- Email: [email protected]
Glucose oxidase bioanodes for glucose conversion and H2O2 production for horseradish peroxidase biocathodes in a flow through glucose biofuel cell design
Journal of Power Sources 392 (2018) 176–180
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Glucose oxidase bioanodes for glucose conversion and H2O2 production for horseradish peroxidase biocathodes in a flow through glucose biofuel cell design
T
Caroline Abreua,b,c, Yannig Nedelleca,b, Olivier Ondelc, Francois Buretc, Serge Cosniera,b, Alan Le Goffa,b, Michael Holzingera,b,∗ a b c
Université Grenoble Alpes, DCM UMR 5250, F-38000, Grenoble, France CNRS, DCM UMR 5250, F-38000, Grenoble, France Université de Lyon, CNRS UMR 5005, Laboratoire Ampère, 36 Avenue Guy de Collongue, 69134, Ecully Cedex, France
H I GH L IG H T S
G R A P H I C A L A B S T R A C T
A Glucose/H O biofuel cell is pre• sented where H O is produced at the 2
2
2
• • • •
2
bioanode. H2O2 is then conducted and consumed at the biocathode in a flow through setup. This biofuel cell has an OCV of 0.6 V and generates 0.7 mW at 0.41 V. 290μW h (1.04 J) is produced during 48 h. The design allows the connection in parallel and in series doubling the power output.
A R T I C LE I N FO
A B S T R A C T
Keywords: Biofuel cells Glucose H2O2 Glucose oxidase Horseradish peroxidase Flow stacks
Bioelectrocatalytic carbon nanotube pellets comprising glucose oxidase (GOx) at the anode and horseradish peroxidase (HRP) at the cathode were integrated in a glucose/H2O2 flow-through fuel cell setup. The porous bioelectrodes, separated with a cellulose membrane, were assembled in a design allowing the fuel/electrolyte flow through the entire fuel cell with controlled direction. An air saturated 5 mmol L−1 glucose solution was directed through the anode where glucose is used for power conversion and for the enzymatic generation of hydrogen peroxide supplying the HRP biocathode with its substrate. This configuration showed an open circuit voltage (OCV) of 0.6 V and provided 0.7 ± 0.035 mW at 0.41 V. Furthermore, different charge/discharge cycles at 500 Ω and 3 kΩ were applied to show the long term stability of this setup producing 290 μW h (1.04 J) of energy after 48 h. The biofuel cell design further allows a convenient assembly of several glucose biofuel cells in reduced volumes and its connection in parallel or in series.
1. Introduction Enzymatic glucose biofuel cells became famous for the conversion of energy to power electronic devices in a living body due to the high
∗
specificity of bocatalysts [1]. GOx is one of the most widely used enzymes for the anodic glucose oxidation of glucose-O2 biofuel cells [2] due to its stable activity and life time over several pH values [3]. GOx consists of two identical subunits (homodimer) wherein the organic
Corresponding author. Université Grenoble Alpes, DCM UMR 5250, F-38000, Grenoble, France. E-mail address: [email protected] (M. Holzinger).
https://doi.org/10.1016/j.jpowsour.2018.04.104 Received 11 January 2018; Received in revised form 26 April 2018; Accepted 28 April 2018 0378-7753/ © 2018 Elsevier B.V. All rights reserved.
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
redox cofactor flavin adenine dinucleotide (FAD) is protected by a thick glycosylated protein shell [4,5]. The GOx catalyzes with high selectivity the oxidation of β-D-glucose to gluconolactone by a 2 e− and 2 H+ process and usually regenerates itself by reducing oxygen to hydrogen peroxide [6] which can typically serve as electrochemical probe for glucose biosensor applications [7]. However, this byproduct is generally considered as disadvantage in biofuel cell applications due to its toxicity and denaturation of proteins [8]. Furthermore, the electrons used for oxygen reduction are not transferred to the electrodes and thus, do not contribute to the power conversion. An efficient way to improve the performances of GOx bioanodes and to reduce the production of H2O2 is to use redox active molecules with appropriate potentials which play the role of electron shuttles from GOx to the electrode material [9]. The Adam Heller group realized pioneering work in tunable osmium hydrogels for efficient mediated electron transfer using GOx as biocatalyst [10–13] with steadily improved glucose biofuels cell performances [14–16]. Beside metal organic complexes, quinones became a prominent alternative due to improved current stabilities and life-time of the bioanodes [17–19]. An elegant approach to clearly reduce the side reaction producing H2O2 is to confine the enzyme in a redox hydrogel where the surface layer consumes oxygen thus providing for the inner layers an anaerobic environment with improved electron transfer rates [20,21]. However, even when very efficient wiring strategies could be developed, the production of H2O2 still remains an issue with GOx as anodic biocatalyst. The introduction of the enzyme catalase which consumes the produced H2O2 forming oxygen and water [22] can further reduce this issue but not totally eliminate it. Nonetheless, an original approach was demonstrated taking advantage of this H2O2 production by GOx for the development of biocathodes using a bi-enzymatic system with HRP as catalysts for the reduction of hydrogen peroxide [23–25]. In this setup, GOx reduces oxygen to H2O2 which is then further reduced to water by HRP. This small enzyme has an easily accessible heme function enabling facilitated direct electron transfer [26]. The advantage of such biocathodes is the lack of inhibition effects by chloride or urate present in body fluids which is a constant drawback for multi copper enzymes like laccase or bilirubin oxidase (BOD) which are often used as cathodic biocatalysts [27–30]. The disadvantage of such biocathodes is the need to combine two enzymes thus increasing the amount of protein to immobilize while only one will be connected. Another disadvantage is the possible depletion of the glucose concentration for the operation of the bioanode. Furthermore, the issue of contaminating H2O2 production at the bioanode is still not resolved. One example of a glucose/H2O2 biofuel cell using GOx as anodic and HRP as cathodic biocatalyst describes a two compartment setup with defined glucose and H2O2 concentrations in the respective compartment to reach 4.7 μWcm−2 [31] without profiting from the natural production of hydrogen peroxide. Here, we present a new approach that benefit from this H2O2 production of remaining unwired GOx at the bioanode. A mixture of GOx, naphthoquinone, and multi-walled carbon nanotubes (MWCNTs) were compressed in a Plexiglas housing to form the bioanode. The biocathode was made in the same way by using a mixture of HRP and MWCNTs. The bioelectrodes were integrated in a recently designed and adapted flow through setup [32,33]. Hydrogen peroxide is here conducted from the anode to a HRP base cathode supplying this enzyme with its natural substrate which is entirely consumed. This setup reduced diffusion issues of oxygen to the cathode and provides very satisfying power outputs and lifetimes.
Saint-Quentin-Fallavier, France. MWCNTs were purchased from Nanocyl, Sambreville, Belgium (> 95% purity, 10 nm diameter, 1.5 μm length). It has to be noted that all experiments were conducted with the same batch of enzymes and MWCNTs to assure the reproducibility of our results and to avoid possible performance fluctuations related to batch-to-batch variations. The microporous gas diffusion layer (GDL) of 210 μm thick, 8 mΩ cm2 of through plane electrical resistance and through plane air permeability of 70 s, was purchased from Paxitech, Échirolles, France (FI2C6). Graphoil® was provided by Panasonic, Kadoma, Japan, with 0.07 mm thickness and 1 kW m−1 K−1 thermal conductivity. The 100% cellulose membrane with 0.13 mm thickness and 64 g m−2 was purchased from FiltraTech, Saran, france. Enzymes were stored at −20 °C. Distilled water was obtained by water purification to a resistivity of 15 MΩ cm using a Millipore, Burlington, Massachusetts, USA, Ultrapure system. Glucose solutions were left to mutarotate overnight to β-D-glucose prior to use. For the electrochemical characterization of the biofuel cell, the anode was set as the working electrode while the cathode was plugged as the counter-reference electrode. All experiments were conducted in a glucose solution of 5 mmol L−1 in McIlvaine, pH 7 unless otherwise mentioned. The biofuel cell was connected to a multichannel potentiostat Biologic®, Seyssinet-Pariset, France, VMP3 running EC-lab software 10.39. Polarization and power curves were recorded after 30-s discharge. All the experiments were done at room temperature. 2.1. Preparation of the bioelectrodes The MWCNT pellets were obtained by soft grinding of an optimized mixture of 100 μL of distilled water, 5 mg of 1,4-naphthoquinone, 15 mg of GOx from Aspergillus Niger (174 U mg−1) and 35 mg of MWCNTs as described in detail in Ref. [34]. Here, catalase which disproportionates H2O2 was not included providing optimal substrate production for the biocathode. The same procedure was employed for the biocathode using 15 mg of HRP from Horseradish (193 U mg−1) and 35 mg of MWCNTs. Each of the obtained homogenous pastes is then compressed directly in a 3 mm thick Plexiglas housing with a hole of 1.3 cm diameter using a hydraulic press [33]. A graphoil disc is connected to one side of the pellet to permit the electrical connection. 2.2. Design of the biofuel cells The biofuel cells were designed according a specific configuration depicted in Fig. 1 to allow the easy assembly of multiple biofuel cells using easy-to-handle in-house fabricated Plexiglas elements. A 3 mm thick Plexiglas slice was designed with a hole in the center of 13 mm in diameter. A microporous GDL was then placed at the backside of the Plexiglas to serve as the electron collector and assures simultaneously the diffusion of the glucose-oxygen solution. After compressing the MWCNT-enzyme pellet in the 13 mm hole of the Plexiglas slice with the GDL on the backside. The open side of the pellet was then covered with a cellulose sheet which served as electrolyte reservoir. A polytetrafluoroethylene (PTFE) sheet of 100 μm thickness with a 16 mm hole was inserted in between the bioanode and the biocathode as a separator and to prevent any leakage of the electrolyte. At the GDL backside of each bioelectrode, a graphoil sheet is placed to allow the electrical connection to an external circuit. To provide a constant glucose flow within the cells, a peristaltic pump or standard glucose perfusion bag, such as those found in hospitals, were used. The 0.2 mL min−1 flow was oriented from the anode to the cathode. 3. Results and discussion
2. Material and methods
3.1. Design of the GBFC
All reagents, glucose oxidase (GOx) from Aspergillus Niger and peroxidase from Horseradish (HRP) enzymes were purchased from Aldrich,
Fig. 1 presents the glucose/H2O2 fuel cell and its device. The flow diffuses from the anode to the cathode to drive the produced hydrogen 177
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
Fig. 1. Schematic presentation of the flow through biofuel cell. At the anode, wired GOx consumes glucose and the involved electrons are transferred to the electrode by the mediator naphthoquinone (NQ). Unwired GOx at the anode consumes glucose and reduces oxygen to hydrogen peroxide. H2O2 is then directed to the cathode and reduced to water by HRP using the electrons directly transferred from the MWCNTs.
With this single cell flow-through configuration, the GBFC showed an open circuit voltage (OCV) of 0.6 V and provided 0.7 ± 0.035 mW at 0.414 V. Under continuous discharge at 6 kΩ (Fig. 3B), the cell voltage rapidly decreased until 75 mV accumulating around 30 μW h (0.18 J) of energy during the first hour and further 20 μW h for the next 23 h. This relative short stability of the performance is clearly attributed to the biocathode since other setups using the same bioanode showed much higher stabilities under these discharge conditions [33]. The reason for this rapid performance decrease might arise from mass transport limitations of HRP at the cathode along the course of the continuous discharge. This limitation can be counterbalanced by operating the fuel cell under discharge/charge cycle mode within a potential range in between 0.5 and 0.4 V at different resistances (Fig. 3C black and red). Under these conditions our setup is perfectly stable during at least 48 h. At 3 kΩ, 290 μW h (1.04 J) was produced during 48 h while at 500 Ω 200 μW h (0.72 J) of energy could be generated. The lower energy yield at lower resistance can be explained by the shorter discharge/charge cycles. At 500 Ω (Fig. 3D, black and red), one cycle takes always 45 min during 48 h. At 3 kΩ (Fig. 3D, blue and green), one cycle takes 65 min at the beginning of the experiment and increased to 90 min for one discharge/charge cycle after 48 h. At 500 Ω, the discharge process is very fast and takes just 2 min which can explain the reduced energy accumulation of 4.1 μW h. At 3 kΩ, the discharge process takes 5 min and during this time, 7.8 μW h of energy is produced. In comparison with the continuous discharge experiment, we assume that during the recharge period at open circuit, the produced electrons are not evacuated at the anode and negative charges are accumulated until the initial potential is reached. Related to this, naphtoquinone is not regenerated and remains in the reduced state. During this time, GOx continues to oxidize glucose and since here is no competitive reaction with the mediator, the regeneration of this enzyme by oxygen is favored and more H2O2 is produced. The more convenient peroxide concentration improves the overall performance of the biocathode.
Fig. 2. Chronoamperometric measurements of the catalytic current of the glucose/H2O2 biofuel cell at imposed cell voltage of 0.55 V and at different glucose concentrations in Mc Ilvaine buffer, pH 7.
peroxide to the HRP which can react together to produce water. 3.2. Dependence of the H2O2 concentration in the system Our glucose fuel cell setup was initially evaluated at different glucose concentrations using chronoamperometry at 0.55 V. As shown in Fig. 2, low currents were measured at low glucose concentrations from 1 mmolL−1 and steadily increased until 20 mmol L−1 where 200 μA was reached. Upon further injection of glucose to reach a final concentration of 50 mmol L−1, the current continued to increase for a short time until it clearly drops. This evolution is most likely due to an overproduction of hydrogen peroxide altering the catalytic activity of HRP [35,36]. One of the most promising applications of glucose biofuel cells is the power generation of implantable medical devices out of body fluids where the glucose concentration of healthy persons is in average 5 mmol L−1 with fluctuation between 4.4 and 6.1 mmol L−1 during a day [37] and the oxygen concentration is around 0.15 mmol L−1 (3 mL/ L) [38]. We thus operated our biofuel cell at 5 mmol L−1 of glucose and air saturated solutions (∼280 mmol L−1 ≘ 6.3mL/L) [39].
3.4. Assembly and stacking of two GBFCs Separated inserts were used to supply individually two GBFCs with the air saturated glucose solution while the flow direction was still directed from the anode to the cathode. The same physiologically relevant 5 mmol L−1 glucose solution was used to supply both GBFCs and was simply separated by T-pieces. Fig. 4A shows the polarization and power curves for the system with in series connection. The power curve was plotted from the polarization curve, recorded by successive 30 s discharges at constant current (0.1 mA–2.8 mA). The performances were monitored at pH 7 in presence of a 5 mmol L−1 air saturated glucose solution flow. With series connection, as expected, the OCV is twice that recorded for one biofuel cell (1.23 V) while the maximum power output and current are
3.3. Performance and stability of one GBFC Fig. 3Ashows the power profile and the polarization curve for 5 mmol L−1 glucose where each point was recorded after 30 s discharge to eliminate the capacitive current. 178
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
Fig. 3. (A) Power profile (blue) and polarization curve (black) for the GBFC at 5 mmol L−1 of glucose in Mc Ilvaine pH 7. Each point represents the value obtained after 30 s discharge between 0.1 and 2.3 mA. (B) Continuous discharge at 6 kΩ (black) and produced energy (blue). (C) Discharge/charge cycles between 0.5 and 0.4 V at 500 Ω (a, black) and produced energy (b, blue), at 3 kΩ (c, red) and produced energy (d, green). (D) Zoom of one cycle chargedischarge at 500 Ω (black) and 3 kΩ (blue) in the beginning (a, black and c, blue respectively) and after 48 h (b, red d, green, respectively). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
1.74 ± 0.086 mW at 0.723 V and 2.4 mA, respectively. While the current remains almost the same, the generated power and the operational potential are even slightly more than two times higher than one
individual GBFC but remain within the standard error of the biofuel cell performances. The stacked glucose biofuel cell connected in parallel (Fig. 4B) exhibits a maximum OCV of 0.757 V and a maximum power
Fig. 4. (A) Scheme (top), power profile (blue) and polarization curve (black) for two GBFCs in series. Each point represents the value obtained after 30 s discharge between 0.1 and 2.8 mA. (B) Scheme (top), power profile (blue) and polarization curve (black) for two GBFCs in parallel. Each point represents the value obtained after 30 s discharge between 0.1 and 5.2 mA. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
179
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
output of 2.24 ± 0.112 mW (at 0.458 V and 4.9 mA). Again, improved performances of these two connected GBFCs can be observed compared to an individual glucose biofuel cell but remains at a reasonable scale. For both connections, a relative standard deviation of around 5% was obtained using the same batches of carbon nanotubes and enzymes. It can thus be concluded that this system is compatible for efficient stacking of many GBFCs and allows maintaining the high performances.
[10] A. Heller, Electrical connection of enzyme redox centers to electrodes, J. Phys. Chem. 96 (9) (1992) 3579–3587. [11] Y. Degani, A. Heller, Electrical communication between redox centers of glucose oxidase and electrodes via electrostatically and covalently bound redox polymers, J. Am. Chem. Soc. 111 (6) (1989) 2357–2358. [12] B.A. Gregg, A. Heller, Cross-linked redox gels containing glucose oxidase for amperometric biosensor applications, Anal. Chem. 62 (3) (1990) 258–263. [13] W. Schuhmann, T.J. Ohara, H.L. Schmidt, A. Heller, Electron transfer between glucose oxidase and electrodes via redox mediators bound with flexible chains to the enzyme surface, J. Am. Chem. Soc. 113 (4) (1991) 1394–1397. [14] N. Mano, F. Mao, A. Heller, A miniature biofuel cell Operating in A Physiological buffer, J. Am. Chem. Soc. 124 (44) (2002) 12962–12963. [15] F. Barrière, P. Kavanagh, D. Leech, A laccase–glucose oxidase biofuel cell prototype operating in a physiological buffer, Electrochim. Acta. 51 (24) (2006) 5187–5192. [16] F. Gao, L. Viry, M. Maugey, P. Poulin, N. Mano, Engineering hybrid nanotube wires for high-power biofuel cells, Nat. Commun. 1 (1) (2010) 2. [17] B. Reuillard, C. Abreu, N. Lalaoui, A. Le Goff, M. Holzinger, O. Ondel, et al., Oneyear stability for a glucose/oxygen biofuel cell combined with pH reactivation of the laccase/carbon nanotube biocathode, Bioelectrochemistry 106 (Part A) (2015) 73–76. [18] R.D. Milton, D.P. Hickey, S. Abdellaoui, K. Lim, F. Wu, B. Tan, et al., Rational design of quinones for high power density biofuel cells, Chem. Sci. 6 (8) (2015) 4867–4875. [19] F. Giroud, C. Gondran, K. Gorgy, V. Vivier, S. Cosnier, An enzymatic biofuel cell based on electrically wired polyphenol oxidase and glucose oxidase operating under physiological conditions, Electrochim. Acta. 85 (0) (2012) 278–282. [20] M. Grattieri, M. Tucci, M. Bestetti, S. Trasatti, P. Cristiani, Facilitated electron hopping in nanolayer oxygen-insensitive glucose biosensor for application in a complex matrix, ChemElectroChem 3 (11) (2016) 1884–1889. [21] A. Prévoteau, N. Mano, How the reduction of O2 on enzymes and/or redox mediators affects the calibration curve of “wired” glucose oxidase and glucose dehydrogenase biosensors, Electrochim. Acta. 112 (2013) 318–326. [22] M. Alfonso-Prieto, X. Biarnés, P. Vidossich, C. Rovira, The molecular mechanism of the catalase reaction, J. Am. Chem. Soc. 131 (33) (2009) 11751–11761. [23] W. Jia, C. Jin, W. Xia, M. Muhler, W. Schuhmann, L. Stoica, Glucose oxidase/ horseradish peroxidase Co-immobilized at a CNT-modified graphite electrode: towards potentially implantable biocathodes, Chem. Eur J. 18 (10) (2012) 2783–2786. [24] C. Agnès, B. Reuillard, A. Le Goff, M. Holzinger, S. Cosnier, A double-walled carbon nanotube-based glucose/H2O2 biofuel cell operating under physiological conditions, Electrochem. Commun. 34 (0) (2013) 105–108. [25] K. Elouarzaki, M. Bourourou, M. Holzinger, A. Le Goff, R. Marks, S. Cosnier, Freestanding HRP-GOx redox buckypaper as oxygen-reducing biocathode for biofuel cell applications, Energy Environ. Sci. 8 (7) (2015) 2069–2074. [26] L. Gorton, A. Lindgren, T. Larsson, F.D. Munteanu, T. Ruzgas, I. Gazaryan, Direct electron transfer between heme-containing enzymes and electrodes as basis for third generation biosensors, Anal. Chim. Acta. 400 (1–3) (1999) 91–108. [27] S.C. Barton, M. Pickard, R. Vazquez-Duhalt, A. Heller, Electroreduction of O2 to water at 0.6 V (SHE) at pH 7 on the ‘wired’ Pleurotus ostreatus laccase cathode, Biosens. Bioelectron. 17 (11–12) (2002) 1071–1074. [28] B. Reuillard, A. Le Goff, M. Holzinger, S. Cosnier, Non-covalent functionalization of carbon nanotubes with boronic acids for the wiring of glycosylated redox enzymes in oxygen-reducing biocathodes, J. Mater. Chem. B. 2 (16) (2014) 2228–2232. [29] E. Katz, K. MacVittie, Implanted biofuel cells operating in vivo - methods, applications and perspectives - feature article, Energy Environ. Sci. 6 (10) (2013) 2791–2803. [30] A. Zebda, S. Cosnier, J.-P. Alcaraz, M. Holzinger, A. Le Goff, C. Gondran, et al., Single glucose biofuel cells implanted in rats power electronic devices, Sci. Rep. 3 (2013) 1516. [31] A. Ramanavicius, A. Kausaite-Minkstimiene, I. Morkvenaite-Vilkonciene, P. Genys, R. Mikhailova, T. Semashko, et al., Biofuel cell based on glucose oxidase from Penicillium funiculosum 46.1 and horseradish peroxidase, Chem. Eng. J. 264 (2015) 165–173. [32] A. Le Goff, Y. Nedellec, C. Abreu, S. Cosnier, M. Holzinger, inventors; Pile à Biocombustible. France, (2017). [33] C. Abreu, Y. Nedellec, A.J. Gross, O. Ondel, F. Buret, A.L. Goff, et al., Assembly and stacking of flow-through enzymatic bioelectrodes for high power glucose fuel cells, ACS Appl. Mater. Interfaces 9 (28) (2017) 23836–23842. [34] B. Reuillard, A. Le Goff, C. Agnes, M. Holzinger, A. Zebda, C. Gondran, et al., High power enzymatic biofuel cell based on naphthoquinone-mediated oxidation of glucose by glucose oxidase in a carbon nanotube 3D matrix, Phys. Chem. Chem. Phys. 15 (14) (2013) 4892–4896. [35] S. Asad, S.-F. Torabi, M. Fathi-Roudsari, N. Ghaemi, K. Khajeh, Phosphate buffer effects on thermal stability and H2O2-resistance of horseradish peroxidase, Int. J. Biol. Macromol. 48 (4) (2011) 566–570. [36] B. Limoges, J.-M. Savéant, D. Yazidi, Quantitative analysis of catalysis and inhibition at horseradish peroxidase monolayers immobilized on an electrode surface, J. Am. Chem. Soc. 125 (30) (2003) 9192–9203. [37] Screening for type 2 diabetes, Diabetes Care 26 (suppl. 1) (2003) s21–s24. [38] R.N. Pittman, San rafael (CA): morgan & claypool life sciences; 2011. Chapter 4. Oxygen transport, in: R.N. Pittman (Ed.), Regulation of Tissue Oxygenation. San Rafael (CA): Morgan & Claypool Life Sciences, 2011. [39] J.H. Carpenter, New measurements of oxygen solubility in pure and natural water, Limnol. Oceanogr. 11 (2) (1966) 264–277.
4. Conclusion A flow through design of a glucose biofuel cell allowing the stacking of two GBFCs and its connection in parallel and in series was used to benefit from the hydrogen peroxide production at the GOx-based bioanode to supply the cathodic biocatalyst HRP with its natural substrate. Each cell was supplied by a glucose flow which is directed from the anode through a cellulose membrane to the cathode. By this way, the biocathode is not dependent on the oxygen concentration in solution since the substrates glucose and oxygen are converted at the bioanode. Furthermore, the produced peroxide is related with the glucose concentration and leads to equilibrated performances of the bioelectrodes. This biofuel cell setup is operational until a glucose concentration of 20 mmolL−1 but we decided to perform the experiments at 5 mmolL−1 relating to the average glucose concentration in body fluids of mammals. Thus, this design is promising for the power supply of implantable medical devices because all products are biocompatible (water and gluconolactone). Two stacked biofuel cells provided a good power output when connected either in series or in parallel. This setup further allows the association of an almost unlimited number GBFC in series and parallel which should make it possible to reach the necessary voltage and current outputs to run devices without the need of supplemental step-up converters or capacitors. However, further efforts are necessary in terms of operational stability before such stacked biofuel cells can be envisioned for the power supply of implantable devices even when stabilities over few days in charge/discharge mode are quite encouraging. Acknowledgment The region Auvergne-Rhône-Alpes is acknowledged for the PhD funding of C. Abreu. The authors would also like to thank the mechanical team of LiPhy at Grenoble for the fabrication of the PTFE and Plexiglass plates. The authors wish further to acknowledge the support from the platform Chimie NanoBio ICMG FR 2607 (PCN-ICMG), from the Institute Carnot PolyNat at Grenoble, and from the LabEx ARCANE (ANR-11-LABX-0003-01). References [1] S. Cosnier, A. Le Goff, M. Holzinger, Enzymatic fuel cells: from design to implantation in mammals, in: E. Katz (Ed.), Implantable Bioelectronics, Wiley-VCH Verlag GmbH & Co. KGaA, 2014, pp. 347–362. [2] M. Rasmussen, S. Abdellaoui, S.D. Minteer, Enzymatic biofuel cells: 30 years of critical advancements, Biosens. Bioelectron. 76 (2016) 91–102. [3] R. Wilson, A.P.F. Turner, Glucose oxidase: an ideal enzyme, Biosens. Bioelectron. 7 (3) (1992) 165–185. [4] G. Wohlfahrt, S. Witt, J. Hendle, D. Schomburg, H.M. Kalisz, H.-J. Hecht, 1.8 and 1.9 A resolution structures of the Penicillium amagasakiense and Aspergillus Niger glucose oxidases as a basis for modelling substrate complexes, Acta Crystallogr. D. 55 (5) (1999) 969–977. [5] O. Courjean, F. Gao, N. Mano, Deglycosylation of glucose oxidase for direct and efficient glucose electrooxidation on a glassy carbon electrode, Angew. Chem. Int. Ed. 48 (32) (2009) 5897–5899. [6] S.B. Bankar, M.V. Bule, R.S. Singhal, L. Ananthanarayan, Glucose oxidase — an overview, Biotechnol. Adv. 27 (4) (2009) 489–501. [7] J. Wang, Electrochemical glucose biosensors, Chem. Rev. 108 (2) (2008) 814–825. [8] B.E. Watt, A.T. Proudfoot, J.A. Vale, Hydrogen peroxide poisoning, Toxicol. Rev. 23 (1) (2004) 51–57. [9] P. Kavanagh, D. Leech, Mediated electron transfer in glucose oxidising enzyme electrodes for application to biofuel cells: recent progress and perspectives, Phys. Chem. Chem. Phys. 15 (14) (2013) 4859–4869.
180
Contents lists available at ScienceDirect
Journal of Power Sources journal homepage: www.elsevier.com/locate/jpowsour
Short communication
Glucose oxidase bioanodes for glucose conversion and H2O2 production for horseradish peroxidase biocathodes in a flow through glucose biofuel cell design
T
Caroline Abreua,b,c, Yannig Nedelleca,b, Olivier Ondelc, Francois Buretc, Serge Cosniera,b, Alan Le Goffa,b, Michael Holzingera,b,∗ a b c
Université Grenoble Alpes, DCM UMR 5250, F-38000, Grenoble, France CNRS, DCM UMR 5250, F-38000, Grenoble, France Université de Lyon, CNRS UMR 5005, Laboratoire Ampère, 36 Avenue Guy de Collongue, 69134, Ecully Cedex, France
H I GH L IG H T S
G R A P H I C A L A B S T R A C T
A Glucose/H O biofuel cell is pre• sented where H O is produced at the 2
2
2
• • • •
2
bioanode. H2O2 is then conducted and consumed at the biocathode in a flow through setup. This biofuel cell has an OCV of 0.6 V and generates 0.7 mW at 0.41 V. 290μW h (1.04 J) is produced during 48 h. The design allows the connection in parallel and in series doubling the power output.
A R T I C LE I N FO
A B S T R A C T
Keywords: Biofuel cells Glucose H2O2 Glucose oxidase Horseradish peroxidase Flow stacks
Bioelectrocatalytic carbon nanotube pellets comprising glucose oxidase (GOx) at the anode and horseradish peroxidase (HRP) at the cathode were integrated in a glucose/H2O2 flow-through fuel cell setup. The porous bioelectrodes, separated with a cellulose membrane, were assembled in a design allowing the fuel/electrolyte flow through the entire fuel cell with controlled direction. An air saturated 5 mmol L−1 glucose solution was directed through the anode where glucose is used for power conversion and for the enzymatic generation of hydrogen peroxide supplying the HRP biocathode with its substrate. This configuration showed an open circuit voltage (OCV) of 0.6 V and provided 0.7 ± 0.035 mW at 0.41 V. Furthermore, different charge/discharge cycles at 500 Ω and 3 kΩ were applied to show the long term stability of this setup producing 290 μW h (1.04 J) of energy after 48 h. The biofuel cell design further allows a convenient assembly of several glucose biofuel cells in reduced volumes and its connection in parallel or in series.
1. Introduction Enzymatic glucose biofuel cells became famous for the conversion of energy to power electronic devices in a living body due to the high
∗
specificity of bocatalysts [1]. GOx is one of the most widely used enzymes for the anodic glucose oxidation of glucose-O2 biofuel cells [2] due to its stable activity and life time over several pH values [3]. GOx consists of two identical subunits (homodimer) wherein the organic
Corresponding author. Université Grenoble Alpes, DCM UMR 5250, F-38000, Grenoble, France. E-mail address: [email protected] (M. Holzinger).
https://doi.org/10.1016/j.jpowsour.2018.04.104 Received 11 January 2018; Received in revised form 26 April 2018; Accepted 28 April 2018 0378-7753/ © 2018 Elsevier B.V. All rights reserved.
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
redox cofactor flavin adenine dinucleotide (FAD) is protected by a thick glycosylated protein shell [4,5]. The GOx catalyzes with high selectivity the oxidation of β-D-glucose to gluconolactone by a 2 e− and 2 H+ process and usually regenerates itself by reducing oxygen to hydrogen peroxide [6] which can typically serve as electrochemical probe for glucose biosensor applications [7]. However, this byproduct is generally considered as disadvantage in biofuel cell applications due to its toxicity and denaturation of proteins [8]. Furthermore, the electrons used for oxygen reduction are not transferred to the electrodes and thus, do not contribute to the power conversion. An efficient way to improve the performances of GOx bioanodes and to reduce the production of H2O2 is to use redox active molecules with appropriate potentials which play the role of electron shuttles from GOx to the electrode material [9]. The Adam Heller group realized pioneering work in tunable osmium hydrogels for efficient mediated electron transfer using GOx as biocatalyst [10–13] with steadily improved glucose biofuels cell performances [14–16]. Beside metal organic complexes, quinones became a prominent alternative due to improved current stabilities and life-time of the bioanodes [17–19]. An elegant approach to clearly reduce the side reaction producing H2O2 is to confine the enzyme in a redox hydrogel where the surface layer consumes oxygen thus providing for the inner layers an anaerobic environment with improved electron transfer rates [20,21]. However, even when very efficient wiring strategies could be developed, the production of H2O2 still remains an issue with GOx as anodic biocatalyst. The introduction of the enzyme catalase which consumes the produced H2O2 forming oxygen and water [22] can further reduce this issue but not totally eliminate it. Nonetheless, an original approach was demonstrated taking advantage of this H2O2 production by GOx for the development of biocathodes using a bi-enzymatic system with HRP as catalysts for the reduction of hydrogen peroxide [23–25]. In this setup, GOx reduces oxygen to H2O2 which is then further reduced to water by HRP. This small enzyme has an easily accessible heme function enabling facilitated direct electron transfer [26]. The advantage of such biocathodes is the lack of inhibition effects by chloride or urate present in body fluids which is a constant drawback for multi copper enzymes like laccase or bilirubin oxidase (BOD) which are often used as cathodic biocatalysts [27–30]. The disadvantage of such biocathodes is the need to combine two enzymes thus increasing the amount of protein to immobilize while only one will be connected. Another disadvantage is the possible depletion of the glucose concentration for the operation of the bioanode. Furthermore, the issue of contaminating H2O2 production at the bioanode is still not resolved. One example of a glucose/H2O2 biofuel cell using GOx as anodic and HRP as cathodic biocatalyst describes a two compartment setup with defined glucose and H2O2 concentrations in the respective compartment to reach 4.7 μWcm−2 [31] without profiting from the natural production of hydrogen peroxide. Here, we present a new approach that benefit from this H2O2 production of remaining unwired GOx at the bioanode. A mixture of GOx, naphthoquinone, and multi-walled carbon nanotubes (MWCNTs) were compressed in a Plexiglas housing to form the bioanode. The biocathode was made in the same way by using a mixture of HRP and MWCNTs. The bioelectrodes were integrated in a recently designed and adapted flow through setup [32,33]. Hydrogen peroxide is here conducted from the anode to a HRP base cathode supplying this enzyme with its natural substrate which is entirely consumed. This setup reduced diffusion issues of oxygen to the cathode and provides very satisfying power outputs and lifetimes.
Saint-Quentin-Fallavier, France. MWCNTs were purchased from Nanocyl, Sambreville, Belgium (> 95% purity, 10 nm diameter, 1.5 μm length). It has to be noted that all experiments were conducted with the same batch of enzymes and MWCNTs to assure the reproducibility of our results and to avoid possible performance fluctuations related to batch-to-batch variations. The microporous gas diffusion layer (GDL) of 210 μm thick, 8 mΩ cm2 of through plane electrical resistance and through plane air permeability of 70 s, was purchased from Paxitech, Échirolles, France (FI2C6). Graphoil® was provided by Panasonic, Kadoma, Japan, with 0.07 mm thickness and 1 kW m−1 K−1 thermal conductivity. The 100% cellulose membrane with 0.13 mm thickness and 64 g m−2 was purchased from FiltraTech, Saran, france. Enzymes were stored at −20 °C. Distilled water was obtained by water purification to a resistivity of 15 MΩ cm using a Millipore, Burlington, Massachusetts, USA, Ultrapure system. Glucose solutions were left to mutarotate overnight to β-D-glucose prior to use. For the electrochemical characterization of the biofuel cell, the anode was set as the working electrode while the cathode was plugged as the counter-reference electrode. All experiments were conducted in a glucose solution of 5 mmol L−1 in McIlvaine, pH 7 unless otherwise mentioned. The biofuel cell was connected to a multichannel potentiostat Biologic®, Seyssinet-Pariset, France, VMP3 running EC-lab software 10.39. Polarization and power curves were recorded after 30-s discharge. All the experiments were done at room temperature. 2.1. Preparation of the bioelectrodes The MWCNT pellets were obtained by soft grinding of an optimized mixture of 100 μL of distilled water, 5 mg of 1,4-naphthoquinone, 15 mg of GOx from Aspergillus Niger (174 U mg−1) and 35 mg of MWCNTs as described in detail in Ref. [34]. Here, catalase which disproportionates H2O2 was not included providing optimal substrate production for the biocathode. The same procedure was employed for the biocathode using 15 mg of HRP from Horseradish (193 U mg−1) and 35 mg of MWCNTs. Each of the obtained homogenous pastes is then compressed directly in a 3 mm thick Plexiglas housing with a hole of 1.3 cm diameter using a hydraulic press [33]. A graphoil disc is connected to one side of the pellet to permit the electrical connection. 2.2. Design of the biofuel cells The biofuel cells were designed according a specific configuration depicted in Fig. 1 to allow the easy assembly of multiple biofuel cells using easy-to-handle in-house fabricated Plexiglas elements. A 3 mm thick Plexiglas slice was designed with a hole in the center of 13 mm in diameter. A microporous GDL was then placed at the backside of the Plexiglas to serve as the electron collector and assures simultaneously the diffusion of the glucose-oxygen solution. After compressing the MWCNT-enzyme pellet in the 13 mm hole of the Plexiglas slice with the GDL on the backside. The open side of the pellet was then covered with a cellulose sheet which served as electrolyte reservoir. A polytetrafluoroethylene (PTFE) sheet of 100 μm thickness with a 16 mm hole was inserted in between the bioanode and the biocathode as a separator and to prevent any leakage of the electrolyte. At the GDL backside of each bioelectrode, a graphoil sheet is placed to allow the electrical connection to an external circuit. To provide a constant glucose flow within the cells, a peristaltic pump or standard glucose perfusion bag, such as those found in hospitals, were used. The 0.2 mL min−1 flow was oriented from the anode to the cathode. 3. Results and discussion
2. Material and methods
3.1. Design of the GBFC
All reagents, glucose oxidase (GOx) from Aspergillus Niger and peroxidase from Horseradish (HRP) enzymes were purchased from Aldrich,
Fig. 1 presents the glucose/H2O2 fuel cell and its device. The flow diffuses from the anode to the cathode to drive the produced hydrogen 177
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
Fig. 1. Schematic presentation of the flow through biofuel cell. At the anode, wired GOx consumes glucose and the involved electrons are transferred to the electrode by the mediator naphthoquinone (NQ). Unwired GOx at the anode consumes glucose and reduces oxygen to hydrogen peroxide. H2O2 is then directed to the cathode and reduced to water by HRP using the electrons directly transferred from the MWCNTs.
With this single cell flow-through configuration, the GBFC showed an open circuit voltage (OCV) of 0.6 V and provided 0.7 ± 0.035 mW at 0.414 V. Under continuous discharge at 6 kΩ (Fig. 3B), the cell voltage rapidly decreased until 75 mV accumulating around 30 μW h (0.18 J) of energy during the first hour and further 20 μW h for the next 23 h. This relative short stability of the performance is clearly attributed to the biocathode since other setups using the same bioanode showed much higher stabilities under these discharge conditions [33]. The reason for this rapid performance decrease might arise from mass transport limitations of HRP at the cathode along the course of the continuous discharge. This limitation can be counterbalanced by operating the fuel cell under discharge/charge cycle mode within a potential range in between 0.5 and 0.4 V at different resistances (Fig. 3C black and red). Under these conditions our setup is perfectly stable during at least 48 h. At 3 kΩ, 290 μW h (1.04 J) was produced during 48 h while at 500 Ω 200 μW h (0.72 J) of energy could be generated. The lower energy yield at lower resistance can be explained by the shorter discharge/charge cycles. At 500 Ω (Fig. 3D, black and red), one cycle takes always 45 min during 48 h. At 3 kΩ (Fig. 3D, blue and green), one cycle takes 65 min at the beginning of the experiment and increased to 90 min for one discharge/charge cycle after 48 h. At 500 Ω, the discharge process is very fast and takes just 2 min which can explain the reduced energy accumulation of 4.1 μW h. At 3 kΩ, the discharge process takes 5 min and during this time, 7.8 μW h of energy is produced. In comparison with the continuous discharge experiment, we assume that during the recharge period at open circuit, the produced electrons are not evacuated at the anode and negative charges are accumulated until the initial potential is reached. Related to this, naphtoquinone is not regenerated and remains in the reduced state. During this time, GOx continues to oxidize glucose and since here is no competitive reaction with the mediator, the regeneration of this enzyme by oxygen is favored and more H2O2 is produced. The more convenient peroxide concentration improves the overall performance of the biocathode.
Fig. 2. Chronoamperometric measurements of the catalytic current of the glucose/H2O2 biofuel cell at imposed cell voltage of 0.55 V and at different glucose concentrations in Mc Ilvaine buffer, pH 7.
peroxide to the HRP which can react together to produce water. 3.2. Dependence of the H2O2 concentration in the system Our glucose fuel cell setup was initially evaluated at different glucose concentrations using chronoamperometry at 0.55 V. As shown in Fig. 2, low currents were measured at low glucose concentrations from 1 mmolL−1 and steadily increased until 20 mmol L−1 where 200 μA was reached. Upon further injection of glucose to reach a final concentration of 50 mmol L−1, the current continued to increase for a short time until it clearly drops. This evolution is most likely due to an overproduction of hydrogen peroxide altering the catalytic activity of HRP [35,36]. One of the most promising applications of glucose biofuel cells is the power generation of implantable medical devices out of body fluids where the glucose concentration of healthy persons is in average 5 mmol L−1 with fluctuation between 4.4 and 6.1 mmol L−1 during a day [37] and the oxygen concentration is around 0.15 mmol L−1 (3 mL/ L) [38]. We thus operated our biofuel cell at 5 mmol L−1 of glucose and air saturated solutions (∼280 mmol L−1 ≘ 6.3mL/L) [39].
3.4. Assembly and stacking of two GBFCs Separated inserts were used to supply individually two GBFCs with the air saturated glucose solution while the flow direction was still directed from the anode to the cathode. The same physiologically relevant 5 mmol L−1 glucose solution was used to supply both GBFCs and was simply separated by T-pieces. Fig. 4A shows the polarization and power curves for the system with in series connection. The power curve was plotted from the polarization curve, recorded by successive 30 s discharges at constant current (0.1 mA–2.8 mA). The performances were monitored at pH 7 in presence of a 5 mmol L−1 air saturated glucose solution flow. With series connection, as expected, the OCV is twice that recorded for one biofuel cell (1.23 V) while the maximum power output and current are
3.3. Performance and stability of one GBFC Fig. 3Ashows the power profile and the polarization curve for 5 mmol L−1 glucose where each point was recorded after 30 s discharge to eliminate the capacitive current. 178
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
Fig. 3. (A) Power profile (blue) and polarization curve (black) for the GBFC at 5 mmol L−1 of glucose in Mc Ilvaine pH 7. Each point represents the value obtained after 30 s discharge between 0.1 and 2.3 mA. (B) Continuous discharge at 6 kΩ (black) and produced energy (blue). (C) Discharge/charge cycles between 0.5 and 0.4 V at 500 Ω (a, black) and produced energy (b, blue), at 3 kΩ (c, red) and produced energy (d, green). (D) Zoom of one cycle chargedischarge at 500 Ω (black) and 3 kΩ (blue) in the beginning (a, black and c, blue respectively) and after 48 h (b, red d, green, respectively). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
1.74 ± 0.086 mW at 0.723 V and 2.4 mA, respectively. While the current remains almost the same, the generated power and the operational potential are even slightly more than two times higher than one
individual GBFC but remain within the standard error of the biofuel cell performances. The stacked glucose biofuel cell connected in parallel (Fig. 4B) exhibits a maximum OCV of 0.757 V and a maximum power
Fig. 4. (A) Scheme (top), power profile (blue) and polarization curve (black) for two GBFCs in series. Each point represents the value obtained after 30 s discharge between 0.1 and 2.8 mA. (B) Scheme (top), power profile (blue) and polarization curve (black) for two GBFCs in parallel. Each point represents the value obtained after 30 s discharge between 0.1 and 5.2 mA. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
179
Journal of Power Sources 392 (2018) 176–180
C. Abreu et al.
output of 2.24 ± 0.112 mW (at 0.458 V and 4.9 mA). Again, improved performances of these two connected GBFCs can be observed compared to an individual glucose biofuel cell but remains at a reasonable scale. For both connections, a relative standard deviation of around 5% was obtained using the same batches of carbon nanotubes and enzymes. It can thus be concluded that this system is compatible for efficient stacking of many GBFCs and allows maintaining the high performances.
[10] A. Heller, Electrical connection of enzyme redox centers to electrodes, J. Phys. Chem. 96 (9) (1992) 3579–3587. [11] Y. Degani, A. Heller, Electrical communication between redox centers of glucose oxidase and electrodes via electrostatically and covalently bound redox polymers, J. Am. Chem. Soc. 111 (6) (1989) 2357–2358. [12] B.A. Gregg, A. Heller, Cross-linked redox gels containing glucose oxidase for amperometric biosensor applications, Anal. Chem. 62 (3) (1990) 258–263. [13] W. Schuhmann, T.J. Ohara, H.L. Schmidt, A. Heller, Electron transfer between glucose oxidase and electrodes via redox mediators bound with flexible chains to the enzyme surface, J. Am. Chem. Soc. 113 (4) (1991) 1394–1397. [14] N. Mano, F. Mao, A. Heller, A miniature biofuel cell Operating in A Physiological buffer, J. Am. Chem. Soc. 124 (44) (2002) 12962–12963. [15] F. Barrière, P. Kavanagh, D. Leech, A laccase–glucose oxidase biofuel cell prototype operating in a physiological buffer, Electrochim. Acta. 51 (24) (2006) 5187–5192. [16] F. Gao, L. Viry, M. Maugey, P. Poulin, N. Mano, Engineering hybrid nanotube wires for high-power biofuel cells, Nat. Commun. 1 (1) (2010) 2. [17] B. Reuillard, C. Abreu, N. Lalaoui, A. Le Goff, M. Holzinger, O. Ondel, et al., Oneyear stability for a glucose/oxygen biofuel cell combined with pH reactivation of the laccase/carbon nanotube biocathode, Bioelectrochemistry 106 (Part A) (2015) 73–76. [18] R.D. Milton, D.P. Hickey, S. Abdellaoui, K. Lim, F. Wu, B. Tan, et al., Rational design of quinones for high power density biofuel cells, Chem. Sci. 6 (8) (2015) 4867–4875. [19] F. Giroud, C. Gondran, K. Gorgy, V. Vivier, S. Cosnier, An enzymatic biofuel cell based on electrically wired polyphenol oxidase and glucose oxidase operating under physiological conditions, Electrochim. Acta. 85 (0) (2012) 278–282. [20] M. Grattieri, M. Tucci, M. Bestetti, S. Trasatti, P. Cristiani, Facilitated electron hopping in nanolayer oxygen-insensitive glucose biosensor for application in a complex matrix, ChemElectroChem 3 (11) (2016) 1884–1889. [21] A. Prévoteau, N. Mano, How the reduction of O2 on enzymes and/or redox mediators affects the calibration curve of “wired” glucose oxidase and glucose dehydrogenase biosensors, Electrochim. Acta. 112 (2013) 318–326. [22] M. Alfonso-Prieto, X. Biarnés, P. Vidossich, C. Rovira, The molecular mechanism of the catalase reaction, J. Am. Chem. Soc. 131 (33) (2009) 11751–11761. [23] W. Jia, C. Jin, W. Xia, M. Muhler, W. Schuhmann, L. Stoica, Glucose oxidase/ horseradish peroxidase Co-immobilized at a CNT-modified graphite electrode: towards potentially implantable biocathodes, Chem. Eur J. 18 (10) (2012) 2783–2786. [24] C. Agnès, B. Reuillard, A. Le Goff, M. Holzinger, S. Cosnier, A double-walled carbon nanotube-based glucose/H2O2 biofuel cell operating under physiological conditions, Electrochem. Commun. 34 (0) (2013) 105–108. [25] K. Elouarzaki, M. Bourourou, M. Holzinger, A. Le Goff, R. Marks, S. Cosnier, Freestanding HRP-GOx redox buckypaper as oxygen-reducing biocathode for biofuel cell applications, Energy Environ. Sci. 8 (7) (2015) 2069–2074. [26] L. Gorton, A. Lindgren, T. Larsson, F.D. Munteanu, T. Ruzgas, I. Gazaryan, Direct electron transfer between heme-containing enzymes and electrodes as basis for third generation biosensors, Anal. Chim. Acta. 400 (1–3) (1999) 91–108. [27] S.C. Barton, M. Pickard, R. Vazquez-Duhalt, A. Heller, Electroreduction of O2 to water at 0.6 V (SHE) at pH 7 on the ‘wired’ Pleurotus ostreatus laccase cathode, Biosens. Bioelectron. 17 (11–12) (2002) 1071–1074. [28] B. Reuillard, A. Le Goff, M. Holzinger, S. Cosnier, Non-covalent functionalization of carbon nanotubes with boronic acids for the wiring of glycosylated redox enzymes in oxygen-reducing biocathodes, J. Mater. Chem. B. 2 (16) (2014) 2228–2232. [29] E. Katz, K. MacVittie, Implanted biofuel cells operating in vivo - methods, applications and perspectives - feature article, Energy Environ. Sci. 6 (10) (2013) 2791–2803. [30] A. Zebda, S. Cosnier, J.-P. Alcaraz, M. Holzinger, A. Le Goff, C. Gondran, et al., Single glucose biofuel cells implanted in rats power electronic devices, Sci. Rep. 3 (2013) 1516. [31] A. Ramanavicius, A. Kausaite-Minkstimiene, I. Morkvenaite-Vilkonciene, P. Genys, R. Mikhailova, T. Semashko, et al., Biofuel cell based on glucose oxidase from Penicillium funiculosum 46.1 and horseradish peroxidase, Chem. Eng. J. 264 (2015) 165–173. [32] A. Le Goff, Y. Nedellec, C. Abreu, S. Cosnier, M. Holzinger, inventors; Pile à Biocombustible. France, (2017). [33] C. Abreu, Y. Nedellec, A.J. Gross, O. Ondel, F. Buret, A.L. Goff, et al., Assembly and stacking of flow-through enzymatic bioelectrodes for high power glucose fuel cells, ACS Appl. Mater. Interfaces 9 (28) (2017) 23836–23842. [34] B. Reuillard, A. Le Goff, C. Agnes, M. Holzinger, A. Zebda, C. Gondran, et al., High power enzymatic biofuel cell based on naphthoquinone-mediated oxidation of glucose by glucose oxidase in a carbon nanotube 3D matrix, Phys. Chem. Chem. Phys. 15 (14) (2013) 4892–4896. [35] S. Asad, S.-F. Torabi, M. Fathi-Roudsari, N. Ghaemi, K. Khajeh, Phosphate buffer effects on thermal stability and H2O2-resistance of horseradish peroxidase, Int. J. Biol. Macromol. 48 (4) (2011) 566–570. [36] B. Limoges, J.-M. Savéant, D. Yazidi, Quantitative analysis of catalysis and inhibition at horseradish peroxidase monolayers immobilized on an electrode surface, J. Am. Chem. Soc. 125 (30) (2003) 9192–9203. [37] Screening for type 2 diabetes, Diabetes Care 26 (suppl. 1) (2003) s21–s24. [38] R.N. Pittman, San rafael (CA): morgan & claypool life sciences; 2011. Chapter 4. Oxygen transport, in: R.N. Pittman (Ed.), Regulation of Tissue Oxygenation. San Rafael (CA): Morgan & Claypool Life Sciences, 2011. [39] J.H. Carpenter, New measurements of oxygen solubility in pure and natural water, Limnol. Oceanogr. 11 (2) (1966) 264–277.
4. Conclusion A flow through design of a glucose biofuel cell allowing the stacking of two GBFCs and its connection in parallel and in series was used to benefit from the hydrogen peroxide production at the GOx-based bioanode to supply the cathodic biocatalyst HRP with its natural substrate. Each cell was supplied by a glucose flow which is directed from the anode through a cellulose membrane to the cathode. By this way, the biocathode is not dependent on the oxygen concentration in solution since the substrates glucose and oxygen are converted at the bioanode. Furthermore, the produced peroxide is related with the glucose concentration and leads to equilibrated performances of the bioelectrodes. This biofuel cell setup is operational until a glucose concentration of 20 mmolL−1 but we decided to perform the experiments at 5 mmolL−1 relating to the average glucose concentration in body fluids of mammals. Thus, this design is promising for the power supply of implantable medical devices because all products are biocompatible (water and gluconolactone). Two stacked biofuel cells provided a good power output when connected either in series or in parallel. This setup further allows the association of an almost unlimited number GBFC in series and parallel which should make it possible to reach the necessary voltage and current outputs to run devices without the need of supplemental step-up converters or capacitors. However, further efforts are necessary in terms of operational stability before such stacked biofuel cells can be envisioned for the power supply of implantable devices even when stabilities over few days in charge/discharge mode are quite encouraging. Acknowledgment The region Auvergne-Rhône-Alpes is acknowledged for the PhD funding of C. Abreu. The authors would also like to thank the mechanical team of LiPhy at Grenoble for the fabrication of the PTFE and Plexiglass plates. The authors wish further to acknowledge the support from the platform Chimie NanoBio ICMG FR 2607 (PCN-ICMG), from the Institute Carnot PolyNat at Grenoble, and from the LabEx ARCANE (ANR-11-LABX-0003-01). References [1] S. Cosnier, A. Le Goff, M. Holzinger, Enzymatic fuel cells: from design to implantation in mammals, in: E. Katz (Ed.), Implantable Bioelectronics, Wiley-VCH Verlag GmbH & Co. KGaA, 2014, pp. 347–362. [2] M. Rasmussen, S. Abdellaoui, S.D. Minteer, Enzymatic biofuel cells: 30 years of critical advancements, Biosens. Bioelectron. 76 (2016) 91–102. [3] R. Wilson, A.P.F. Turner, Glucose oxidase: an ideal enzyme, Biosens. Bioelectron. 7 (3) (1992) 165–185. [4] G. Wohlfahrt, S. Witt, J. Hendle, D. Schomburg, H.M. Kalisz, H.-J. Hecht, 1.8 and 1.9 A resolution structures of the Penicillium amagasakiense and Aspergillus Niger glucose oxidases as a basis for modelling substrate complexes, Acta Crystallogr. D. 55 (5) (1999) 969–977. [5] O. Courjean, F. Gao, N. Mano, Deglycosylation of glucose oxidase for direct and efficient glucose electrooxidation on a glassy carbon electrode, Angew. Chem. Int. Ed. 48 (32) (2009) 5897–5899. [6] S.B. Bankar, M.V. Bule, R.S. Singhal, L. Ananthanarayan, Glucose oxidase — an overview, Biotechnol. Adv. 27 (4) (2009) 489–501. [7] J. Wang, Electrochemical glucose biosensors, Chem. Rev. 108 (2) (2008) 814–825. [8] B.E. Watt, A.T. Proudfoot, J.A. Vale, Hydrogen peroxide poisoning, Toxicol. Rev. 23 (1) (2004) 51–57. [9] P. Kavanagh, D. Leech, Mediated electron transfer in glucose oxidising enzyme electrodes for application to biofuel cells: recent progress and perspectives, Phys. Chem. Chem. Phys. 15 (14) (2013) 4859–4869.
180