Fully Enzymatic Membraneless Glucose|Oxygen Fuel Cell That

Jan 11, 2016 - A maximum power density of 275 μW cm–2 is obtained in 5 mM glucose in PBS, the highest to date under these conditions, providing suf...
1 downloads 10 Views 1MB Size
Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

A fully enzymatic membrane-less glucose|oxygen fuel cell provides 0.275 mA cm in 5 mM glucose operates in human physiological solutions and powers transmission of sensing data #2

Peter Ó Conghaile, Magnus Falk, Domhnall MacAodha, MARIA Evgenievna YAKOVLEVA, Christoph Gonaus, Clemens K Peterbauer, Lo Gorton, Sergey Shleev, and Donal Leech Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b03745 • Publication Date (Web): 11 Jan 2016 Downloaded from http://pubs.acs.org on January 15, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A fully enzymatic membrane-less glucose|oxygen fuel cell provides 0.275 mA cm‒2 in 5 mM glucose, operates in human physiological solutions, and powers transmission of sensing data Peter Ó Conghaile,† Magnus Falk,‡ Domhnall MacAodha,† Maria E. Yakovleva,§ Cristoph Gonaus,∥ Clemens K. Peterbauer,∥ Lo Gorton,§ Sergey Shleev‡ and Dónal Leech*,† †

School of Chemistry & Ryan Institute, National University of Ireland, Galway, Ireland. Department of Biomedical Science, Faculty of Health and Society, Malmö University, 20560 Malmö, Sweden § Department of Biochemistry and Structural Biology, Lund University, PO Box 124, 221 00 Lund, Sweden. ∥Food Biotechnology Lab, Department of Food Sciences and Technology, BOKU-University of Natural Resources and Life Sciences ‡

KEYWORDS Enzymatic fuel cell; biofuel cell; glucose; osmium; mediator.

ABSTRACT: Co-immobilization of pyranose dehydrogenase as an enzyme catalyst, osmium redox polymers, [Os(4,4'-dimethoxy2,2'-bipyridine)2(poly-vinylimidazole)10Cl]+ or [Os(4,4'-dimethyl-2,2'-bipyridine)2(poly-vinylimidazole)10Cl]+ as mediators, and carbon nanotube conductive scaffolds, in films on graphite electrodes provides enzyme electrodes for glucose oxidation. The recombinant enzyme and a de-glycosylated form, both expressed in Pichia pastoris, are investigated and compared as biocatalysts for glucose oxidation using flow injection amperometry and voltammetry. In the presence of 5 mM glucose in phosphate buffered saline (50 mM phosphate buffer solution, pH 7.4, with 150 mM NaCl), higher glucose oxidation current densities, 0.41 mA cm−2, are obtained from enzyme electrodes containing the de-glycosylated form of the enzyme. The optimized glucose-oxidizing anode, prepared using de-glycosylated enzyme co-immobilized with [Os(4,4'-dimethyl-2,2'-bipyridine)2(poly-vinylimidazole)10Cl]+ and carbon nanotubes, was coupled with an oxygen-reducing, bilirubin oxidase on gold nanoparticle dispersed on gold electrode, as a biocathode to provide a membrane-less fully enzymatic fuel cell. A maximum power density of 275 µW cm−2 is obtained in 5 mM glucose in PBS, the highest to date under these conditions, providing sufficient power to enable wireless transmission of a signal to a data-logger. When tested in whole human blood and un-stimulated human saliva maximum power densities of 73 µW cm−2 and 6 µW cm−2 are obtained for the same fuel cell configuration, respectively.

Development of prototype biodevices with wireless capability and an ability to self-power is of scientific and practical importance for application as self-sustained medical and/or portable devices. Implantable or semi-implantable (i.e. attachable, floating, etc), self-powered wireless sensor-systems could allow for real-time biomedical monitoring or intervention. Research on self-contained biodevices is an interdisciplinary endeavor spanning scientific, computing, engineering and medical disciplines, with significant research focus and a recent rapid growth in the number of publications in this area of bioelectronics.1–5 Use of enzymes as catalysts to oxidize glucose at an anode and to reduce O2 at a cathode, that when combined act as a glucose|O2 enzymatic fuel cell (EFC), shows promise as a technology for conversion of in-vivo available chemical energy to electrical power.6–8 Immobilization of substrate specific enzymes at electrode surfaces opens up the possibility of device miniaturization, by eliminating the requirement of a separating ion-exchange interface between anolyte and catholyte: a membrane-less EFC.

Pyranose dehydrogenase (PDH, EC 1.1.99.29) a glycosylated extracellular oxidoreductase, can be isolated from the litter-decomposing fungus Agaricus meleagris,9 and carries a flavin adenine dinucleotide (FAD) prosthetic group covalently bound to a polypeptide chain of the protein. This enzyme is capable of oxidizing a range of non-phosphorylated sugars at their C2 and C3 carbon,9 with recent studies by Tan et al.10 showing that C2 is the principal site of oxidation of glucose by PDH.10 The enzyme does not utilize oxygen as an electron acceptor,11 compared to, for example, glucose oxidase (GOx) and pyranose oxidase (POx). This makes PDH an attractive enzyme for adoption as a catalyst in a glucose|oxygen membrane-less EFC, because of issues associated with competition for electrons by oxygen12, and with toxicity of the H2O2 product of oxygen reduction when using GOx and POx.13 PDH also possesses a wider range of substrate specificity and regioselectively compared to GOx, attributed to the unique structure of the region surrounding the flavin pocket.10,11 For example, Tasca et al. reported broad specificity for enzyme electrodes prepared with PDH and an [Os(4,4'-dimethyl-2,2'bipyridine)2(poly-vinyl imidazole)10Cl]+ (Os(dmbpy)PVI)

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

redox polymer as electron transfer mediator.14 Production of greater amounts of the enzyme is achieved using heterologous expression in Pichia pastoris. De-glycosylation of this recombinant PDH (dgPDH),15 resulted in a 2-fold increase in the current output for enzyme electrodes over the nondeglycosylated form of PDH, using the same redox polymer as mediator.16 We recently reported that the dgPDH turns into a form that appears to have fragmented into a 50 kDa and a 20 kDa fragment under denaturing conditions.17 The resulting enzyme has higher specific activity for glucose oxidation compared to PDH, and enzyme electrodes prepared using the dgPDH co-immobilized with Os(dmbpy)PVI produce ~0.15 mA cm−2 in 5 mM glucose solutions compared to ~0.01 mA cm−2 for PDH-modified electrodes.17 The addition of multiwalled carbon nanotubes (MWCNT) as a component in glucose-oxidizing enzyme electrodes results in increased current and stability in their application as biofuel cell electrodes,18, 19 with evidence that this increase is related to improved retention of enzyme activity within such films.20 In addition, we have reported that utilization of glutaraldehyde (GA) vapors to crosslink GOx and osmium-based redox polymer, for application to biofuel cell anodes oxidizing glucose, provides higher current densities for glucose oxidation, and an improved signal stability, over an approach using a diepoxide crosslinker.18 Here we report on a comparison of glucose oxidation by enzyme electrodes prepared using redox polymers, MWCNT and PDH variants (PDH and dgPDH), co-immobilized using GA vapors on graphite electrodes, to provide for current generation for a glucose-oxidizing anodic half-cell for application to EFCs. Furthermore, assembly of these PDH-based enzyme electrodes as anodes with oxygen-reducing cathodes based on films of a Myrothecium verrucaria (BOx) adsorbed on gold nanoparticles (AuNPs) on gold electrodes21, 22, 23 is undertaken for testing of EFC performance in human blood and saliva. Finally, we report on demonstration that the EFC that provides the highest power output, when placed in 5 mM PBS, can power a device24 capable of wireless transmission of sensing data. EXPERIMENTAL SECTION Materials. All chemicals and biochemicals were, unless otherwise stated, purchased from Sigma-Aldrich (Ireland) and used as received. All solutions are made from Milli-Q grade water unless otherwise stated. Multiwalled carbon nanotubes (MWCNTs, from Sigma-Aldrich) were purified by refluxing in nitric acid for 6 h (20 mg mL–1 in HNO3), with the treated MWCNTs isolated by filtration and washed repeatedly with distilled water until complete nitric acid removal. Synthesis of the redox polymers was achieved by adapting literature procedures.25, 26 Human resting saliva (pH 7.3) and blood (pH 7.5) was donated from a healthy volunteer. Blood was collected using BD Vacutainers® containing 0.109 mM sodium citrate to prevent blood coagulation. The glucose concentration within the blood is estimated as 5.4 mM using a glucose analyser 201+ from HemoCue AB (Ängelholm, Sweden). Sugar-oxidizing enzymes. Glycosylated PDH (EC 1.1.99.29) 27 identified from Agaricus meleagris was recombinantly expressed in Pichia pastoris according to the previously reported procedure 15 to yield 23.8 mg mL–1 (Bradford assay) protein with volumetric activity of 335 U mL–1 using a ferricenium coupled glucose oxidation assay at 20 °C, as described below. Deglycosylation of PDH was performed by incubation

Page 2 of 9

of 60 mg gPDH with 80 NEB units Endo Hf (1000000 NEB units mL–1, New England Biolabs, Bionordiska AB, Stockholm, Sweden) in a 124 mM sodium citrate buffer (pH 5.5) at 37 °C for 6 hours. Deglycosylated PDH was concentrated with 30 kDa Amicon ultrafiltration tube (EMD Millipore Corporation, USA), washed with 50 mM sodium phosphate buffer (pH 6.5) containing 150 mM NaCl and purified by S300 Sephacryl size exclusion chromatography (GE Healthcare, USA) according to manufacturer’s recommendations. The resulting pool was concentrated with a 30 kDa Amicon tube and washed with 50 mM sodium phosphate buffer (pH 6.5) to provide a protein content of 20.4 mg mL–1 (Bradford assay) and volumetric activity of ~342 U mL–1 using the ferricenium coupled glucose oxidation assay.17 Enzyme activity assay. The enzyme activity of PDH preparations was evaluated using a spectrophotometric activity assay of ferricenium reduction,28 molar absorptivity 4.3 mM–1 cm–1, with a UV-2401 PC spectrophotometer (Shimadzu Deutschland GmbH) at 20 °C. Briefly, an aliquot of the enzyme was added to a mixture of 0.1 mM sodium phosphate buffer containing 137 mM NaCl, pH 7.4, 0.05 mM D-(+)glucose and 0.4 µM Fc+PF6– (freshly prepared in 5 mM HCl) in a one millimeter cuvette and the absorbance was monitored at 300 nm. One unit of enzyme activity is equal to the amount of enzyme required for reduction of 2 µmol of Fc+ per minute at 20 °C. Preparation of anodes. Graphite rods, with a diameter of 3.05 mm (Ringsdorff Werke GmbH, Bonn) were polished on emery paper, rinsed with Milli-Q water and sonicated for 5 min in Milli-Q water and dried at room temperature. Deposition of enzyme electrode films was achieved by pipetting 9.6 µL of a 46.25 mg mL–1 dispersion of acid treated MWCNT, 9.6 µL of a 5 mg mL–1 redox polymer aqueous solution and sufficient volume to deposit 0.048 mg of enzyme (in an aqueous solution) on the surface of the graphite electrode disk. Cross-linking treatment of the films was achieved by placing the enzyme electrodes in a sealed headspace of glutaraldehyde (GA) vapors for 30 min, as previously reported.18, 29 The GA treated electrodes were subsequently immersed in 100 mM NaBH4 solution for 5 s, removed and rinsed in Milli-Q water, to reduce the Schiff bases formed upon GA crosslinking.30, 31 All current densities are normalized to the geometric surface of the electrode. Preparation of AuNP based BOx biocathode. AuNPs were prepared following a previously reported citrate reduction procedure.21 The AuNP size was estimated to be ~20 nm, based on the UV–Vis spectrum 33 and by SEM.21 The AuNPs were concentrated by centrifugation and 98% of the supernatant was removed. The precipitated AuNPs were re-suspended by ultrasonication and stored as a 50 times concentrated AuNPs dispersion at 4 °C. Polycrystalline gold-disk electrodes (Bioanalytical Systems, USA) were polished with 1 µm alumina slurry cleaned in Piranha solution (mixture of volumetric 3:1 concentrated sulfuric acid and 30% hydrogen peroxide solution: CAUTION: Piranha solution reacts violently with organics and proper caution should be taken when handling) for 3 min and then polished with 0.05 µm alumina slurry. Any debris on the electrode was then removed by placing electrodes in a solution in an ultrasonic bath. Three-dimensional AuNP-modified electrodes were fabricated by solution casting 22, 23 1.5 µL of concentrated AuNP dispersion onto the Au electrode surface and allowing to dry. This procedure was undertaken 3 times for each electrode. The electrodes were subse-

ACS Paragon Plus Environment

Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

quently subjected to 20 potential cycles between -0.3 and 1.5 V vs. Ag/AgCl at a scan rate of 0.1 V s−1 in 0.5 M H2SO4 until well-defined voltammograms were obtained.34 Thereafter, the electrodes were rinsed with water and a 10 µL of a BOx solution (1 mg mL−1) was added to the electrode surface and left for incubation for 2 h. The electrodes were rinsed with 50 mM phosphate buffer, pH 7.4 prior to testing. Electrochemical methods. Flow injection measurements were performed with a flow-through wall-jet amperometric cell, containing a platinum wire as the counter and an Ag/AgCl (0.1 M KCl), as a reference electrode. The applied potential was controlled by a three-electrode potentiostat (Zäta Electronics, Höör, Sweden). The response of graphite disc working electrodes was registered with a recorder (BD 112, Kipp & Zonen, Utrecht, Netherlands). For the introduction of the samples an injector (Rheodyne, type 7125 LabPR, Cotati, CA, USA) with a 50 µL loop was used. All concentrations in the injected samples were corrected for the dispersion factor of the FIA system, determined to be 1.088. Linear sweep voltammetry was performed with an Autolab PGSTAT 30 from Eco Chemie (Utrecht, The Netherlands) electrochemical system using graphite disks as working electrode, an Ag/AgCl reference electrode and a Pt foil as counter electrode. Electrochemical cells were studied at room temperature in gently stirred air-saturated solutions. RESULTS AND DISCUSSION Enzyme electrodes for operation as anodes are developed based on previous results29 by co-immobilization of PDH or dgPDH enzyme with Os(dmbpy)PVI or [Os(4, 4'-dimethoxy2,2'-bipyridine)2(poly-vinylimidazole)10Cl]+ (Os(dmobpy)PVI) redox polymers and MWCNTs on graphite disk electrodes. The electron-shuttling redox polymers are able to electrically connect the PDH cofactor to the surface of the electrode, with electron transfer by collisions between the reduced and oxidized forms of the mobile osmium redox centers that are tethered to the polymer backbone (Figure 1).35, 36 The redox polymers, Os(dmobpy)PVI and Os(dmbpy)PVI, are selected for screening of mediation of glucose oxidation by co-immobilized PDH variants in films on carbon electrodes, because of their relatively low redox potentials, of −20 mV and 110 mV vs. Ag/AgCl, respectively, providing the possibility of higher cell voltages for EFCs compared to selection of redox polymers with more positive redox potentials. The relative mass of redox polymer, enzyme and MWCNT deposited are selected to match that determined to be appropriate for glucose oxidation by enzyme electrodes of glucose oxidase (GOx) and [Os(2,2'-bipyridine)2(poly-vinylimidazole)10Cl]+ as mediating redox polymer.18 Effect of applied potential on enzyme electrode response. Selection of the potential to be applied in a flow injection assay (FIA) system is based on comparison of the normalized current density for oxidation of 5 mM glucose (in PBS, pH 7.4) as a function of applied potential using enzyme electrode prepared by co-immobilization of either Os(dmbpy)PVI or Os(dmobpy)PVI with dgPDH enzyme and MWCNTs, Figure 2. Applied potentials of 0 V and 0.18 V vs. Ag/AgCl (0.1 M KCl) were selected, based on producing 80% of the maximum current for enzyme electrodes using Os(dmobpy)PVI and Os(dmbpy)PVI, respectively, Figure 2. This potential provides a compromise to permit evaluation of current whilst operating at low potentials, in selecting an anode for EFC assembly.

Figure 1. Simplified schematic representation of glucose oxidation by enzyme electrodes containing PDH and a PVI-bound osmium mediator (structure presented in the inset, where R represents a methoxy- or methyl- group.

Figure 2. Dependence of catalytic current, normalized to the maximum observed current for each, of the Os(dmbpy)PVI (●) or Os(dmobpy)PVI (■) modified graphite electrodes, MWCNT and dgPDH as a function of applied potential. Experiments were performed in 50 mM phosphate buffer at pH 7.4 containing 5 mM glucose. The flow rate was 0.5 mL min–1.

Selection of optimized anode. Enzyme electrodes for glucose oxidation prepared by addition of MWCNTs demonstrate at least a 7-fold increase in current densities over those prepared without MWCNT. Such an increase in glucose oxidation currents upon inclusion of MWCNT into polymer-bound osmium mediator and enzyme films on electrodes has been reported previously for GOx-based systems.18, 37 Under pseudo-physiological conditions selected, 50 mM phosphate buffer at pH 7.4 containing 5 mM glucose, films prepared using PDH produce lower glucose oxidation current

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

density, figures 3a and 3b. For example, current density of 0.03 ± 0.01 mA cm−2 is produced in 5 mM glucose and PBS by enzyme electrodes based on MWCNT, Os(dmbpy)PVI and the PDH compared to 0.41 ± 0.04 mA cm−2 using dgPDH. For films prepared using Os(dmobpy)PVI and MWCNT the highest glucose oxidation current densities, though lower than those obtained with Os(dmbpy)PVI, are also recorded for films prepared with dgPDH, providing 0.30 ± 0.05 mA cm−2, compared to those prepared with PDH which produced 0.02 ± 0.01 mA cm−2. For comparison, glucose oxidation current densities for films of Os(dmobpy)PVI or Os(dmbpy)PVI, each co-immobilized with MWCNT and an FAD-dependent glucose dehydrogenase (GDH) on graphite electrodes produced 0.20 or 0.80 mA cm−2, respectively, under non-FIA, quiescent, conditions.29 In addition we previously reported38 glucose oxidation current densities of 0.30 and 0.42 mA cm−2 in quiescent 5 mM glucose PBS, for films of PQQ-dependent GDH and MWCNT co-immobilized with the redox complexes [Os(4,4'-dimethoxy-2,2'-bipyridine)2(4aminomethylpyridine)Cl]+ and [Os(4,4'-dimethyl-2,2'bipyridine)2(4-aminomethylpyridine)Cl]+, respectively.

Figure 3. Dependence of the current densities on the glucose concentration of films of PDH (■) and dgPDH (●) co-immobilized with Os(dmbpy)PVI (a) or with Os(dmobpy)PVI (b) in the presence of MWCNT (n=3). Measured with the FIA system in 50 mM PBS, pH 7.4, at a flow rate of 0.5 mL min–1 at an applied potential of 0 mV and 180 mV vs. Ag/AgCl (0.1 M KCl) for Os(dmbpy)PVI and with Os(dmobpy)PVI, respectively.

Page 4 of 9

The selected enzyme mass (0.048 mg) in film preparation equates to specific glucose oxidation activity (using the spectrophotometric assay) of 0.68 U for PDH enzyme electrodes, compared to 5.3 U for dgPDH enzyme electrodes. Glucose oxidation current density can thus be normalized by activity for comparison under the pseudo-physiological conditions selected. Higher activity normalized current density is obtained for enzyme electrodes based on MWCNT, Os(dmbpy)PVI and dgPDH, providing 0.08 mA U-1 cm-2 compared to 0.05 mA U-1 cm-2 for the PDH enzyme electrodes. A similar trend is observed for enzyme electrodes based on the Os(dmobpy)PVI redox polymer. For comparison, a Zafar et al.39 reported glucose oxidation currents, normalized to glucose activity, of 0.03 mA U–1 cm–2 for films of Os(dmbpy)PVI co-immobilized with glucose dehydrogenase from Glomerella cingulata in neutral buffer containing glucose. Comparison of the slopes of linear portions of the calibration curve normalized to activity provides an insight into the sensitivity of the different enzyme electrodes to glucose. Electrodes prepared using dgPDH provide the highest value of 15.9 ± 1.5 µA mM-1 U-1 cm-2 using Os(dmbpy)PVI mediator or 9.9 ± 2.6 µA mM-1 U-1 cm-2 using Os(dmobpy)PVI mediator compared to PDH enzyme electrode response of 9.3 ± 1.1 µA mM-1 U-1 cm-2 and 5.5 ± 1.9 µA mM-1 U-1 cm-2, respectively. Under similar conditions, enzyme electrodes of Glomerella cingulata glucose dehydrogenase and Os(dmbpy)PVI39 produced 2.8 µA mM-1 U-1 cm-2. Assembled enzymatic fuel cells. Assembly of membraneless fully enzymatic fuel cells operating in 5 mM glucose containing PBS, pH 7.4 was implemented by coupling glucoseoxidizing PDH electrodes to an oxygen-reducing enzyme electrode. The oxygen-reducing cathode consists of BOx deposited on AuNPs at Au disk electrodes (BOx/AuNPs/Au) producing current densities of ~0.40 mA cm–2 in air-saturated PBS, pH 7.4 (Supplementary Information, Figure S2).21 The cathode may thus limit the power output in a fuel cell composed of PDH enzyme electrodes as anodes coupled to cathodes of same area at the anode. Anodes of geometric area 0.0175 cm2 are therefore coupled to cathodes of 0.08 cm2 geometric area, selected to ensure anode limitation of the fuel cell power output (Supplementary Information, Figure S1 and S2) for investigation of EFC power output, with power density normalized to anode geometric area. From cell polarization studies recorded potentiodynamically (0.1 mV s–1) the initial drop in cell voltage at low current densities is associated with kinetic polarization, Figure S3 (Supporting Information). Further cell voltage losses, in proportion to the current density, is indicative of ohmic drop. At higher current densities, concentration overpotential (mass transport) results in steep cell voltage loss. As expected, EFCs based on anodes containing the dgPDH provide higher maximum power outputs than those based on anodes containing PDH, Figure 4 and 5. The use of the Os(dmbpy)PVI as redox polymer in the anode provides double the maximum power density compared to use of Os(dmobpy)PVI, for both the dgPDH and PDH enzyme electrodes in an EFC, as expected from the comparison of steady state current densities (Figure 3). The highest maximum power density of 275 ± 50 µW cm–2, at a cell voltage of 0.30 V is obtained using anodes based on MWCNT, Os(dmbpy)PVI and dgPDH, Figure 5b. The observed power densities for the membrane-less EFC system based on combining macroscopic enzyme electrodes are almost double the maximum power density of 145 µW cm–2 obtained under similar conditions

ACS Paragon Plus Environment

Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

using films of GOx crosslinked to Os(dmbpy)PVI and Myceliophthora thermophila laccase crosslinked to [Os(2,2'bipyridine)2(poly-vinylimidazole)10Cl]+ (Os(bpy)PVI) on graphite electrodes with MWCNT as anode and cathode, respectively.40 Soukharev et al.41 report an EFC prototype using GOx from Aspergillus niger and a fungal laccase, coimmobilized with osmium redox polymers on 7 µm diameter, 2 cm long, carbon fibres, producing 350 µW cm–2 power density in 15 mM glucose, O2 saturated pH 5 buffered solutions, dropping to a quarter of that maximum power, 90 µW cm–2, when tested at 5 mM glucose concentration. Replacement of GOx from Aspergillus niger with a GOx from Penicillium pinophilum resulted in a maximum power density of 280 µW cm–2 for the same EFC configuration operating at 37 °C in 5 mM glucose, but at pH 5 not pH 7.4.42

Figure 5. Average power curves (n=3) with error bars, recorded by 0.1 mV s–1 linear sweep voltammetry in 50 mM PBS containing 5 mM glucose for membrane-less EFCs based on BOx on AuNPs as a cathode combined with either (a) PDH or (b) dgPDH co-immobilized with Os(dmbpy)PVI, MWCNT as anodes.

Figure 4. Average power curves (n=3) with error bars, recorded by 0.1 mV s–1 linear sweep voltammetry in 50 mM PBS containing 5 mM glucose for membrane-less EFCs based on BOx on AuNPs as a cathode combined with anodes based on coimmobilization of Os(dmobpy)PVI and MWCNTS with PDH (a) or dgPDH (b).

Wireless transmission device system. An obvious target for applications based on enzymatic fuel cell research is for in vivo deployment where fuel can be withdrawn from the flow of blood to provide a power supply for small electronic devices implanted or semi-implanted into the body. In this context, a wireless self-powered sensing device was designed and created, where an electronic circuit was constructed along with computer control software and USB radio receiver that enabled transmission of data captured by a sensor, both powered by an EFC, to a remote (~3 m) receiver.24 However, single, non-stacked EFCs deliver low voltages which therefore need to be boosted up to levels so that micro-potentiostats and radio units can be operated. Southcott et al.43 investigated the use of an EFC, with human serum spiked with glucose flowing over buckypaper electrodes, to power a pacemaker and demonstrated the use of a charge pump and DC–DC converter interface circuit to boost the cell voltage, so such an arrangement was used here. A small radio prototype with an integrated antenna and a USB interface, connected to a PC and serving as the receiver radio unit, is used to evaluate the wireless transmission system setup, Figure 6.24

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. Biodevice test set-up, showing an EFC connected to the wireless operational unit and a PC with the developed control software and receiver.

The self-contained device require approximately 44 µA at 0.57 V to charge. When functioning the device is powered by the EFC and sends a signal to a remote receiver when sufficient power is accumulated to operate the signal transmitter. To test for EFC powering of the device an assembled EFC consisting of dgPDH co-immobilized with MWCNT and Os(dmbpy)PVI as anode of geometric area 0.282 cm2 was coupled to a BOx/AuNPs/Au cathode of geometric area 0.24 cm2. It should be noted that a substantial power density of 90 ± 40 µW cm–2 is still produced at a cell voltage of 0.5 V for fuel cells prepared using Os(dmbpy)PVI and dgPDH in the anode enzyme electrode (Figure 5b). This ensures that sufficient current can be produced to charge the capacitor when placed in unstirred pseudo-physiological solutions of 5 mM glucose in PBS solution, 100 mL volume. The initial test permitted the transfer of a signal, once the micro-potentiostat was fully powered up, from the device over approximately 45 minutes resulting in a signal received approximately every 8 minutes. Operation of EFCs in physiological fluids. There has been little reported on the testing of EFC systems in plasma or blood samples. In 2010, Coman et al.44 reported on the assembly of an EFC using cellobiose dehydrogenase and BOx enzyme electrodes, which relied on direct electron transfer between the active site and the electrode for glucose oxidation and oxygen reduction, that provided power density of 3 µW cm–2 in both PBS and human serum. Recently, MacAodha et al.40 described an EFC assembled using a GDH enzyme and Os(dmbpy)PVI at the anode and a biocathode consisting of a Myceliophthora thermophila laccase and Os(bpy)PVI, each co-immobilized onto a graphite electrode in the presence of MWCNTs, that produced a power density of 110 µW cm–2 in PBS and 60 µW cm–2 when operated in artificial plasma.40 EFCs assembled based on enzyme electrodes of Streptomyces coelicolor laccase co-immobilized with a redox complex [Os(2,2′-bipyridine)2(4-aminomethylpyridine)Cl]+ for the cathode, and PQQ-GDH co-immobilized with a redox complex [Os(4,4'-dimethoxy-2,2'-bipyridine)2(4aminomethylpyridine)Cl]+, for the glucose oxidizing anode produced 66 µW cm−2 in 5 mM glucose containing PBS, pH 7.4 at 37°C, dropping to 37 µW cm−2 in human serum. The performance of the fully enzymatic, membrane-less EFCs is evaluated, for the dgPDH-based enzyme electrode as an anode combined with a BOx/AuNPs/Au cathode in stirred

Page 6 of 9

human blood at room temperature. The glucose concentration within the blood is estimated to be 5.4 mM, which is a normal blood glucose value.45 Slow scan linear sweep voltammograms (LSVs) at enzyme electrodes based on dgPDH co-immobilized with Os(dmbpy)PVI and MWCNTs, and at BOx/AuNPs/Au cathode in human blood show onset potentials for current at −80 mV and 480 mV vs. Ag/AgCl, respectively (see Supplementary Information, Figures S2 and S3). The anode current density is significantly higher than the cathode current density, as a result electrode areas were adjusted to ensure an anode limited response. The assembled EFC therefore consists of anodes of geometric area 0.0175 cm2 producing glucose oxidation currents of up to 2.5 µA, lower than the 4.8 µA current for oxygen reduction at cathodes of geometric areas of 0.08 cm2 (Supporting Information Figures S1 and S2). As was the case for the EFCs operating in buffer solutions, the maximum power density in human blood is higher for EFCs based on Os(dmbpy)PVI as anode mediator, producing 73 ± 7 µW cm−2, compared to 34 ± 3 µW cm−2 for those based on Os(dmobpy)PVI, Figure 7. Overall the power density recorded for EFCs operating in blood is approximately a quarter of that observed for identical EFCs operating in glucose containing PBS, Figures 4b, 5b and 7. A differences in power output between operation in blood and PBS has been observed by others, and has been attributed to the presence of anti-oxidants and enzyme-inhibiting compounds present in physiological fluids.43, 46, 47 Biofouling of the electrodes may also contribute to the power difference. For comparison, Wang et al. reported a maximum power density of 2.8 µW cm–2 for an EFC based on direct electron transfer between electrodes and adsorbed cellobiose dehydrogenase, at the anode, and BOx at the cathode, in human blood, slightly lower than the maximum power density observed in PBS.21 Since the EFC is designed to utilize glucose as a fuel and oxygen as an oxidant, the operation of the EFC in human saliva was also investigated. Saliva as a source for glucose and oxygen may prove a more realistic option than blood as it will require no invasive surgery, when healthy, provides a more accessible source of oxygen and consists of a less complex matrix.48 When operated in human saliva, collected from a healthy volunteer, EFC assembled from enzyme electrodes based on Os(dmbpy)PVI co-immobilized with dgPDH and MWCNT at the anode and BOx/AuNPs/Au cathode yield maximum power density of 6.0 ± 0.5 µW cm−2 at 0.45 V, Figure 8. The lower power output of this EFC relative to that recorded for an identical EFC operating in 5 mM glucose in PBS and operated in blood may be attributed to the much lower glucose concentration present in saliva compared to that in blood, i.e. as low as 50 µM in unstimulated saliva.42 Interestingly the 6.0 ± 0.5 µW cm−2 power density at 0.45 V is similar to that recorded for the EFC operating in PBS containing 50 µM glucose (red trace in Figure 8), indicating little impact on power density output in between saliva and the buffer solution. For comparison, Falk et al.49 reported on operation of an EFC in human saliva providing a maximum power density of 2.1 µW cm−2 at an operational voltage of 0.16 V. The peak at 0.16 V, also observed by Falk et al.,49 most likely corresponds to oxidation of compounds such as ascorbic acid present in saliva.

ACS Paragon Plus Environment

Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry Measuring operational stability is important for assessment of the potential for deployment of enzyme-based electrodes, with application to continuous-use biosensors or EFCs. Stability of the EFC assembled from enzyme electrodes based on Os(dmbpy)PVI co-immobilized with dgPDH and MWCNT at the anode and BOx/AuNPs/Au cathode was examined in physiological media and compared to stabilty in a buffer solution. Operational stability testing was performed on EFCs using a continuous voltage load for maximum power and monitoring current flow, subsequent to the cell polarization tests, so that stability testing commenced 2 hours after the immersion of the electrodes into the testing media. Reference to initial power densities refer to the initial response recorded at the commencement of cell testing. After 12 h of immersion, 67%, 22% and 40% of the initial power remained when cells were operating in PBS, human blood or un-stimulated saliva, respectively, yielding half-life, using simple first-order decay model, estimated from the plot in S4 of 13.8 h, 3 h and 24.7 h. For comparison, an EFC based on direct electron transfer between electrodes and adsorbed cellobiose dehydrogenase, at the anode, and BOx at the cathode showed no residual power after 9 hours operation in human blood, compared to 80% power being retained in buffer solution.21 The observed lower stability of the EFCs in human blood relative to that observed in buffer is proposed to be as a result of the electrode surface becoming covered by blood cells,21 with the exact mechanism(s) as yet unclear, warranting further investigation.

Figure 7. Average power curves (n=3) with error bars, recorded by 0.1 mV s–1 linear sweep voltammetry in human blood for membrane-less EFCs based on BOx/AuNPs/Au as a cathode combined with either (a) Os(dmbpy)PVI or (b) Os(dmobpy)PVI co-immobilized with dgPDH, MWCNT as anodes.

CONCLUSIONS Properties of PDH enzymes for glucose oxidation were investigated and compared using flow injection amperometry. The dgPDH is an excellent candidate for mediated oxidation of glucose at the anode of an assembled EFC, yielding current densities of up to 0.41 ± 0.04 mA cm−2 from enzyme electrodes of Os(dmbpy)PVI, MWCNT and dgPDH at 0 V vs Ag/AgCl in 5 mM glucose, 150 mM NaCl, phosphate buffer solution. In an EFC configuration, when paired with BOx on AuNP substrate as cathode, power densities of up to 325 µW cm−2 were achieved, in PBS which contained 5 mM glucose, with an average of 275 ± 50 µW cm−2 (n=3), providing enough power to enable wireless transmission of sensing data by a prototype device. When tested in whole human blood a maximum power density of 73 ± 7 µW cm−2 was achieved, the highest reported power generation in human blood to date. These fuel cells can also produce power densities of up to 6.0 ± 0.5 µW cm−2 on operation in un-stimulated human saliva.

ASSOCIATED CONTENT Supporting Information Supporting Information. Figures S1-S4 of linear sweep voltammograms, polarization curve and power over time. This material is available free of charge via the Internet at http://pubs.acs.org. Figure 8: Power curves, recorded by 0.1 mV s–1 linear sweep voltammetry in un-stimulated human saliva for membrane-less glucose/O2 EFCs based on BOx on AuNPs as a cathode combined with Os(dmbpy)PVI co-immobilized with dgPDH, MWCNT as anode in air-saturated stirred solutions: in 50 mM PBS, 0.05 mM glucose (black); in un-stimulated human saliva (red).

AUTHOR INFORMATION Corresponding Author [email protected] Notes The authors declare no competing financial interest.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACKNOWLEDGMENT The authors would like to thank the ERA-Chemistry award through the Irish Research Council, a Technology Innovation Development Award through SFI Ireland (06/CP/E006) and The Swedish Research Council, for financial support. We also thank Conan Mercer for his contribution towards the graphical abstract.

REFERENCES (1) Kavanagh, P.; Leech, D. Enzymatic Fuel Cells. In Bioelectrochemistry: Fundamentals, Applications and Recent Developments; Alkire, R. C., Kolb D. M., Lipkowski, J., Eds.; Wiley-VCH: Weinheim, 2011; Vol. 13, pp. 229–267. (2) Wang, J. Chem. Rev. 2008, 108, 814–825. (3) Turner, A. P. F. Chem. Soc. Rev. 2013, 42, 3184–3196. (4) Falk, M.; Narváez Villarrubia, C. W.; Babanova, S.; Atanassov, P.; Shleev, S. ChemPhysChem 2013, 14, 2045–2058. (5) Sapsford, K. E.; Algar, W. R.; Berti, L.; Gemmill, K. B.; Casey, B. J.; Oh, E.; Stewart, M. H.; Medintz, I. L. Chem. Rev. 2013, 113, 1904–2074. (6) Barton, S. C.; Gallaway, J.; Atanassov, P. Chem. Rev. 2004, 104, 4867–4886. (7) Leech, D.; Kavanagh, P.; Schuhmann, W. Electrochim. Acta 2012, 84, 223–234. (8) Davis, F.; Higson, S. P. J. Biosens. Bioelectron. 2007, 22, 1224–1235. (9) Sygmund, C.; Kittl, R.; Volc, J.; Halada, P.; Kubátová, E.; Haltrich, D.; Peterbauer, C. K. J. Biotechnol. 2008, 133, 334–342. (10) Tan, T. C.; Spadiut, O.; Wongnate, T.; Sucharitakul, J.; Krondorfer, I.; Sygmund, C.; Haltrich, D.; Chaiyen, P.; Peterbauer, C. K.; Divne, C. PLoS One 2013, 8, e53567. (11) Baron, R.; Riley, C.; Chenprakhon, P.; Thotsaporn, K.; Winter, R. T.; Alfieri, A.; Forneris, F.; van Berkel, W. J. H.; Chaiyen, P.; Fraaije, M. W.; Mattevi, A.; McCammon, J. A. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 10603–10608. (12) Prévoteau, A.; Mano, N. Electrochim. Acta 2013, 112, 318326. (13) Milton, R. D.; Giroud, F.; Thumser, A. E.; Minteer, S. D.; Slade, R. C. T. Phys. Chem. Chem. Phys. 2013, 15, 19371. (14) Tasca, F.; Gorton, L.; Kujawa, M.; Patel, I.; Harreither, W.; Peterbauer, C. K.; Ludwig, R.; Nöll, G. Biosens. Bioelectron. 2010, 25, 1710–1716. (15) Sygmund, C.; Gutmann, A.; Krondorfer, I.; Kujawa, M.; Glieder, A.; Pscheidt, B.; Haltrich, D.; Peterbauer, C.; Kittl, R. Appl. Microbiol. Biotechnol. 2012, 94, 695–704. (16) Yakovleva, M. E.; Killyéni, A.; Ortiz, R.; Schulz, C.; MacAodha, D.; Ó Conghaile, P.; Leech, D.; Popescu, I. C.; Gonaus, C.; Peterbauer, C. K.; Gorton, L. Electrochem. commun. 2012, 24, 120– 122. (17) Yakovleva, M. E.; Killyéni, A.; Seubert, O.; O Conghaile, P.; Macaodha, D.; Leech, D.; Gonaus, C.; Popescu, I. C.; Peterbauer, C. K.; Kjellstrom, S.; Gorton, L. Anal. Chem. 2013, 85, 9852-9858. (18) MacAodha, D.; Ferrer, M. L.; Ó Conghaile, P.; Kavanagh, P.; Leech, D. Phys. Chem. Chem. Phys. 2012, 14, 14667–14672. (19) Tran, T. O.; Lammert, E. G.; Chen, J.; Merchant, S. A.; Brunski, D. B.; Keay, J. C.; Johnson, M. B.; Glatzhofer, D. T.; Schmidtke, D. W. Langmuir 2011, 27, 6201–6210. (20) Osadebe, I.; Leech, D. ChemElectroChem 2014, 11, 1988– 1993. (21) Wang, X.; Falk, M.; Ortiz, R.; Matsumura, H.; Bobacka, J.; Ludwig, R.; Bergelin, M.; Gorton, L.; Shleev, S. Biosens. Bioelectron. 2012, 31, 219–225. (22) Murata, K.; Kajiya, K.; Nakamura, N.; Ohno, H. Energy Environ. Sci. 2009, 2, 1280-1285. (23) Murata, K.; Kajiya, K.; Nukaga, M.; Suga, Y.; Watanabe, T.; Nakamura, N.; Ohno, H. Electroanalysis 2010, 22, 185–190. (24) Falk, M.; Alcalde, M.; Bartlett, P. N.; De Lacey, A. L.; Gorton, L.; Gutierrez-Sanchez, C.; Haddad, R.; Kilburn, J.; Leech, D.; Ludwig, R.; Magner, E.; Mate, D. M.; Conghaile, P. Ó.; Ortiz, R.; Pita, M.; Pöller, S.; Ruzgas, T.; Salaj-Kosla, U.; Schuhmann, W.;

Page 8 of 9

Sebelius, F.; Shao, M.; Stoica, L.; Sygmund, C.; Tilly, J.; Toscano, M. D.; Vivekananthan, J.; Wright, E.; Shleev, S. PLoS One 2014, 9, e109104. (25) Kober, E. M.; Caspar, J. V.; Sullivan, B. P.; Meyer, T. J. Inorg. Chem. 1988, 27, 4587–4598. (26) Forster, R. J.; Vos, J. G. Macromolecules 1990, 23, 4372– 4377. (27) Kittl, R.; Sygmund, C.; Halada, P.; Volc, J.; Divne, C.; Haltrich, D.; Peterbauer, C. K. Curr. Genet. 2008, 53, 117–127. (28) Kujawa, M.; Volc, J.; Halada, P.; Sedmera, P.; Divne, C.; Sygmund, C.; Leitner, C.; Peterbauer, C.; Haltrich, D. FEBS J. 2007, 274, 879–894. (29) MacAodha, D.; Ó Conghaile, P.; Egan, B.; Kavanagh, P.; Sygmund, C.; Ludwig, R.; Leech, D. Electroanalysis 2013, 25, 94– 100. (30) Chaikin, S.; Brown, W. J. Am. Chem. 1949, 3425, 122– 125. (31) Gacesa, P.; Whish, W. J. Biochem. J. 1978, 175, 349–352. (32) Frens, G. Nat. Phys. Sci. 1973, 241, 20–22. (33) Haiss, W.; Thanh, N. T. K.; Aveyard, J.; Fernig, D. G. Anal. Chem. 2007, 79, 4215–4221. (34) Finklea, H. O.; Avery, S.; Lynch, M.; Furtsch, T. Langmuir 1987, 3, 409–413. (35) Gregg, B. A; Heller, A. Anal. Chem. 1990, 62, 258–263. (36) Gregg, B. A.; Heller, A. J. Phys. Chem. 1991, 95, 5970– 5975. (37) Ó Conghaile, P.; Pöller, S.; MacAodha, D.; Schuhmann, W.; Leech, D. Biosens. Bioelectron. 2013, 43, 30–37. (38) Ó Conghaile, P.; MacAodha, D.; Egan, B.; Kavanagh, P.; Leech, D. J. Electrochem. Soc. 2013, 160, G3165–G3170. (39) Zafar, M. N.; Beden, N.; Leech, D.; Sygmund, C.; Ludwig, R.; Gorton, L. Anal. Bioanal. Chem. 2012, 402, 2069-2077. (40) MacAodha, D.; Ó Conghaile, P.; Egan, B.; Kavanagh, P.; Leech, D. ChemPhysChem 2013, 14, 2302–2307. (41) Soukharev, V.; Mano, N.; Heller, A. J. Am. Chem. Soc. 2004, 126, 8368–8369. (42) Mano, N. Chem. Commun. 2008, 2221–2223. (43) Southcott, M.; Macvittie, K.; Halámek, J.; Halámková, L.; Jemison, W. D.; Lobel, R.; Katz, E. Phys. Chem. Chem. Phys. 2013, 15, 6278–6283. (44) Coman, V.; Ludwig, R.; Harreither, W.; Haltrich, D.; Gorton, L.; Ruzgas, T.; Shleev, S. Fuel Cells 2010, 10, 9–16. (45) Polonsky, K. S.; Given, B. D.; Van Cauter, E. J. Clin. Invest. 1988, 81, 442–448. (46) Kang, C.; Shin, H.; Zhang, Y.; Heller, A. Bioelectrochemistry 2004, 65, 83–88. (47) Magner, E. Analyst 2001, 126, 861–865. (48) Edgar, M.; Dawes, C.; O’Mullane, D. Saliva Oral Heal. 2004. (49) Falk, M.; Blum, Z.; Shleev, S. Electrochim. Acta 2012, 82, 191–202.

ACS Paragon Plus Environment

Page 9 of 9

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment