, methane, and carboxylates can be produced at biocathodes via different mechanisms (Fig. 1). Although each component, in theory, can be produced directly at the cathode (direct electrode-based reactions), in many cases the compounds are produced indirectly via intermediates (secondary reactions). The main intermediate is hydrogen, which can shuttle electrons between the electrode and microorganisms or from one microorganism to the other [15]. In some cases, other electron mediators such as formate are being formed, either biologically or electrochemically [16]. Mediators such as 9,10-anthraquinone-2,7-disulphonic acid Table 1 Characteristics for the use of enzymes, pure cultures, and mixed cultures at biocathodes Product specificity Enzymes High Microorganisms Mixed Low–high cultures Pure High cultures

Lifetime Low

Operational cost High

Flexibility (type of substrate) Low

Sensitivity to changing conditions High

High

Low

High

Low

High

Moderate

Low

High

Fig. 1 Mechanisms for bioproduction of hydrogen, methane, and carboxylates: direct electrodebased reactions and secondary reactions

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates

207

and methyl viologen have been added to the catholyte to facilitate electron transfer between electrodes and microorganisms [17].

1.2

Thermodynamics of Bioproduction

The change in Gibb’s free energy change (ΔG) for a reaction determines the maximum available energy at equilibrium. The Gibb’s free energy for biocathodes is determined by the cathode potential (electron donor) in combination with the electron-accepting components involved in the reaction and their concentrations. As cathode potential becomes more negative, a higher overpotential is applied, meaning that more energy is provided to drive the cathodic reduction reaction. Figure 2 links the cathode potential to the Gibb’s free energy of the reactions for hydrogen, ethanol, acetate, and methane, expressed per mole of electrons transferred. All values have been calculated at pH ¼ 7, 30 C, 50 mM carboxylates and 50 mM HCO3 concentration, and 1 bar gas pressures, as explained elsewhere [18, 19]. In this chapter, all potentials are expressed vs Normal Hydrogen Electrode (NHE). Thermodynamically, the direct production of methane from HCO3 occurs at 240 mV vs NHE, and is the most favorable reaction of all reactions addressed in this chapter. It thus occurs at the least negative cathode potential. Direct production of acetate would require a cathode potential 40 mV more negative, whereas ethanol requires a cathode potential 60 mV more negative. Hydrogen production is the least favorable reaction because it requires a cathode potential 170 mV more negative than direct production of methane. It has to be noted here that the Gibb’s free energy is strongly dependent on both pH and the gas partial pressures. Local pH and local gas pressures are often not measured and are unknown; these affect the thermodynamic values considerably.

Fig. 2 Gibb’s free energy change of the reactions increases with more negative cathode potentials (calculated at pH 7, 30 C, 50 mM carboxylates and 50 mM HCO3 concentration, and 1 bar gas pressure). These cathode potentials are equilibrium values, so at zero net reaction rate. A higher overpotential is needed for microbial growth and maintenance, and to drive the reactions

208

A. ter Heijne et al.

The minimal Gibb’s free energy change for microbial growth and maintenance is normally assumed to be in the range 15 to 20 kJ/mol reaction [20, 21]. Taking this minimum energy for microbial reactions into account, an additional overpotential is required for the reactions to occur. Thermodynamic calculations represent the system in equilibrium, meaning that the forward and backward reactions occur at equal rates, and the net reaction rate is zero. It therefore reflects the minimum required energy input. However, when reactions are running, the system moves away from equilibrium, resulting in higher energy inputs than thermodynamics dictate. This higher energy input is required to overcome internal resistances in the system [22]. These internal resistances are related to anode overpotential, cathode overpotential, ion transport across the membrane, and losses as a result of ionic conductivity; examples of typical cell voltages required to overcome internal resistances are given in the following sections. All these factors together determine the eventual performance of the system in terms of conversion rate and the related energy input [22].

2 Sources for Mixed Cultures Used in Bioproduction of Hydrogen, Methane, and Carboxylates Mixed microbial communities have been widely used by researchers to carry out biological transformations in BESs. Although single species can provide better controlled environments and higher specificity, mixed cultures offer higher flexibility and resilience. Thus, they are in many cases better candidates for future applications and scaling-up. Mixed cultures adapt to the operational conditions and those species that are able to carry out the process of interest have a competitive advantage and therefore constitute the microbiome of the process [23]. Mixed cultures can develop syntrophy in BESs as they do in anaerobic digestion and other environments [24], so the whole process is the result of a series of interactions between the different microorganisms present in the final microbiome. Syntrophic relations can be established among the microbial community from many perspectives, which include electron transfer, oxygen removal, and by-product transformation [24, 25]. Without these interactions, the operation of BESs with mixed cultures may not be successful. There is a wide range of sources from which microorganisms can be extracted to develop microbiomes for the production of valuable compounds. These sources can be divided into two categories: natural environments and enriched environments. The natural environment usually refers to wastewater treatment sludge, sediments, and soil, as these sources contain naturally-selected methanogens, fermentative and acetogenic microorganisms. The enriched environment usually means to use inoculum from other already running BESs to make sure that electrochemically active microorganisms are present, or to use pre-enrichments of specific microorganisms,

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates

209

especially when there is knowledge about the target species and enrichment procedures. Table 2 presents a summary of the sources of inoculum typically used for the production of hydrogen, methane, and carboxylates. Developing bioelectrochemical activity from mixed cultures is a rather simple process, but it may take a long time to get high and stable production rates [48, 49]. Therefore, many studies used previously enriched inocula from parent BESs or other systems to obtain bioelectrochemical activity more quickly. It has to be noted that hydrogen and methane producing activity is usually observed prior to acetogenic activity [41]. Therefore, enriching a microbiome intended for carboxylates production is a rather complex process. This is because acetogens are often outcompeted in mixed cultures by the presence of other microorganisms, especially methane producers, which can act as electron scavengers (hydrogenotrophic methanogens) or consume the organic products to produce methane (acetoclastic methanogens) under both mesophilic and thermophilic conditions. Therefore, methane inhibition techniques, such as low pH, short hydraulic retention times, thermal shock, or addition of inhibitors (i.e., 2-bromoethanesulfonate) [54], are usually applied in BESs to carry out MES or bio-hydrogen-mediated production of organic compounds.

Table 2 Summary of the microorganism sources typically used in BESs for the production of hydrogen, methane, and organic acids Product Hydrogen

Methane

Carboxylates

a

Inoculum sourcea Anaerobic sludge (anaerobic digestion) Activated sludge (wastewater treatment) Pond sediments Parent BES-enriched Enriched culture-hydrogenophilic dechlorinating culture Anaerobic sludge (distillery wastewater) Anaerobic sludge (winery digester) Anaerobic sludge (wastewater treatment) Parent BES-enriched Enriched culture – Methanobacterium palustre Anaerobic basin (brewery wastewater) Anaerobic basin (manure) Anaerobic sludge (wastewater treatment) Pond sediments Parent BES-enriched Enriched culture – Clostridiales

Inoculum source used in the BESs, either entirely or mixed with others

Reference [26] [27] [26] [7, 27–31] [32] [11] [33] [34–36] [37–39] [8, 25, 40] [41] [42] [23, 43–47] [43, 47] [48–52] [47, 53]

210

A. ter Heijne et al.

3 Hydrogen Production at Abiotic and Biotic Cathodes 3.1

Products: Hydrogen

Hydrogen production at a cathode is a way to transfer electrons at high energy level (or a highly reduced component) into the electrolyte. Hydrogen can be further used in the previously described secondary bioproduction processes, but hydrogen by itself is already a valuable product [55]. It can be used as a chemical, for example, in ammonia production and petroleum processing, as fuel, for example, in power production and vehicles, and as a raw material to be used in the food and pharmaceutical industry [55, 56]. Predictions of future hydrogen consumption may increase up to 22  1018 J in 2050, mostly depending on its application in transportation [57]. As hydrogen is not present in large quantities on Earth, it is an energy carrier rather than an energy source, and needs to be produced using an energy and proton source. Nowadays, 50% hydrogen is produced from fossil fuels such as natural gas, although more innovative technologies are being explored to produce hydrogen in eco-friendly ways. MECs are an example of such technology, which can use biomass as a source for both protons and (part of the) energy [3]. The required external energy input can be provided using renewable energy. In 2005, two groups simultaneously and independently reported on the bioelectrochemical mediated hydrogen production from organic waste materials [51, 52]. Both proofof-principle reactor systems made use of a platinum catalyst on the cathode to enable hydrogen production. At that time, hydrogen production in the MECs was mainly considered to be an abiotic process occurring in the absence of microorganisms. In 2007, it was shown that an enriched mixed culture of electrochemically active microorganisms (biocathode) could successfully catalyze hydrogen evolution [29]. The biocathode was first grown as a bioanode on acetate and was later “switched” to a biocathode by stepwise changing the polarity of the electrode from 0.2 to 0.8 V [29]. This biocathode produced hydrogen at a rate of 0.63 m3/m3 reactor per day at a potential of 0.7 V. Later attempts successfully demonstrated the direct startup of such a cathode after inoculation without changing the polarity [7]. Furthermore, it was also shown that addition of a carbon source in the form of acetate effectively decreased the startup time of such a biocathode compared to the use of bicarbonate [28].

3.2

Typical Potentials and Production Rates for Hydrogen Production

Thermodynamically, the production of hydrogen at a biocathode, when combined with a bioanode oxidizing acetate, requires an applied cell voltage of 0.12 V [3]. This applied voltage is significantly lower than the demand for hydrogen generation from water electrolysis, which is 1.21 V (at pH 7). The oxidation of

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates

211

organic material at the bioanode thus provides an advantage of MECs over electrolysis, as the chemical energy in organic material can provide part of the required energy input for hydrogen production in the MEC. The actual required energy input is generally higher than the thermodynamic value of 0.12 V, as a result of irreversible losses produced by the internal resistance of the system [22]. In practice, it was observed that applied voltages higher than 0.2 V are needed for substantial hydrogen production at cathode potentials more negative than 0.65 V [3]. Since 2006, many researchers have managed to boost the development of MECs from lab scale (1 L) [58]. The rate at which hydrogen can be produced has increased from 0.02 to 50 m3/m3 reactor per day [59, 60]. In MECs, the hydrogen production rate is affected by many parameters, such as biomass biodegradability, electrolyte conductivity, reactor configuration (membrane, distance between anode and cathode), external electricity input, and electrode materials (biocompatibility, specific surface area) [61]. The rate at which the hydrogen evolution reaction (HER) itself proceeds depends on the cathode overpotential, which is defined as the additional potential on top of the thermodynamic potential of the HER. This cathodic overpotential is determined by two factors: mass transfer limitations and charge transfer limitations. Mass transfer limitations are determined by the transfer of substrates and products of the reaction toward and from the reaction. Because protons play a dominant role in this reaction, it is mostly determined by the buffer concentration and especially the pKa of this buffer and the flow speed of the catholyte [62, 63]. The pKa of the buffer species determines the speciation of the specific buffer and therefore determines the concentration gradient between electrolyte bulk and the electrode surface [64, 65]. Another way to reduce mass transfer limitations is through reduction of the diffusion layer by either increased mixing in the system or the use of electrodes with higher surface to volume ratio, for example, 3D structured electrodes [60, 66]. The charge transfer from electrode to electrolyte is strongly affected by the electrode material and, if applicable, by the catalyst. Besides its ability to produce hydrogen, some other important criteria for the choice of material are cost and stability of the material under the desired potential. Table 3 gives an overview of some of the materials tested as suitable cathodes in MECs. The different electrode materials and catalysts as alternatives for the most commonly used platinum [71, 72] tested for the HER in MECs include carbon-based materials coated with palladium nanoparticles [68] or molybdenum disulfide [73], stainless steel [66, 67, 74, 75], and nickel and nickel alloys [67, 75–77]. The potential range applied for cathodes is mostly between 0.5 and 0.9 V vs NHE. Hydrogen production rates up to 50 m3/m3/day have been reported at 0.9 V vs NHE for abiotic cathodes, and biocathodes have been reported to produce 2.2 m3/m 3 /day at 0.7 V vs NHE. Coulombic efficiencies for hydrogen-producing cathodes are between 50 and 95%. Losses of hydrogen occur through diffusion through the membrane, but leakage from reactors and tubing is difficult to prevent.

212

A. ter Heijne et al.

Table 3 Overview of hydrogen production using biocathodes and abiotic cathodes with different materials Cathode material Biocathode on graphite felt Biocathode on graphite felt Carbon cloth + NiOx Carbon cloth + Ni Carbon cloth + Pd Stainless steel brush Ti + Pt Ni foam Carbon cloth + NiW Carbon cloth + NiMo Carbon cloth + MoS2

3.3

Ecat (V vs NHE) 0.7 0.7 0.8 0.5 0.55 0.71 0.87 0.91 0.5 0.5 0.9

Production rate (m3/m3/day) 0.63 2.2 0.76 0.79 2.6 (L/m2/day) 1.7 2.1 50 1.5 2.0 0.6

CE (%) 49 61 43 84 83 73 75 95

Reference [29] [28] [67] [67] [68] [66] [22] [60] [69] [69] [70]

Microorganisms for Biotic Hydrogen Production

A wide variety of microorganisms has been reported to catalyze the hydrogen evolution reaction in MECs, for example, Geobacter and Desulfovibrio spp. [78, 79]. All microbial hydrogen production involves three groups of hydrogenases which can be categorized according to their redox active metal site; Ni-Fe, Fe-Fe, and Fe-hydrogenases [80, 81]. In many cases the detected hydrogenases genes code for bidirectional Hox-type hydrogenases [82]. It has been shown that the diversity in biocathode population can vary depending on inoculum, design, and operation of the used set-up and carbon source. For an overview of the types of inoculum used in hydrogen producing MECs, see Sect. 2. In terms of carbon source, hydrogen producing bacteria may be limited if only HCO3 is supplied. It has been shown that through addition of acetate, a faster biofilm development was observed than when CO2 was used as carbon source [28]. Finally, it has been shown that a limitation in sulfur led to an almost pure culture of Promicromonospora [82].

4 Methane Production at Biocathodes 4.1

Products: Methane

The capacity of Archaea to produce methane directly from CO2 was first investigated by Cheng and co-workers, proving that a single methanogen, Methanobacterium palustre, could use electric current for direct biological reduction of CO2 into methane by a process called electromethanogenesis [8]. At first,

Mixed Culture Biocathodes for Production of Hydrogen, Methane, and Carboxylates

213

such a methane-producing biocathode was coupled to a bioanode, making the overall conversion of acetate into CO2 and methane similar to anaerobic digestion [33]. This reaction needed a small applied voltage to overcome internal resistances in the system. Later research showed that the system could be used to convert electricity into methane, so this could be a new technology for energy storage or to produce methane without the need for biomass waste streams [11, 83]. Methanogens have three metabolic pathways for methane production (Fig. 3): the CO2 reduction pathway, the methylotrophic pathway, and the acetoclastic pathway. As shown in Fig. 3, eight electrons are needed for the CO2 reduction pathway, which is four times more than both the methylotrophic pathway and the acetate pathway. Most studies on methane-producing BESs used CO2 or bicarbonate as sole carbon source. As a consequence, almost all methanogens detected in methane-producing BESs use the CO2 reduction pathway, except Methanosarcina barkeri, which can include all three pathways [84] and a few thermophilic methanogens, for example, Methanosarcina thermophila and Methanomethylovorans hollandica, which are only capable of utilizing the acetoclastic [85] and methylotrophic pathways [86], respectively.

Fig. 3 Scheme of the major substrates (in red) and the respective pathways utilized for methanogenesis (modified from [6])

214

4.2

A. ter Heijne et al.

Typical Potentials and Production Rates in Methane-Producing BESs

Many research groups have investigated the production of methane in BES, following different approaches. Most studies use a double-chamber BES separated by a membrane, and a variety of operating conditions and materials have been explored. Performance of methane-producing BESs has been reviewed in detail in Geppert et al. [83]. The applied cathode potential is one important factor that determines the performance of methane-producing BESs [18, 87, 88] because (1) it affects the mechanisms of electron transfer, and (2) cathode potential is related to the current and therefore influences the methane production rate. It was shown that the more negative the cathode potential (the higher the overpotential), the higher the current. Performance of methane-producing biocathodes is summarized in Table 4. Values for the methane production rate in the selected studies are between 0.2 and 61.7 mmol of carbon per liter of cathode volume per day (mmol C/Lcat/day) (at 298 K and 1 bar), with a tendency to increase at more negative cathode potentials. In addition to cathode potential, clearly other factors are important in determining methane production rates. The type of electron donor, system design, operational parameters, and microbial community, but also experimental time frame and the specific goal of the studies are different [83]. The highest volumetric methane production rate (60.7 mmol C/Lcat/day) has been realized within a BES containing carbon granules as cathode electrode material at a cathode potential of 0.9 V vs NHE [33], whereas lower methane production rates have been recorded in studies with BESs containing a carbon brush (0.2 mmol C/Lcat/day) [90] or a Pt-coated graphite block (0.3 mmol C/Lcat/day) [35] as cathode material, although these were operated at more positive cathode potentials. Current-to-methane efficiencies are typically between 20 and 100%, although values higher than 100% have also been reported, arising from microbial corrosion of the cathode. Efficiencies lower than 100% may be caused by, for example, side reactions as a result of oxygen diffusion to the cathode, microbial growth and maintenance, and methane losses through the membrane [83]. It has to be noted that comparisons between studies are challenging as there is no clear standardization of the experimental methodology and procedures that need to be followed to set up and carry out experiments [93]. Standardization approaches are addressed in more detail in [94] and help the field move forward to understand better and compare the results obtained in different studies. Methane can be generated via direct or mediated electron transfer, and determining the mechanisms of electron transfer is challenging. In the case of direct electron transfer, electrons might be transferred to the microorganisms via direct contact of the electrode and the microorganisms’ surface (e.g., cytochromes) [95]. During the bioelectrochemical synthesis of methane it was found that hydrogen [8, 40], acetate [91], and formate [91, 96] can act as intermediate components. The difficulty in distinguishing this indirect pathway is that the intermediates are

Water Acetate

Acetate (no membrane)d Acetate (no membrane)

0.6 0.7