Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 Redox-Polymer-Wired [NiFeSe] Hydrogenase Variants with Enhanced O2 Stability for Triple-Protected High-Current- Density H2-Oxidation Bioanodes Adrian Ruff,*[a] Julian Szczesny,[a] Maria Vega,[b] Sonia Zacarias,[c] Pedro M. Matias,[c, d] S8bastien Gounel,[e] Nicolas Mano,[e] InÞs A. C. Pereira,[c] and Wolfgang Schuhmann*[a] Dedicated to Prof. Bernd Speiser on the occasion of his retirement. Variants of the highly active [NiFeSe] hydrogenase from D. vul- tion, establishing a triple protection for the bioanode. The use garis Hildenborough that exhibit enhanced O2 tolerance were of gas-diffusion bioanodes provided current densities for H2- used as H2-oxidation catalysts in H2/O2 biofuel cells. Two oxidation of up to 6.3 mA cm@2. Combination of the gas-diffu- [NiFeSe] variants were electrically wired by means of low-po- sion bioanode with a bilirubin oxidase-based gas-diffusion O2- tential viologen-modified redox polymers and evaluated with reducing biocathode in a membrane-free biofuel cell under respect to H2-oxidation and stability against O2 in the immobi- anode-limiting conditions showed unprecedented benchmark lized state. The two variants showed maximum current densi- power densities of 4.4 mW cm@2 at 0.7 V and an open-circuit ties of (450 : 84) mA cm@2 for G491A and (476 : 172) mA cm@2 voltage of 1.14 V even at moderate catalyst loadings, outper- for variant G941S on glassy carbon electrodes and a higher O2 forming the previously reported system obtained with the tolerance than the wild type. In addition, the polymer protect- [NiFeSe] wild type and the [NiFe] hydrogenase from D. vulgaris ed the enzyme from O2 damage and high-potential inactiva- Miyazaki F. Introduction Bioelectrocatalysis has gained huge importance in the past de- velopment of novel sustainable energy conversion and storage cades and is considered a promising research field for the de- systems[1] as well as for the green production of value-added chemicals and (solar) fuels.[1, 2] In particular, the use of biocata- lysts to produce sustainable H2/O2-powered biofuel cells with [a] Dr. A. Ruff, J. Szczesny, Prof. Dr. W. Schuhmann Analytical Chemistry—Center for Electrochemical Sciences (CES) high power output has become a major research area in this Faculty of Chemistry and Biochemistry context. Although the biocathodes used for such devices are Ruhr-University Bochum mainly based on rather robust, stable, and easy-to-wire multi- Universit-tsstr. 150, 44780 Bochum (Germany) copper oxidases such as bilirubin oxidase or laccase,[3] the bio- E-mail: adrian.ruff@rub.de wolfgang.schuhmann@rub.de catalyst at the bioanode, that is, highly active H2-oxidizing Ni/ [b] M. Vega Fe-based hydrogenases, typically suffers from pronounced O2 Facultat de BiociHncies sensitivity and fast inactivation at high potentials especially Universitat Autknoma de Barcelona (UAB) under anode-limiting conditions.[4] Hence, the preparation of Carrer de la Vall Moronta, 08193 Bellaterra (Spain) hydrogenase-based bioanodes requires the implementation of [c] Dr. S. Zacarias, Dr. P. M. Matias, Prof. Dr. I. A. C. Pereira specific protection systems, for example, based on O2-reducing Instituto de Tecnologia Qu&mica e Bioljgica Antjnio Xavier (ITQB NOVA) Universidade NOVA de Lisboa low-potential redox polymers, or the use of O2-tolerant but Av. da Repfflblica, 2780-157 Oeiras (Portugal) usually less active hydrogenases, such as, for instance, the hy- [d] Dr. P. M. Matias drogenases from Escherichia coli,[5] Ralstonia eutropha,[6] or Instituto de Biologia Experimental e Tecnoljgica (iBET) Aquifex aeolicus.[7] The introduction of O2-reducing redox poly- Apartado 12, 2780-901 Oeiras (Portugal) mers for electrical wiring and protection of air-sensitive hydro- [e] Dr. S. Gounel, Dr. N. Mano genases was successfully demonstrated for various hydroge- CNRS, CRPP, UMR 5031 Univ. Bordeaux nases including [NiFe] hydrogenase from D. vulgaris Miyaza- 33600 Pessac (France) ki F,[8] [FeFe] hydrogenase from Chlamydomonas reinhardtii,[9] Supporting Information and the ORCID identification number(s) for the and [NiFeSe] hydrogenase from D. vulgaris Hildenborough.[10] author(s) of this article can be found under: Moreover, effective protection was even observed for thin https://doi.org/10.1002/cssc.202000999. films[11] and polymer/enzyme-modified gas-diffusion electrodes T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA. (GDEs), which could be incorporated in membrane-free biofuel This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in cells that exhibited benchmark power densities for polymer- any medium, provided the original work is properly cited. based systems.[12] In addition, the low-potential redox polymer ChemSusChem 2020, 13, 3627 – 3635 3627 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 not only acts as an O2-quenching matrix but also as a Nernst the H2/2 H + couple [approximately @450 mV vs. standard hy- buffer system for the biocatalyst, preventing inactivation at drogen electrode (SHE) at pH 7].[27] The enhanced stability of high potentials.[8] Hence, low-potential redox polymers provide the altered proteins was attributed to a physical blocking a double protection system for such sensitive materials, which effect of the O2 molecule in a hydrophilic channel that con- was even transposable to synthetic catalysts.[13] Moreover, hy- nects the active site of the protein with the enzyme surface, drogenases deactivated under aerobic conditions may be reac- thus preventing oxidation of a specific active-site cysteine tivated by the low-potential polymer matrix, as was shown ligand (Figure 1 a; for a more comprehensive description of the previously for [NiFe][14] and [NiFeSe][10] hydrogenases. structural changes inside the protein shell of the enzyme that Although O2-tolerant hydrogenases derived from the above- lead to the desired O2-blocking effect, see Ref. [27]). This effect mentioned microorganisms can be operated in the presence of distinct levels of O2,[15] inactivation at high potentials still re- mains an issue.[5–7] Furthermore, the overpotential for H2-oxida- tion is often at more positive values than those of O2-sensitive [NiFe] and [FeFe] hydrogenases.[5, 15, 16] Evidently, this will limit the power output of a related biofuel cell because this value will directly affect the maximum open-circuit voltage (OCV). Hence, a combination of redox polymers (double protection shield) and hydrogenase variants with enhanced O2 tolerance would allow the development of triple-protected bioanodes. We want to emphasize that although the redox polymer re- quires a slightly more positive redox potential than the hydro- genase itself to ensure successful electron exchange at oxida- tive conditions, the open-circuit potential of the corresponding bioanode and, consequently, the OCV of the cell will not be limited by the mid-point potential of the redox polymer owing to the pseudo-capacitive properties of the polymer matrix.[17] This effect was shown for glucose oxidase/bilirubin oxidase[18] as well as hydrogenase/bilirubin oxidase-based biofuel cells,[12] in which the anodic catalyst (glucose oxidase, hydrogenase) was electrically wired by means of a redox polymer. The properties of enzymes can be modulated by enzyme en- gineering.[19] For instance, glucose oxidase that uses O2 as a natural electron acceptor could be turned into an O2-insensi- tive enzyme by site-directed mutagenesis[20] and cofactor rede- sign.[21] Artificial maturation of [FeFe] hydrogenase allows for the fine tuning of the properties of the active center of the enzyme.[22] The synthetic nature of the active site allows for the incorporation of specific ligands or the alteration of the overall ligand sphere to adjust the properties of the whole enzyme.[23, 24] Following this approach, the stability of [FeFe] hy- drogenases could be enhanced, and the overall activity of the enzyme can be controlled.[23, 24] Although the maturation pro- Figure 1. (a) Cartoon showing the mechanism of the different behavior of cess occurs spontaneously without any helper proteins or addi- the wild type (1) and the variants (2) of [NiFeSe] hydrogenase when exposed tional cofactors,[25] the preparation of the active site of a [FeFe] to O2. In the wild-type [NiFeSe] hydrogenase, oxygen reaches the active hydrogenase requires distinct synthetic efforts,[26] and the ma- center of the enzyme (1), whereas in the variants G491A and G491S, the turation process is an additional step in the preparation of the pathway to the active center is partially blocked by altered amino acid resi- dues, which hamper the access of O2 (2). For a detailed structural and mech- active enzyme. In contrast, once developed, variants of an anistic description, see Ref. [27]. (b) Schematic of the high-current-density enzyme showing altered properties are produced directly from carbon cloth gas-diffusion H2-oxidation bioanode equipped with a polymer/ the living organisms. The additional maturation step and com- hydrogenase layer. For the immobilization of the active P(N3MA-BA-GMA)- plex synthesis is thus not required. vio/hydrogenase layer, a carbon cloth-based gas-diffusion layer was first modified with an adhesion layer, that is, the redox polymer P(GMA-BA- Recently, it was shown that variants (G491A and G491S) of PEGMA)-vio, which shows a higher hydrophobic monomer content (for a de- the [NiFeSe] hydrogenase from D. vulgaris Hildenborough with tailed description of the electrode architecture and electrochemical gas dif- modification in a specific amino acid close to the active site fusion cell, see Ref. [12]). By combining the low-potential redox polymer led to enhanced stability of the biocatalyst in the presence of with a more O2-tolerant hydrogenase variant, a stable high-current-density bioanode is obtained, which can be operated in a membrane-free H2/O2 bio- O2 while retaining a high activity for H2-oxidation [G491A: up fuel cell. (a, b) The structure of the wt-[NiFeSe] from D. vulgaris Hildenbor- to (4080 : 80) s@1; G491S: up to (2810 : 150) s@1, wild type: ough (5JSH)[28] was used as a representative enzyme model; not drawn to & (4850 : 260) s@1] and a redox potential that is still close to scale. ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3628 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 was evidenced by protein film electrochemistry conducted in a direct electron transfer (DET) regime and in the presence and absence of O2.[27] However, the DET mode does not provide any protection against high-potential inactivation or against high O2 concentrations and is hence impractical for potential applications. Nevertheless, it shows that variants of this type of hydrogenase can be prepared with enhanced O2 stability. In this contribution, we combined the advantages of the en- hanced O2-stable variants of [NiFeSe] hydrogenase from D. vul- garis Hildenborough and the O2 quenching and Nernst buffer properties of low-potential viologen-modified redox polymers to fabricate high-current-density bioanodes, which could be successfully incorporated into H2/O2-powered biofuel cells ex- hibiting benchmark power densities at moderate catalyst load- ings. Results and Discussion Electrical wiring The redox polymer P(N3MA-BA-GMA)-vio [poly(3-azido-propyl methacrylate-co-butyl acrylate-co-glycidyl methacrylate)-vio, with vio = 1-(5-hexyn-1-yl)-1’-methyl-4,4’-bipyridinium; see Fig- ure S1 in the Supporting Information] was used previously for Figure 2. Cyclic voltammograms recorded at a scan rate of 10 mV s@1 of productive electrical wiring and protection of wild-type (a) P(N3MA-BA-GMA)-vio/G491A and (b) P(N3MA-BA-GMA)-vio/G491S films [NiFeSe] hydrogenase from D. vulgaris Hildenborough (wt- immobilized on glassy carbon disk electrodes (3 mm 1) in the absence [NiFeSe]).[10] Benchmark H2-oxidation currents of approximately (black lines, 100 % Ar, purged through solution) and presence of H2 (red 1.7 mA cm@2 at optimized conditions were obtained for lines, 100 % H2, purged through solution). Working electrolyte: phosphate buffer, 0.1 m, pH 7.3, room temperature. P(N3MA-BA-GMA)-vio/wt-[NiFeSe]-modified glassy carbon elec- trodes.[10] Stimulated by our previous findings, we exploited the possibility to electrically wire two O2-tolerant [NiFeSe] var- (1918 : 119) s@1 and G491S: (2416 : 387) s@1; note that the iants, namely G491A and G491S, to the same electrode materi- standard deviations are overlapping], which are below the ac- al. tivity of the wild-type enzyme (4850 : 260 s@1)[27] and below Indeed, cyclic voltammograms of drop-cast P(N3MA-BA- the maximum values reported previously for the variants (see GMA)-vio/G419A and P(N3MA-BA-GMA)-vio/G419S films mea- values above).[27] sured under turnover conditions, that is, under H2 atmosphere The steady-state current observed at high potentials (> (Figure 2 a, b, red curves), showed pronounced catalytic H2-oxi- @0.2 V vs. SHE, Figure 2, red lines) under turnover conditions dation waves with half-wave potentials (a and b: approximate- indicates that the variants can also be effectively protected ly @0.32 V vs. SHE), which matches closely the mid-point po- against high-potential inactivation in contrast to the operation tential of the polymer-bound viologen units ( & 0.34 V vs. SHE, under DET conditions for this type of hydrogenases.[29, 30] black curves). The behavior is in line with the results measured for the wild type (Figure S2 in the Supporting Information). We Oxygen tolerance hence conclude that the [NiFeSe] variants can also be produc- tively wired through the redox polymer P(N3MA-BA-GMA)-vio Because the low-potential redox polymer acts as an O2- in a mediated electron-transfer regime. Moreover, long-term quenching matrix and thus protects the enzyme from O2- chronoamperometric measurements over 7 h under continu- damage,[8] the stability of the variants and the wild type ous turnover conditions showed similar operational stability against oxygen in the immobilized sate was measured in the for the wild type and the two variants (Figure S3 in the Sup- absence of H2 in chronoamperometric experiments. Under porting Information). these conditions, electrons from H2-oxidation are absent and The maximum current densities Jmax measured for the indi- cannot be used by the polymer matrix to reduce incoming vidual, freshly prepared electrodes were calculated as average O2.[8, 31] values from three electrodes as (450 : 84) mA cm@2 for G491A Figure 3 shows chronoamperometric experiments at an ap- and (476 : 172) mA cm@2 for G941S. The wild type shows a Jmax plied potential (Eappl) of + 160 mV (vs. SHE) under alternating value of (752 : 259) mA cm@2 (note that for the wild type and gas-mixture atmospheres. The O2 content in the gas feed was the variants, storage of the modified electrodes leads to a de- stepwise increased after each H2 cycle. To ensure that all H2 crease in overall electrode activity). The results are in line with had been removed before the O2 was added to the gas feed, the measured activities of the used enzyme batches [G491A: the cell was purged with argon. After the background current ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3629 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 (G491A: & 20 % of the initial H2-oxidation current; G491S: & 35 %). The results demonstrate that the variants indeed ex- hibit an increased O2 tolerance compared with the wild-type enzyme owing to a partial blocking of molecular oxygen (ham- pered access) based on the altered amino acid residues in the variants (Figure 1 a) and, by this, that the variants provide an additional protection for the proposed H2-oxidation bioanodes. The electrochemical results obtained with the polymer/enzyme films are in line with the results reported for operating the same hydrogenases in the DET regime.[27] However, strong var- iations in the residual currents were observed after exposure to O2, which is attributed to variations in film thickness and in- homogeneities of the catalytic layers, leading to different diffu- sion profiles of O2. However, in all experiments the variants showed a higher stability towards O2. Reactivation For the wild-type [NiFeSe] hydrogenase, which was deactivated under aerobic conditions, reactivation is known to occur quick- ly at rather negative potentials.[29, 30] Moreover, the reduced low-potential polymer matrix P(N3MA-BA-GMA)-vio is also able to reactivate the inactive [NiFeSe] hydrogenase.[10] To evaluate a possible reactivation behavior of the two [NiFeSe] hydrogenase variants, glassy carbon electrodes modi- fied with P(N3MA-BA-GMA)-vio/G491A and P(N3MA-BA-GMA)- vio/G491S films were exposed to O2 until complete inactivation occurred, as evidenced by the current dropping back to back- ground values. Application of a negative potential of @440 mV (vs. SHE; polymer is fully reduced, inactive mediator form) for 500 s leads to reactivation of the enzyme (Figure 4) as indicat- ed by the oxidative currents, which were observed again when the potential was stepped back to + 160 mV (vs. SHE; t > 500 s, mediator is oxidized, active form). Both potentials were applied under a 90 % H2/10 % Ar gas feed. The wild type shows the same behavior (see Figure S4 in the Supporting Information and Ref. [10]). For practical applications, this effect is highly desirable be- cause a potentially necessary exchange of a deactivated elec- trode in a device can be prevented. Instead, a short cathodic Figure 3. Representative chronoamperometric experiments with (a) P(N3MA- potential pulse might reactivate the electrode, and operation BA-GMA)-vio/wt-[NiFeSe], (b) P(N3MA-BA-GMA)-vio/G491A, and (c) P(N3MA- can be resumed. BA-GMA)-vio/G491S films immobilized on glassy carbon disk electrodes (3 mm 1) under alternating gas feeds (90 % H2/10 % Ar, 100 % Ar as well as 5 % O2/95 %Ar, 10 % O2/90 % Ar, and 15 % O2/85 % Ar). Working conditions: Membrane-free H2/O2 biofuel cells phosphate buffer, 0.1 m, pH 7.3, room temperature; Eappl = + 160 mV vs. SHE. Oxygen-tolerant hydrogenases typical display higher redox po- tentials, which will decrease the maximum OCV of a corre- was reached, the film was exposed to an O2/Ar mixture with sponding biofuel cell compared with their O2-sensitive ana- varying O2 content (5, 10, and 15 %, gray shaded areas in logues. However, because the O2-tolerant variants G491A and Figure 3). The wild type shows a steady decrease of the H2-oxi- G941S show similar H2-oxidation potentials as the wild type,[27] dation activity over all O2/Ar cycles. After exposure to 15 % O2, the electrical wiring is possible with the same polymer the electrode remains inactive when switching the gas feed (Figure 2). Hence, we expect similar OCV values for related bio- back to H2. This is consistent with a fast in-diffusion of O2 to fuel cells as for those based on the wild-type hydrogenase. To the active center of the enzyme (Figure 1 a). In contrast, both evaluate the performance of the bioanodes in a biofuel cell, variants show a rather constant current output after the 5 and polymer/hydrogenase-modified glassy carbon electrodes were 10 % O2 cycle. Moreover, even after exposure to 15 % O2, both combined with a gas-diffusion O2-reducing bilirubin oxidase- electrodes still show a remarkable activity towards H2-oxidation based biocathode. The use of a gas-diffusion system at the ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3630 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 nase-modified glassy carbon electrodes exhibiting maximum absolute currents < 80 mA for all hydrogenases. Membrane-free biofuel cells prepared with the gas-diffusion Bp-BOD-based biocathode in combination with P(N3MA-BA- MA)-vio/wt-[NiFeSe] (Figure S6 a in the Supporting Informa- tion), P(N3MA-BA-MA)-vio/G491A (b), and P(N3MA-BA-MA)-vio/ G491S (c) bioanodes showed OCV values of 1.06, 1.05, and 1.06 V, respectively. As expected, all biofuel cell assemblies show similar or even identical OCV values (note that all electro- des show very similar enzyme and polymer loadings), indicat- ing again the benefits of the two variants compared with typi- cally used O2-tolerant hydrogenases with higher redox poten- tials. The maximum power density with respect to the elec- trode surface area of the bioanode was observed at 0.8 V for all fuel cells and was estimated to be 340 mW cm@2 (wt- [NiFeSe]), 325 mW cm@2 (G491A), and 271 mW cm@2 (G491S). The values are similar to other polymer-based biofuel cells using [NiFe][8] and [FeFe][9] hydrogenases. Recently, we showed that the use of gas-diffusion layers modified with polymer/wt-[NiFeSe] and polymer/[NiFe] films displayed enhanced power output owing to an enhanced mass transport of the gaseous substrate H2 towards the bioa- node.[12] Current densities for the bioanode of close to 8 mA cm@2 and power densities of 3.8 mW cm@2 for biofuel cells with a bilirubin oxidase-modified gas-diffusion biocathode were observed.[12] To demonstrate the relevance of the O2-tol- erant [NiFeSe] variants, carbon cloth-based gas-diffusion layers were first modified with P(GMA-BA-PEGMA)-vio [poly(glycidyl methacrylate-co-butyl acrylate-co-poly(ethylene glycol)meth- acrylate)-vio; for the structure and synthesis of this polymer, see Figure S1 in the Supporting Information and Ref. [12], re- spectively] films followed by the immobilization of an active P(N3MA-BA-GMA)-vio/G491S layer [Figure 1 b; for a detailed de- scription of the preparation process see the Experimental Sec- Figure 4. Chronoamperometric experiments with aerobically deactivated tion; electrodes are denoted as P(GMA-BA-PEGMA)-vio// (a) P(N3MA-BA-GMA)-vio/G491A and (b) P(N3MA-BA-GMA)-vio/G491S films P(N3MA-BA-GMA)-vio/G491S; owing to the limited amount of immobilized on glassy carbon disk electrodes (3 mm 1). First, a potential of enzyme, only the variant G491S was used for the preparation @440 mV vs. SHE was applied for 500 s to fully reduce the viologen-modi- of a H2-oxidation gas-diffusion layer]. fied polymer (the enzyme is reactivated during reduction via the polymer). After switching the potential to + 160 mV vs. SHE (t > 500 s), H2-oxidation The less hydrophilic viologen-modified polymer P(GMA-BA- currents indicate successful reactivation. Working conditions: phosphate PEGMA)-vio acts as an adhesion layer between the hydrophilic buffer, 0.1 m, pH 7.3, room temperature, electrodes were deactivated by ex- active layer and the hydrophobic carbon cloth surface.[12] More- tensive exposure to O2 until any H2-oxidation current was absent. over, the underlying redox polymer layer prevents contribution from DET between the enzyme and the porous electrode sur- cathode side, in which mass transport is not limiting, ensures face and excludes high-potential inactivation.[12] bioanode-limiting conditions. The biocathode was prepared Under gas-diffusion conditions, the bioanode showed abso- with bilirubin oxidase from Bacillus pumilus (Bp-BOD),[32] a lute H2-oxidation currents of approximately 0.8 mA (Figure 5 a). stable multi-copper oxidase used previously in biofuel cells,[18] The modified surface area of the carbon cloth-based bioanode by drop-casting a Bp-BOD stock solution (borate buffer, has a diameter of approximately 4 mm, which results in a sur- 50 mm, pH, pH 9, 54.75 mg mL@1) onto a carbon cloth-based face area of the active layer of approximately 0.126 cm@2, and gas-diffusion layer equipped with a conducting microporous thus maximum current densities of 6.3 mA cm@2 were achieved. Nafion/Teflon/carbon layer with enhanced surface area (for a The values are similar to previously reported polymer-based detailed description of the immobilization process, see the Ex- gas-diffusion systems equipped with wt-[NiFeSe] and [NiFe] hy- perimental Section). In cyclic voltammograms, maximum abso- drogenases (Table S1 in the Supporting Information). However, lute currents for O2 reduction of approximately 180 mA were care must be taken when comparing current densities mea- observed when the gas-diffusion electrode was exposed to air sured with porous electrodes. Because of the 3D structure of (Figure S5 in the Supporting Information). These values are sig- the electrodes, the real surface is often unknown. Hence, the nificantly higher than those obtained for the polymer/hydroge- catalyst loading is a better value for comparison. For the ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3631 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 Figure 5. Characterization of (a) the P(GMA-BA-PEGMA)-vio//P(N3MA-BA-GMA)-vio/G491S-based gas-diffusion bioanode and (b, c) biofuel cells equipped with a Mv-BOD-based gas-diffusion O2-reducing biocathode (operated in 100 % O2) in 0.1 m phosphate buffer, pH 7.4. (a) Cyclic voltammograms of the gas-diffusion P(N3MA-BA-GMA)-vio/G491S bioanode in the absence (black curve) and presence of H2 (red curve); scan rate = 5 mV s@1. (b) Current density (red squares, right ordinate) and power density (black squares, left ordinate, with respect to the geometric surface area of the modified part of the bioanode, & 0.126 cm@2). (c) Operational stability of the biofuel cell over 10 h at 0.7 V. Nominal biocatalyst loading: 8.4 nmol cm@2/1.06 nmol electrode@1. G491S-based electrodes, the catalyst loading is 8.4 nmol cm@2/ The Mv-BOD was then immobilized by means of a convention- 1.06 nmol electrode@1. Interestingly, for the wt-[NiFeSe] hydro- al drop-casting process and was operated in the DET genase, current densities of only 5.3 mA cm@2 were observed regime.[12, 33] A high catalyst loading was used to ensure anode- with a substantially higher catalyst loading of 27.0 nmol cm@2/ limiting conditions (nominal enzyme loading: 1.2 mg elec- 3.4 nmol electrode@1 as reported in our previous work (see trode@1). Absolute currents under gas-diffusion conditions Ref. [12] and Table S1 in the Supporting Information), which (100 % O2) reached approximately 2 mA (Figure S7 in the Sup- largely exceeds the values of the G491S variant at almost iden- porting Information), which largely outperforms the bioanode tical overall polymer loading [wt-[NiFeSe]: 230 mg electrode@1 ( & 0.8 mA, Figure 5 a). The fully assembled H2/O2 biofuel cell (previous work, Ref. [12]); G491S: 260 mg electrode@1]. At a (Figure 5 b) showed an OCV of 1.14 V, which is slightly higher lower catalyst loading of 12.1 nmol cm@2/1.53 nmol electrode@1 than the values obtained on glassy carbon electrodes (1.05– (polymer loading 230 mg electrode@1), which is only slightly 1.06 V); this might be attributed to the slightly lower overpo- higher than the loading of the G491S enzyme, the wt-[NiFeSe] tential for O2 reduction of Mv-BOD compared with Bp-BOD.[34] shows a Jmax value of only 3.6 mA cm@2 (see Ref. [12]). This The maximum power density was reached at 0.7 V and was es- effect might be related to an improved incorporation of the timated to be 4.4 mW cm@2. This value even outperforms our G941S variant in the polymer film when immobilized on the previously reported value for the [NiFe]-based biofuel cell rather hydrophobic carbon cloth-based electrodes. A stronger (3.6 mW cm@2)[12] and—to the best of our knowledge—sets a interaction prevents leaching of the enzyme and thus ensures new benchmark for a biofuel cell using redox-polymer-based a higher local concentration of the biocatalyst during the ex- bioanodes (Table S1 in the Supporting Information). Moreover, periment. In addition, a loss of activity in the immobilized state the catalyst loading is significantly lower than the [NiFe] for the wild type may also contribute to a reduced electrode system (31.8 nmol cm@2/4 nmol electrode@1)[12] reported previ- activity. An effect of different polymer-to-enzyme ratios can be ously (Table S1 in the Supporting Information). ruled out because almost identical polymer loadings were Cyclic voltammograms (Figure S8 in the Supporting Informa- used for all experiments. However, the effect seems to be spe- tion) measured before and after biofuel cell operation showed cific for the porous, hydrophobic carbon cloth electrodes be- similar values for the bioanode, with the slightly higher cur- cause the wild type shows a higher activity on flat glassy rents after the biofuel cell test most likely as a result of carbon electrodes (see Figure 2 and Figure S2 in the Support- changed diffusion properties inside the polymer/enzyme layer, ing Information). for example, owing to swelling and/or slightly changed local To evaluate the performance of the gas-diffusion P(GMA-BA- pH values, which will affect the overall activity of the enzyme. PEGMA)-vio//P(N3MA-BA-GMA)-vio/G491S electrode in an all- In contrast, the current of the biocathode was slightly de- gas-diffusion membrane-free H2/O2 biofuel cell, the bioanode creased after the biofuel evaluation (Figure S7 in the Support- was combined with an O2-reducing biocathode modified with ing Information). This again highlights the high stability of the bilirubin oxidase from Myrothecium verrucaria (Mv-BOD, for bioanode in a membrane-free biofuel cell under anode-limiting comparison purposes because it was used in our previously re- conditions. The operational stability of the biofuel cell was ported experiments[12]). For the immobilization of Mv-BOD, the tested at a constant load of 0.7 V (Figure 5 c). After 10 h of con- carbon cloth was first modified with 2-ABA (2-amino benzoic tinuous operation, 75 % of the initial current density remained. acid) to ensure a proper orientation of the enzyme on the elec- Cyclic voltammograms measured after the long-term experi- trode surface. The modifier was anchored in an electrochemical ment showed significantly lower currents for the bioanode grafting process by applying an oxidative potential pulse.[33] (Figure S8 in the Supporting Information) and the biocathode ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3632 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 (Figure S7 in the Supporting Information). We want to empha- sis and characterization of the less hydrophilic redox polymer poly- size that the amplitudes of the polymer signals (Figure S7 in (glycidyl methacrylate-co-butyl acrylate-co-poly(ethylene glycol) the Supporting Information, dashed black curve) were also de- methacrylate)-vio [P(GMA-BA-PEGMA)-vio] was described previous- ly in Ref. [12]. It was used as an aqueous solution with a concentra- creased compared with the voltammograms measured with tion of 7.5 mg mL@1. the freshly prepared electrode. Thus, not only does deactiva- tion/decomposition of the enzyme contribute to the decreased activity after long-term operation, but the loss of immobiliza- tion matrix may also have an effect. Nevertheless, the bio- Enzymes anode shows an outstanding performance and demonstrates the potential applicability of G941S (enhanced O2 tolerance) as The isolation and purification of the wild-type [NiFeSe] hydroge- a highly active and stable catalyst in a membrane-free biofuel nase from D. vulgaris Hildenborough (wt-[NiFeSe]) was described cell device. Moreover, the proposed H2-oxidation bioanodes previously in Ref. [28]. The activity for H2 uptake was measured as (4850 : 260) s@1.[27] The enzyme was stored in Tris-HCl buffer, combine the advantages of the protection matrix (O2 quench- 20 mm, pH 7.6 at @80 8C at a concentration of 170 mm. The prepa- ing; no high-potential deactivation) and the enhanced enzyme ration of the [NiFeSe] variants G491A and G491S is described in stability of the hydrogenase variants (blocking of O2 access) in Ref. [27]. Their H2 uptake was measured to be (1918 : 119) s@1 accordance with the mechanism depicted in Figure 1 b and (G491A) and (2416 : 387) s@1 (G491S). The [NiFeSe] variants were thus demonstrate a triple-protection system for the high-cur- stored in Tris-HCl, 20 mm, pH 7.6 at @80 8C (G491A: 82.96 mm; rent-density H2-oxidation bioanodes. G491S: 53 mm). Bilirubin oxidase from Bacillus pumilus (Bp-BOD) was isolated and purified according to protocols reported in Ref. [32]. The protein Conclusions was stored in 50 mm borate buffer, pH 9, at @80 8C; concentra- tion = 54.75 mg mL@1; activity = 713 U mg@1. Bilirubin oxidase from The two [NiFeSe] variants show a higher O2 tolerance than the Myrothecium verrucaria (Mv-BOD, lyophilized powder, 15–65 U mg@1 wild type in the immobilized state, which is in line with results protein) was obtained from Sigma–Aldrich and stored at @20 8C as reported for the enzymes operated in a direct electron transfer a powder. For electrode modification, the enzyme was dissolved in regime.[27] In combination with the redox polymer-based pro- phosphate buffer, 0.1 m, pH 7.3, at a concentration of 15 mg mL@1. tection matrix, the proposed bioanodes reveal a triple-protec- tion system that ensures stable operation. Moreover, we could demonstrate that the two [NiFeSe] variants show similar per- formance as the wild type and as [NiFe] as well as [FeFe] hy- Electrochemical experiments drogenases when incorporated into a conventional redox-poly- All electrochemical experiments were conducted under the corre- mer-based biofuel cell. In addition, the use of gas-diffusion sponding atmosphere (argon, hydrogen, oxygen, and their mix- layers ensured high substrate transport towards the active tures) and at room temperature by using a Gamry Reference 600 polymer/enzyme layer, allowing H2-oxidation currents of ap- potentiostat in a three-electrode configuration with an Ag/AgCl/ 3 m KCl reference electrode. All potentials were rescaled to the proximately 6.3 mA cm@2 for the G491S variant at comparative- standard hydrogen electrode (SHE) according to the equation ly low catalyst loadings. Combination of the gas-diffusion bio- ESHE = EAg/AgCl/3 m KCl + 210 mV. Phosphate buffer (0.1 m, pH 7.3) was anode with a gas-diffusion O2-reducing biocathode allowed for used as electrolyte for all experiments. For cyclic voltammetric and the fabrication of a H2/O2-powered biofuel cell with benchmark chronoamperometric experiments, a Pt counter electrode and performance in a membrane-free configuration. We conclude modified glassy carbon disk working electrodes (3 mm) were used. that the novel O2-tolerant [NiFeSe] variants are promising can- The latter were polished by using, first, diamond particles (3 mm) didates for biofuel cell applications, demonstrating that followed by Al2O3 powder (1 mm, then 0.3 mm) following standard enzyme engineering is indeed a powerful tool, which may be protocols. Measurements under gas-diffusion conditions were per- formed in a homemade glass cell[12] with carbon cloth-based gas- used to not only overcome sensitivity issues but also to further diffusion electrodes [MTI, carbon foam sheet, porous C, 0.454 mm enhance the activity and stability of biocatalysts. thick, & 10 mL cm@2 s@1, porosity & 31 mm coated on one side with a conductive Nafion/Teflon-based microporous film (50 mm), carbon content 5 mg cm@2, EQ-bcgdl-1400S-LD]. Thermal mass flow Experimental Section controllers (GFC17, Aalborg Instruments and Controls) were used to adjust the desired atmosphere and gas mixtures with prede- Chemicals and materials fined compositions (for compositions of the gas feed, see the main All chemicals and materials were purchased from Sigma–Aldrich, text and figures). The back of the gas-diffusion electrode was ex- Alfa-Aesar, VWR, Acros-Organics, or Fisher Scientific and were used posed to the corresponding gas atmosphere (bioanode) or to air/ as received (reagent or analytical grade) except where otherwise O2 (biocathode). During the experiments in gas-breathing mode, noted. For the preparation of all aqueous solutions, deionized the electrochemical cell/electrolyte was continuously purged with water from a Millipore water-purification system was used. The an argon stream to prevent permeation of O2 into the bulk electro- synthesis and characterization of the redox polymer poly(3-azido- lyte. For characterization of the biofuel cells, power curves were propyl methacrylate-co-butyl acrylate-co-glycidyl methacrylate)-vio measured by stepped potential chronoamperometric experiments [P(N3MA-BA-GMA)-vio, with vio = (1-(5-hexyn-1-yl)-1’-methyl-4,4’-bi- to minimize contributions from capacitive charging currents. After pyridinium] was described previously in Ref. [10]. It was used as an each potential step, steady-state currents were used to calculate aqueous solution with a concentration of 7.3 mg mL@1. The synthe- the corresponding power values. ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3633 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 Modification of glassy carbon electrodes with hydrogenase/ search and innovation programme (GA 810856). N.M. thanks the polymer films funding of the ANR project BIO3 (ANR-16-CE19-0001-03). All films were prepared by means of a standard drop-casting pro- cess. For this, stock solutions of the polymer P(N3MA-BA-GMA)-vio and the corresponding hydrogenase variant were prepared: 4 mL Conflict of interest of an aqueous P(N3MA-BA-GMA)-vio solution (7.3 mg mL@1) were mixed with 3 mL of phosphate buffer (0.1 m, pH 7.3), and the corre- The authors declare no conflict of interest. sponding hydrogenase was added (wt-[NiFeSe]: 0.5 mL, 170 mm in Tris-HCl buffer, 20 mm, pH 7.6; G491A: 1 mL, 82.96 mm in Tris-HCl Keywords: bioelectrocatalysis · biofuel cells · enzyme buffer; G491S: 1.56 mL, 53 mm in Tris-HCl buffer). For electrode modification, 1.3 mL of the stock solution was drop-cast onto the engineering · hydrogenases · redox polymers 3 mm glassy carbon disk electrode. The modified electrodes were incubated overnight at 4 8C and air dried for 1 h prior to use. In a [1] A. Ruff, F. Conzuelo, W. Schuhmann, Nat. Catal. 2020, 3, 214 – 224. typical experiment, three electrodes were modified from the same [2] H. Chen, F. Dong, S. D. Minteer, Nat. Catal. 2020, 3, 225 – 244. stock solution. [3] N. Mano, A. de Poulpiquet, Chem. Rev. 2018, 118, 2392 – 2468. [4] W. Lubitz, H. Ogata, O. Rediger, E. Reijerse, Chem. Rev. 2014, 114, 4081 – 4148. [5] A. F. Wait, A. Parkin, G. M. Morley, L. dos Santos, F. A. Armstrong, J. Phys. Modification of carbon cloth electrodes with hydrogenase/ Chem. C 2010, 114, 12003 – 12009. polymer films [6] M. Ludwig, J. A. Cracknell, K. A. Vincent, F. A. Armstrong, O. Lenz, J. Biol. Chem. 2009, 284, 465 – 477. First, 20 mL of the polymer P(GMA-BA-PEGMA)-vio (7.5 mg mL@1 in [7] M.-E. Pandelia, V. Fourmond, P. Tron-Infossi, E. Lojou, P. Bertrand, C. water) was drop-cast on the microporous side of the carbon cloth L8ger, M.-T. Giudici-Orticoni, W. Lubitz, J. Am. Chem. Soc. 2010, 132, electrode and dried overnight at room temperature. Subsequently, 6991 – 7004. 20 mL of G491S (53 mm in Tris-HCl, 20 mm, pH 7.6) was mixed with [8] N. Plumer8, O. Rediger, A. A. Oughli, R. Williams, J. Vivekananthan, S. Pçller, W. Schuhmann, W. Lubitz, Nat. Chem. 2014, 6, 822 – 827. 15 mL of P(N3MA-BA-GMA)-vio (7.3 mg mL@1 in water) and drop-cast [9] A. A. Oughli, F. Conzuelo, M. Winkler, T. Happe, W. Lubitz, W. Schuh- onto the already existing polymer spot (diameter of & 4 mm). The mann, O. Rediger, N. Plumer8, Angew. Chem. Int. Ed. 2015, 54, 12329 – electrode was dried overnight at 4 8C. 12333; Angew. Chem. 2015, 127, 12506 – 12510. [10] A. Ruff, J. Szczesny, S. Zacarias, I. A. C. Pereira, N. Plumer8, W. Schuh- mann, ACS Energy Lett. 2017, 2, 964 – 968. Modification of carbon cloth electrodes with Bp-BOD [11] H. Li, D. Buesen, S. Dementin, C. L8ger, V. Fourmond, N. Plumer8, J. Am. Chem. Soc. 2019, 141, 16734 – 16742. The bare carbon cloth was pre-wetted with ethanol on both sides [12] J. Szczesny, N. Marković, F. Conzuelo, S. Zacarias, I. A. C. Pereira, W. and rinsed with water. Then, the microporous side of the gas-diffu- Lubitz, N. Plumer8, W. Schuhmann, A. Ruff, Nat. Commun. 2018, 9, 4715. sion layer was modified with 20 mL of the Bp-BOx solution [13] A. Ruff, S. Janke, J. Szczesny, S. Alsaoub, I. Ruff, W. Lubitz, W. Schuh- mann, ACS Appl. Energy Mater. 2019, 2, 2921 – 2929. (54.75 mg mL@1 in borate buffer, 50 mm, pH 9). The electrode was [14] a) A. Ruff, J. Szczesny, N. Marković, F. Conzuelo, S. Zacarias, I. A. C. Per- dried overnight at 4 8C. eira, W. Lubitz, W. Schuhmann, Nat. Commun. 2018, 9, 3675; b) S. Shirai- wa, K. So, Y. Sugimoto, Y. Kitazumi, O. Shirai, K. Nishikawa, Y. Higuchi, K. Kano, Bioelectrochemistry 2018, 123, 156 – 161. Modification of carbon cloth electrodes with Mv-BOD [15] K. A. Vincent, A. Parkin, F. A. Armstrong, Chem. Rev. 2007, 107, 4366 – 4413. For the preparation of the Mv-BOD-based cathode, the micropo- [16] I. Mazurenko, K. Monsalve, P. Infossi, M.-T. Giudici-Orticoni, F. Topin, N. rous side of the ethanol-treated carbon cloth gas-diffusion elec- Mano, E. Lojou, Energy Environ. Sci. 2017, 10, 1966 – 1982. trode was first modified with 2-amino benzoic acid (2-ABA) in an [17] a) D. Pankratov, F. Conzuelo, P. Pinyou, S. Alsaoub, W. Schuhmann, S. electrochemical grafting process in 0.1 m KCl/5 mm 2-ABA/water by Shleev, Angew. Chem. Int. Ed. 2016, 55, 15434 – 15438; Angew. Chem. 2016, 128, 15660 – 15664; b) F. Conzuelo, N. Marković, A. Ruff, W. Schuh- applying a potential pulse of + 0.8 V vs. Ag/AgCl/3 m KCl for 60 s mann, Angew. Chem. Int. Ed. 2018, 57, 13681 – 13685; Angew. Chem. according to procedures reported in Refs. [12, 33]. The modified 2018, 130, 13870 – 13875. electrode was rinsed with water and further modified with 120 mL [18] S. Alsaoub, F. Conzuelo, S. Gounel, N. Mano, W. Schuhmann, A. Ruff, of an aqueous Mv-BOD solution (10 mg mL@1, nominal enzyme ChemElectroChem 2019, 6, 2080 – 2087. loading: 1.2 mg electrode@1) and dried at 4 8C overnight. [19] J. M. Le, K. L. Bren, ACS Energy Lett. 2019, 4, 2168 – 2180. [20] E. Tremey, C. Stines-Chaumeil, S. Gounel, N. Mano, ChemElectroChem 2017, 4, 2520 – 2526. [21] E. Tremey, E. Suraniti, O. Courjean, S. Gounel, C. Stines-Chaumeil, F. Acknowledgements Louerat, N. Mano, Chem. Commun. 2014, 50, 5912 – 5914. [22] a) G. Caserta, S. Roy, M. Atta, V. Artero, M. Fontecave, Curr. Opin. Chem. This work was financially supported by the Deutsche Forschungs- Biol. 2015, 25, 36 – 47; b) F. Wittkamp, M. Senger, S. T. Stripp, U.-P. Apfel, gemeinschaft (DFG) within the framework of the Cluster of Excel- Chem. Commun. 2018, 54, 5934 – 5942. [23] J. F. Siebel, A. Adamska-Venkatesh, K. Weber, S. Rumpel, E. Reijerse, W. lence RESOLV (EXC-2033; project number 390677874) as well as Lubitz, Biochemistry 2015, 54, 1474 – 1483. by the projects RTI2018-095090-B-100 (Spanish MCIU/AEI and Eu- [24] C. Papini, C. Sommer, L. Pecqueur, D. Pramanik, S. Roy, E. J. Reijerse, F. ropean Union FEDER), a SFRH/BD/116515/2016 fellowship, grant Wittkamp, V. Artero, W. Lubitz, M. Fontecave, ACS Catal. 2019, 9, 4495 – PTDC/BBB-BEP/2885/2014 (Fundażo para a CiÞncia e Tecnologia, 4501. [25] a) J. Esselborn, C. Lambertz, A. Adamska-Venkatesh, T. Simmons, G. Portugal), and the R&D unit LISBOA-01-0145-FEDER-007660 Berggren, J. Noth, J. Siebel, A. Hemschemeier, V. Artero, E. Reijerse, M. (Most-Micro) cofounded by FCT/MCTES and FEDER funds through Fontecave, W. Lubitz, T. Happe, Nat. Chem. Biol. 2013, 9, 607 – 609; b) G. COMPETE2020/POCI, and a European Union’s Horizon 2020 re- Berggren, A. Adamska, C. Lambertz, T. R. Simmons, J. Esselborn, M. Atta, ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3634 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Full Papers ChemSusChem doi.org/10.1002/cssc.202000999 S. Gambarelli, J.-M. Mouesca, E. Reijerse, W. Lubitz, T. Happe, V. Artero, [32] S. Gounel, J. Rouhana, C. Stines-Chaumeil, M. Cadet, N. Mano, J. Biotech- M. Fontecave, Nature 2013, 499, 66 – 69. nol. 2016, 230, 19 – 25. [26] Y. Li, T. B. Rauchfuss, Chem. Rev. 2016, 116, 7043 – 7077. [33] H.-q. Xia, K. So, Y. Kitazumi, O. Shirai, K. Nishikawa, Y. Higuchi, K. Kano, J. [27] S. Zacarias, A. Tempor¼o, M. D. Barrio, V. Fourmond, C. L8ger, P. M. Power Sources 2016, 335, 105 – 112. Matias, I. A. C. Pereira, ACS Catal. 2019, 9, 8509 – 8519. [34] S. Gentil, S. M. Che Mansor, H. Jamet, S. Cosnier, C. Cavazza, A. Le Goff, [28] M. C. Marques, C. Tapia, O. Guti8rrez-Sanz, A. R. Ramos, K. L. Keller, J. D. ACS Catal. 2018, 8, 3957 – 3964. Wall, A. L. de Lacey, P. M. Matias, I. A. C. Pereira, Nat. Chem. Biol. 2017, 13, 544 – 550. [29] A. Parkin, G. Goldet, C. Cavazza, J. C. Fontecilla-Camps, F. A. Armstrong, J. Am. Chem. Soc. 2008, 130, 13410 – 13416. [30] P. Ceccaldi, M. C. Marques, V. Fourmond, I. C. Pereira, C. L8ger, Chem. Commun. 2015, 51, 14223 – 14226. Manuscript received: April 16, 2020 [31] V. Fourmond, S. Stapf, H. Li, D. Buesen, J. Birrell, O. Rediger, W. Lubitz, Revised manuscript received: April 25, 2020 W. Schuhmann, N. Plumer8, C. L8ger, J. Am. Chem. Soc. 2015, 137, Accepted manuscript online: April 27, 2020 5494 – 5505. Version of record online: June 8, 2020 ChemSusChem 2020, 13, 3627 – 3635 www.chemsuschem.org 3635 T 2020 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim