doi.org/10.26434/chemrxiv.13602869.v1

Following Electroenzymatic Hydrogen Production by Rotating Ring DiskElectrochemistry and Mass SpectrometryJaloliddin Khushvakov, Robin Nussbaum, Cécile Cadoux, Jifu Duan, Sven T. Stripp, Ross Milton

Submitted date: 18/01/2021 • Posted date: 19/01/2021Licence: CC BY-NC-ND 4.0Citation information: Khushvakov, Jaloliddin; Nussbaum, Robin; Cadoux, Cécile; Duan, Jifu; Stripp, Sven T.;Milton, Ross (2021): Following Electroenzymatic Hydrogen Production by Rotating Ring Disk Electrochemistryand Mass Spectrometry. ChemRxiv. Preprint. https://doi.org/10.26434/chemrxiv.13602869.v1

We report on two new approaches to study H2-producing metalloenzymes using electrochemistry and massspectrometry, where H+ reduction is driven by hydrogenase within an electrochemically active polymer (redoxpolymer). Researchers have established electrochemical approaches to utilize the H2-processingmetalloenzyme hydrogenase at electrode surfaces. However, it is more-than-often the case that hydrogenaseelectrodes are employed for H2 oxidation. There is significant interest in using renewable electrical energy todrive low-potential reductive reactions such as H2 evolution and N2 fixation, particularly with metalloenzymes.However, much work is required to understand metalloenzymes.The use of rotating ring disk electrochemistrywith hydrogenase is innovative in that it provides a live method to quantify the H2 being produced by theenzyme. This method will be valuable in determining product distributions for such enzymes in real-time, atelectrode surfaces.There is also significant interest in utilizing isotopes of enzymatic substrates whenperforming electrochemistry, since the rate of the reaction corresponds to the current at the electrode.However, researchers of electroenzymatic H+ reduction have yet to utilize online mass spectrometry toanalyze the products of hydron reduction. We report on the ability to follow and differentiate the formation ofH2, HD and D2 in real-time, permitting the calculation of apparent kinetic isotope effects. This approach will bevaluable to characterizing rate-limiting steps involving H+, as well as for other gas-processingmetalloenzymes.

File list (3)

download fileview on ChemRxiv170121 MS file.pdf (1.13 MiB)

download fileview on ChemRxivTOC.png (212.99 KiB)

download fileview on ChemRxiv170121 Supporting information.pdf (1.82 MiB)

Following electroenzymatic hydrogen production by rotating ring 1

disk electrochemistry and mass spectrometry 2

Jaloliddin Khushvakov,a† Robin Nussbaum,a† Cécile Cadoux,a Jifu Duan,b Sven T. Strippc and 3

Ross D. Miltona* 4

a Department of Inorganic and Analytical Chemistry, University of Geneva, Sciences II, Quai 5

Ernest-Ansermet 30, 1211 Geneva 4, Switzerland. 6

b Faculty of Biology and Biotechnology, Photobiotechnology, Ruhr-Universität Bochum, 7

Universitätsstrasse 150, 44801 Bochum, Germany. 8

c Department of Physics, Bioinorganic Chemistry, Freie Universität Berlin, 10623 Berlin, 9

Germany. 10

* Email: ross.milton@unige.ch 11

12

† These authors contributed equally to this work. 13

Abstract 14

Gas-processing metalloenzymes are of interest to new future biotechnologies and bioinspired 15

technologies. Of particular importance are hydrogenases and nitrogenases, which both produce 16

molecular hydrogen (H2) from proton (H+) reduction. Here, we report on the use of rotating 17

ring disk electrochemistry (RRDE) and mass spectrometry (MS) to follow the production of 18

H2 and isotopes produced from deuteron (D+) reduction (HD and D2) using a model hydrogen-19

evolving metalloenzyme [FeFe]-hydrogenase from Clostridium pasteurianum. This facilitates 20

enzymology studies independent of non-innocent chemical reductants. We anticipate that these 21

approaches will be of value in resolving the catalytic mechanisms of H2-producing 22

metalloenzymes and the design of bioinspired catalysts for H2 production and N2 fixation. 23

24

Keywords 25

Enzymatic electrochemistry; [FeFe]; Hydrogen; Hydrogenases; Kinetic isotope effect; 26

Metalloenzymes. 27

Introduction 28

Due to ever-increasing environmental awareness, there is considerable interest to develop 29

efficient electrocatalysts to produce renewable fuels (such as molecular hydrogen, H2) or to 30

supplement/delocalize global industrial processes (such as the production of ammonia fertilizer 31

from molecular nitrogen, N2).1 Gas-processing metalloenzymes are attractive for new 32

electrochemical biotechnologies and provide inspiration for the design of new catalysts due to 33

desirable catalytic properties such as high selectivity, the use of non-precious and abundant 34

metals in their catalytic cores.2 Further, their optimal catalytic activities are often found under 35

mild conditions (ambient temperature and pressure, near-neutral pH). Two particular enzymes 36

of interest are hydrogenases and nitrogenases. Hydrogenases are iron- and sulfur-dependent 37

metalloenzymes that are found in all kingdoms of life, which catalyze the reversible reduction 38

of protons (2H+) to H2 (E0’ = −0.414 V vs. SHE).3 Nitrogenases are also iron- and sulfur-39

dependent metalloenzymes that are found in select archaea and bacteria, which catalyze the 40

fixation of N2 to NH3 (E0’ = +0.274 V vs. SHE),4 as well as the reduction of H+ to H2.5 The 41

production of one equivalent of H2 per N2 reduced is thought to be necessary in order to activate 42

the catalytic cofactor of nitrogenase for N2 fixation.6 Thus, there is considerable interest to 43

understand how nitrogenase evolves H2 for activation as well as how nitrogenase can 44

theoretically divert up to 75% of its electrons toward N2 fixation over H+ reduction in aqueous 45

media, a reaction that plagues abiotic N2−reducing catalytic systems.6–8 46

While hydrogenases and nitrogenases exchange reducing equivalents with small 47

metalloproteins such as ferredoxins or flavodoxins in vivo, electrodes have been employed to 48

drive the artificial reduction of H+ and N2 by these enzymes in vitro.9–13 Such an approach is 49

attractive not only for new biotechnologies, but also for mechanistic interrogation of their 50

complex catalytic mechanisms where electron transfer to these gas-processing metalloenzymes 51

can be controlled. Here, we demonstrate the use of rotating ring disk electrochemistry (RRDE) 52

as a technique to follow the production of H2 from [FeFe]-hydrogenase from Clostridium 53

pasteurianum (CpI) wired to a carbon electrode surface within a redox polymer (Figure 1). 54

Second, we investigated electroenzymatic hydron reduction by CpI in a 50%/50% H/D 55

buffered electrolyte. We demonstrate the use of online mass spectrometry (MS) to follow H2, 56

molecular deuterium (D2) and deuterium hydride (HD) production by an enzyme electrode for 57

the first time, enabling kinetic isotope effect (KIE) studies. These “online” methods provide 58

approaches that can be translated to interrogate other metalloenzymes that produce H2, as well 59

as characterizing catalytic biases for substrate reduction, i.e., N2 vs. H+ reduction by 60

nitrogenase. 61

62

Figure 1. The [FeFe]-hydrogenase from C. pasteurianum (CpI) is immobilized within a cobaltocene-containing redox polymer (Cc-BPEI) at the surface of a rotating carbon electrode. The electroenzymatically generated H2 gas is detected by electrocatalytic oxidation at a rotating Pt ring electrode.

Results 63

While standard [FeFe]-hydrogenases are particularly active in the H+ reduction reaction 64

direction,14 different hydrogenases have previously been reported to undergo H+ reduction 65

when entrapped within a bis(cyclopentadienyl)cobalt(II)-grafted redox polymer at electrode 66

surfaces.15 This “cobaltocene” mediator has a reduction potential (E0) of −0.91 V vs. SHE and 67

is well-suited to facilitate electron transfer to hydrogenase for H+ reduction (E0’ = −0.414 V 68

vs. SHE).16 We employed branched poly(ethylenimine)-grafted cobaltocene as the redox 69

polymer (Cc-BPEI)17 alongside CpI,18 which can be recombinantly expressed in Escherichia 70

coli and activated in vivo and in vitro (Figure S1).18–21 In the first step, we investigated the 71

activity of CpI and integrity of the “H-cluster” active site cofactor by ATR FTIR spectroscopy. 72

The CO and CN– ligands of the H-cluster absorb strongly in the frequency regime between 73

2150 – 1750 cm-1 and do not overlap with the absorbance bands of liquid water and protein. 74

Absolute spectra of hydrated sample (0.5 mM CpI or 0.5 mM CpI + 2 mg/mL Cc-BPEI) suggest 75

no degradation of the H-cluster (Figure S2). To probe the reactivity of the hydrogenase within 76

the polymer, we recorded ATR FTIR difference spectra triggering the reduction of CpI by 77

changing the atmosphere above the sample from 100% N2 to 90% N2 and 10% H2. Figure 2a 78

depicts the decrease of the oxidized state Hox (negative bands) over the increase of one-79

electron reduced states (Hred´ and Hred) and two-electron reduced states (Hhyd and Hsred). 80

This is the typical behavior of pure [FeFe]-hydrogenase at near-neutral pH (Figure S3), as 81

reported earlier.22–24 Following the reduction and auto-oxidation of CpI in time-resolved 82

experiments further confirms the unperturbed activity of enzyme within the Cc-BPEI polymer 83

(Figure 2b and Figure S3). Recent 84

work shows that [FeFe]-hydrogenases 85

can even be reconstituted within redox 86

polymers.25 87

Next, the ability to 88

electrochemically follow H2 produced 89

by a CpI + Cc-BPEI-functionalized 90

electrode was investigated by rotating 91

ring disk electrochemistry (RRDE). 92

While Pt electrodes efficiently reduce 93

H+ and oxidize H2 and RRDE can 94

therefore be employed to follow H2 95

production at Pt ring electrodes,26,27 this 96

technique has not yet been utilized to 97

study mechanisms of H2-producing 98

metalloenzymes. Figure 3a presents a 99

cyclic voltammogram for Cc-mediated 100

H+ reduction by CpI, where a reductive 101

catalytic “wave” commencing at around <−0.4 V vs. SHE was observed representing 102

electroenzymatic H+ reduction to H2 at a glassy carbon (GC) disk electrode. A significant H2 103

oxidative catalytic “wave” was not observed due to (i) the absence of H2 in the glovebox (and 104

electrochemical cell) environment, and (ii) the reductive bias imparted by the low reduction 105

potential of the Cc-BPEI redox polymer. Simultaneously, the neighboring Pt ring electrode was 106

poised at a potential sufficiently positive for electrocatalytic H2 oxidation (i.e., 0 V vs. SHE) 107

(Figure S4). In order to confirm that the disk reductive currents and the ring oxidative currents 108

Figure 2. Integrity of the H-cluster and activity of CpI probed by ATR FTIR spectroscopy. (a) The figure depicts H2 – N2 difference spectra of CpI + 2 mg mL−1 Cc-BPEI. Negative bands represent the H-cluster states that prevail under N2, positive bands are assigned to states accumulated under 10% H2. The observed pattern is virtually identical to pure CpI. (b) Following the time-dependent evolution of states under N2 and 10% H2 further confirms the unperturbed activity of CpI within the Cc-BPEI polymer.

did indeed correspond to 109

electroenzymatic H2 turnover, a control 110

experiment was performed where we 111

exploited the extreme sensitivity of 112

[FeFe]-hydrogenases to O2.28 By taking 113

a CpI-modified electrode and 114

deactivating the enzyme by rotating the 115

electrode in an O2-containing 116

electrolyte solution, the reductive (GC 117

disk) and oxidative (Pt ring) 118

electrocatalytic currents were almost 119

entirely abolished (Figure S5). 120

RRDEs have associated 121

collection efficiencies (CEs) 122

corresponding to the quantity of species 123

produced at the disk that is 124

subsequently detected at the ring; this 125

value is specified to be 24.9% for the 126

RRDE used in this study (further 127

information in the Supporting 128

Information). The CE of this setup was 129

first confirmed using the ferrocyanide/ferricyanide couple and found to be around 27% 130

between rotation rates of 500 – 2500 rpm. For the detection of H2 produced by CpI we observed 131

the CE to reach up to 23% (Figure 3b, Figure S6 and Figure S7). Interestingly, the CE was 132

typically found to decrease with increasing rotation rates (Figure 3b), which was attributed to 133

Figure 3. (a) Cyclic voltammogram for electroenzymatic H2 production (black line) and its subsequent detection by the Pt ring of a RRDE (red line). The scan rate of the CpI-functionalized GC working electrode was 10 mV s−1. The potential of the Pt ring electrode was fixed at 0 V vs. SHE during the experiment, although the current is plotted vs. the potential applied at the GC electrode. The RRDE was rotated at 1000 rpm in an argon-filled glovebox. Electrolyte = 0.5 M phosphate buffer, pH 6.5. (b) A comparison of collection efficiencies obtained for H2 detection (black line) vs. ferricyanide reduction (red line) as a function of rotation rate. Data collected from 2x amperometric i-t experiments. For H2 collection efficiency, the CpI-functionalized disk electrode was poised at −0.8 V and the Pt ring electrode was poised at 0 V vs. SHE. Electrolyte = 0.5 M phosphate buffer, pH 6.5. Ferricyanide generation was performed at a disk electrode poised at +0.8 V and ferricyanide reduction was performed at a Pt ring electrode poised at +0.1 V vs. SHE. Electrolyte = 0.1 M KCl containing 2 mM ferrocyanide.

the relatively low solubility of H2 in 134

aqueous solution and increasingly poor 135

H2 adsorption at high rotation rates. 136

Further, subsequent electrochemical 137

cleaning of the Pt ring surface 138

(Supporting Information) after 139

performing hydrogenase experiments in 140

phosphate buffer revealed a sharp 141

oxidative peak that disappeared after 142

the first scan (Figure S8). We 143

hypothesize that this peak could result 144

from the oxidative stripping of an 145

unknown species that inhibits H2 146

adsorption. Further, high phosphate 147

concentrations were found to 148

significantly impact O2 reduction on Pt 149

electrodes (Figure S9), while the CE of 150

the ferrocyanide/ferricyanide couple by 151

RRDE was not significantly impacted. 152

We next evaluated the KIE for 153

hydron reduction by CpI within the Cc-BPEI redox polymer. Initially, the RRDE approach was 154

employed where a significant decrease in the magnitude of the reductive current (diminished 155

H2 production) was observed in different fractions of D2O-based electrolyte (Figure 4a). In the 156

case where electron transfer to CpI is rate-determining, a change in the electrocatalytic current 157

Figure 4. (a) RRDE cyclic voltammograms for electroenzymatic hydron reduction (black line) with varying ratios of H/D, and subsequent product detection (red line). The scan rate of the CpI-functionalized glassy carbon working electrode was 10 mV s−1. The potential of the Pt ring electrode was fixed at 0 V vs. SHE during the experiment. The RRDE was rotated at 1000 rpm in an argon-filled glovebox. Electrolyte = 0.5 M phosphate buffer, pH/pD = 6.5. (b) Amperometric i-t curve demonstrating the sharp decrease in hydron reductive current upon the introduction of 50% D (final) to the electrolyte. Buffer hydron composition and electrode rotation rates are given above the traces. The glassy carbon working electrode was poised at −0.6 V vs. SHE. Electrolyte = 0.5 M phosphate buffer, pH/pD = 6.5.

(which is proportional to the rate 158

constant) would not be expected and 159

KIE = 𝑖"#$/𝑖"#$ = 1. However, the 160

significant decrease in the rate (𝑘 ∝ 𝑖) 161

of hydron reduction indicates that 162

KIE ≠ 1 and we hypothesized that the 163

rate-determining step for hydron 164

reduction could be associated with (i) 165

one or more hydrons, (ii) product 166

release (i.e., H2/D2/HD)29 and/or (iii) 167

hydron mass transport (either in the 168

bulk, within the redox polymer film or 169

within CpI). Amperometric i-t was 170

subsequently employed to confirm 171

whether bulk mass transport was rate-172

limiting (Figure 4b). We found that 173

the titration of an equivalent D2O-174

based phosphate buffer electrolyte (to 175

a final H:D ratio of 1:1) immediately 176

resulted in a decrease in the magnitude of the reductive current by approximately 37%. 177

Increasing the rotation rate of the electrode from 1000 – 3000 rpm did not significantly recover 178

the reductive catalytic current, indicating that bulk mass-transport was not rate-limiting. 179

Further, the subsequent addition of H2O-based phosphate buffer electrode (with a final H:D 180

ratio of 3:1) resulted in an increase in the magnitude of the reductive current, demonstrating 181

that this effect is reversible. 182

Figure 5. (a) Electrolytic evolution of H2 (black line), HD (red line) and D2 (blue line) following the application of −0.6 V vs. SHE at a CpI-functionalized carbon paper electrode for 3600 s in 1:1 H:D buffer. Each product was followed simultaneously (in the same run) by mass spectrometry (m/z = 2, 3 and 4). Uncalibrated gas quantities are given as baseline corrected changes in ppm. Following electrolysis and in between standard injections, the electrochemical cell was flushed with argon gas (black dashed line) to displace H2/HD/D2 from the reaction headspace. Electrolyte = 0.5 M phosphate buffer, pH/pD = 6.5. The injection of individual or mixed gas standards demonstrates the differentiation of these products in the same experiment. (b) Calibrated quantities of H2 (black), HD (red), D2 (blue) and the sum (magenta) produced during 3600 s of bulk electrolysis (from (a)). The x-axis was corrected to the start of bulk electrolysis.

Upon performing 183

electroenzymatic hydron reduction with 184

CpI in a mixed H/D buffer the expected 185

gaseous products would be H2, D2 and 186

HD. While we have shown that RRDE is 187

effective in observing the H2 produced 188

by enzyme electrodes, differentiating 189

between H2, D2 and HD at a Pt ring 190

electrode was not expected to be facile. 191

To this end, we utilized an online 192

residual gas analyzer to differentiate 193

between H2, D2 and HD. A carbon paper 194

electrode functionalized with CpI and 195

the Cc-BPEI redox polymer was 196

introduced to an electrolyte comprised 197

of 50% deuterated phosphate buffer electrolyte. Bulk electrolysis was performed at −0.6 V vs. 198

SHE while simultaneously and continuously following the formation of products (H2, D2 and 199

HD) in the headspace of the vial with m/z values of 2, 3 and 4 (Figure 5, Figures S10 and 200

S11). As demonstrated in Figure 5a, it was possible to repeatedly flush the electrochemical 201

cell and make subsequent injections of gas standards for calibration, where H2, D2 and HD can 202

either be followed together or individually. After calibration of the cell (Figure 5b), the 203

observed product distribution was found to be approximately 0.44:0.42:0.14 for H2:HD:D2 at 204

steady-state (Figure 6a). Here, we assume that hydrons are sequentially delivered to the H-205

cluster active site of CpI,30 arriving at possible pathways for product formation of: 206

Pathway 1: [FeFe] -𝐇/0⎯2[FeFe]"

-𝐇/0⎯2 [FeFe] + 𝐇𝟐 207

Figure 6. (a) Observed product distribution during electrolysis for 3600 s using a CpI-functionalized carbon paper electrode in 1:1 H:D phosphate buffer, pH 6.5. (b) Calculated kinetic isotope effect (KIE) of the CpI-functionalized electrode, based on the observed fraction of each product in (A).

Pathway 2: [FeFe] -𝐇/0⎯2[FeFe]"

-𝐃/0⎯2 [FeFe] + 𝐇𝐃 208

Pathway 3: [FeFe] -𝐃/0⎯2 [FeFe]6

-𝐇/0⎯2 [FeFe] + 𝐇𝐃 209

Pathway 4: [FeFe] -𝐃/0⎯2 [FeFe]6

-𝐃/0⎯2 [FeFe] +𝐃𝟐 210

211

The probabilities (𝑝) of producing H2, HD (pathways 2 and 3 combined) or D2 can be 212

then expressed as a function of a KIE and the mole fractions (𝑓9), as outlined in the Supporting 213

Information.31 As shown in Figure 6b, the KIEs for the overall CpI electrode as calculated for 214

the observed product distribution are ~2 (for H2), ~2.3 (for HD) and ~1.7 (for D2). Figure S11 215

reports the amperometric i-t trace and observed product distribution for a control CpI carbon 216

paper electrode that was prepared under oxic conditions in order to render the enzyme inactive; 217

significantly diminished catalytic currents (x20 fold) and products (x17 fold) were observed, 218

consistent with active CpI being necessary for the observed production of H2, HD and D2. 219

220

Conclusions 221

We report on the use of RRDE and MS as two different approaches to interrogate H2 production 222

by gas-processing metalloenzymes at electrode surfaces. The [FeFe]-hydrogenase CpI was 223

used as a “model” metalloenzyme and a cobaltocene-based redox polymer was used to 224

immobilize the enzyme and mediate electrons for H+ reduction. The Pt ring electrode of an 225

RRDE was shown to be effective at monitoring real-time H2 production. Further, online MS 226

was employed to follow the production of H2 and isotopes (HD, D2) that are produced by 227

hydron reduction at the gas-processing metalloenzyme electrode in a mixed H/D buffer 228

electrolyte. The combination of these techniques will be important to interrogating 229

metalloenzyme mechanisms, notably independent of non-innocent chemical reductants such as 230

dithionite or europium compounds. Future work will seek to identify the origin of the observed 231

KIE for CpI under these conditions. Moreover, this technique will also be employed to 232

investigate other H2-producing metalloenzymes for which H+ reduction is of mechanistic 233

importance, such as nitrogenase. 234

Acknowledgements 235

We thank Alexandre Jolly and Darren Martin for assistance with mass spectrometry and 236

enzymatic activity assays. We thank the James Swartz (Stanford University) for sharing the 237

E. coli strain used for CpI expression. We thank Thomas Happe for sharing the sample of CpI 238

used for the ATR FTIR experiments. RDM thanks the COMAD of the University of Geneva, 239

the Ernest Boninchi Foundation, the Academic Society of Geneva and the Ernst and Lucie 240

Schmidheiny Foundation for support. STS acknowledges funding by the Deutsche 241

Forschungsgemeinschaft through priority program 1927 (grant agreement 1554/5-1). 242

References 243

(1) Holade, Y.; Servat, K.; Tingry, S.; Napporn, T. W.; Remita, H.; Cornu, D.; Kokoh, K. 244

B. Advances in Electrocatalysis for Energy Conversion and Synthesis of Organic 245

Molecules. ChemPhysChem, 2017, 18 (19), 2573–2605. 246

(2) Jenner, L. P.; Butt, J. N. Electrochemistry of Surface-Confined Enzymes: Inspiration, 247

Insight and Opportunity for Sustainable Biotechnology. Current Opinion in 248

Electrochemistry, 2018, 8, 81–88. 249

(3) Lubitz, W.; Ogata, H.; Rüdiger, O.; Reijerse, E. Hydrogenases. Chem. Rev., 2014, 114 250

(8), 4081–4148. 251

(4) Guo, W.; Zhang, K.; Liang, Z.; Zou, R.; Xu, Q. Electrochemical Nitrogen Fixation and 252

Utilization: Theories, Advanced Catalyst Materials and System Design. Chemical 253

Society Reviews, 2019, 48 (24), 5658–5716. 254

(5) Foster, S. L.; Bakovic, S. I. P.; Duda, R. D.; Maheshwari, S.; Milton, R. D.; Minteer, 255

S. D.; Janik, M. J.; Renner, J. N.; Greenlee, L. F. Catalysts for Nitrogen Reduction to 256

Ammonia. Nature Catalysis, 2018, 1 (7), 490–500. 257

(6) Rohde, M.; Sippel, D.; Trncik, C.; Andrade, S. L. A.; Einsle, O. The Critical E4 State 258

of Nitrogenase Catalysis. Biochemistry, 2018, 57 (38), 5497–5504. 259

(7) Rutledge, H. L.; Tezcan, F. A. Electron Transfer in Nitrogenase. Chemical Reviews, 260

2020, 120 (12), 5158–5193. 261

(8) Seefeldt, L. C.; Hoffman, B. M.; Peters, J. W.; Raugei, S.; Beratan, D. N.; Antony, E.; 262

Dean, D. R. Energy Transduction in Nitrogenase. Accounts of Chemical Research, 263

2018, 51 (9), 2179–2186. 264

(9) Milton, R. D.; Minteer, S. D. Nitrogenase Bioelectrochemistry for Synthesis 265

Applications. Accounts of Chemical Research, 2019, 52 (12), 3351–3360. 266

(10) Poudel, S.; Colman, D. R.; Fixen, K. R.; Ledbetter, R. N.; Zheng, Y.; Pence, N.; 267

Seefeldt, L. C.; Peters, J. W.; Harwood, C. S.; Boyd, E. S. Electron Transfer to 268

Nitrogenase in Different Genomic and Metabolic Backgrounds. Journal of 269

Bacteriology, 2018, 200 (10). 270

(11) Chen, J. G.; Crooks, R. M.; Seefeldt, L. C.; Bren, K. L.; Bullock, R. M.; Darensbourg, 271

M. Y.; Holland, P. L.; Hoffman, B.; Janik, M. J.; Jones, A. K.; Kanatzidis, M. G.; 272

King, P.; Lancaster, K. M.; Lymar, S. V; Pfromm, P.; Schneider, W. F.; Schrock, R. R. 273

Beyond Fossil Fuel–Driven Nitrogen Transformations. Science, 2018, 360 (6391), 274

eaar6611. 275

(12) Gorton, L.; Meredith, M. T.; Minteer, S. D. S. D. D.; Hu, Y. J.; Lee, C. W. C.; Ribbe, 276

M. W.; Hoffman, B. M. B. M.; Lukoyanov, D.; Yang, Z. Y.; Dean, D. R. D. R.; et al. 277

Nitrilotriacetic Acid Degradation under Microbial Fuel Cell Environment. Biosens. 278

Bioelectron., 2010, 5 (1), 241–248. 279

(13) Cadoux, C.; Milton, R. D. Recent Enzymatic Electrochemistry for Reductive 280

Reactions. ChemElectroChem, 2020, 7 (9), 1974–1986. 281

(14) Land, H.; Senger, M.; Berggren, G.; Stripp, S. T. Current State of [FeFe]-Hydrogenase 282

Research: Biodiversity and Spectroscopic Investigations. ACS Catalysis, 2020, 10 283

(13), 7069–7086. 284

(15) Ruth, J. C.; Milton, R. D.; Gu, W.; Spormann, A. M. Enhanced Electrosynthetic 285

Hydrogen Evolution by Hydrogenases Embedded in a Redox-Active Hydrogel. 286

Chemistry - A European Journal, 2020, 26 (32), 7323–7329. 287

(16) Khanova, L. A.; Topolev, V. V.; Krishtalik, L. I. Effect of the Aqueous-Organic 288

Solvent Structure on the Cobalticenium-Cobaltocene Redox Potential: The Redox 289

Couple as a Basis for Determination of the Single Ion Transfer Energies. Chemical 290

Physics, 2006, 326 (1), 33–42. 291

(17) Tapia, C.; Milton, R. D.; Pankratova, G.; Minteer, S. D.; Åkerlund, H. E.; Leech, D.; 292

De Lacey, A. L.; Pita, M.; Gorton, L. Wiring of Photosystem I and Hydrogenase on an 293

Electrode for Photoelectrochemical H2 Production by Using Redox Polymers for 294

Relatively Positive Onset Potential. ChemElectroChem, 2017, 4 (1), 90–95. 295

(18) Kuchenreuther, J. M.; Grady-Smith, C. S.; Bingham, A. S.; George, S. J.; Cramer, S. 296

P.; Swartz, J. R. High-Yield Expression of Heterologous [FeFe] Hydrogenases in 297

Escherichia Coli. PLoS ONE, 2010, 5 (11), e15491. 298

(19) Posewitz, M. C.; King, P. W.; Smolinski, S. L.; Zhang, L.; Seibert, M.; Ghirardi, M. L. 299

Discovery of Two Novel Radical S-Adenosylmethionine Proteins Required for the 300

Assembly of an Active [Fe] Hydrogenase. Journal of Biological Chemistry, 2004, 279 301

(24), 25711–25720. 302

(20) Berggren, G.; Adamska, A.; Lambertz, C.; Simmons, T. R.; Esselborn, J.; Atta, M.; 303

Gambarelli, S.; Mouesca, J. M.; Reijerse, E.; Lubitz, W.; Happe, T.; Artero, V.; 304

Fontecave, M. Biomimetic Assembly and Activation of [FeFe]-Hydrogenases. Nature, 305

2013, 499 (7456), 66–69. 306

(21) Esselborn, J.; Lambertz, C.; Adamska-Venkatesh, A.; Simmons, T.; Berggren, G.; 307

Noth, J.; Siebel, J.; Hemschemeier, A.; Artero, V.; Reijerse, E.; Fontecave, M.; Lubitz, 308

W.; Happe, T. Spontaneous Activation of [FeFe]-Hydrogenases by an Inorganic [2Fe] 309

Active Site Mimic. Nature Chemical Biology, 2013, 9 (10), 607–609. 310

(22) Senger, M.; Mebs, S.; Duan, J.; Wittkamp, F.; Apfel, U. P.; Heberle, J.; Haumann, M.; 311

Stripp, S. T. Stepwise Isotope Editing of [FeFe]-Hydrogenases Exposes Cofactor 312

Dynamics. Proceedings of the National Academy of Sciences of the United States of 313

America, 2016, 113 (30), 8454–8459. 314

(23) Winkler, M.; Senger, M.; Duan, J.; Esselborn, J.; Wittkamp, F.; Hofmann, E.; Apfel, 315

U. P.; Stripp, S. T.; Happe, T. Accumulating the Hydride State in the Catalytic Cycle 316

of [FeFe]-Hydrogenases. Nature Communications, 2017, 8 (1), 1–7. 317

(24) Duan, J.; Mebs, S.; Laun, K.; Wittkamp, F.; Heberle, J.; Happe, T.; Hofmann, E.; 318

Apfel, U. P.; Winkler, M.; Senger, M.; Haumann, M.; Stripp, S. T. Geometry of the 319

Catalytic Active Site in [FeFe]-Hydrogenase Is Determined by Hydrogen Bonding and 320

Proton Transfer. ACS Catalysis, 2019, 9 (10), 9140–9149. 321

(25) Felbek, C.; Hardt, S.; Papini, C.; Pramanik, D.; Artero, V.; Fontecave, M.; Fourmond, 322

V.; Plumere, N.; Léger, C. Artificial Maturation of [FeFe] Hydrogenase in a Redox 323

Polymer Film. Chemical Communications, 2021, accepted (DOI: 324

10.1039/D0CC08168J). 325

(26) Goyal, A.; Marcandalli, G.; Mints, V. A.; Koper, M. T. M. Competition between CO2 326

Reduction and Hydrogen Evolution on a Gold Electrode under Well-Defined Mass 327

Transport Conditions. Journal of the American Chemical Society, 2020, 142 (9), 4154–328

4161. 329

(27) Ahmed, M. E.; Dey, S.; Darensbourg, M. Y.; Dey, A. Oxygen-Tolerant H2 Production 330

by [FeFe]-H2ase Active Site Mimics Aided by Second Sphere Proton Shuttle. Journal 331

of the American Chemical Society, 2018, 140 (39), 12457–12468. 332

(28) Stripp, S. T.; Goldet, G.; Brandmayr, C.; Sanganas, O.; Vincent, K. A.; Haumann, M.; 333

Armstrong, F. A.; Happe, T. How Oxygen Attacks [FeFe] Hydrogenases from 334

Photosynthetic Organisms. Proceedings of the National Academy of Sciences of the 335

United States of America, 2009, 106 (41), 17331–17336. 336

(29) Sanchez, M. L. K.; Sommer, C.; Reijerse, E.; Birrell, J. A.; Lubitz, W.; Dyer, R. B. 337

Investigating the Kinetic Competency of CrHydA1 [FeFe] Hydrogenase Intermediate 338

States via Time-Resolved Infrared Spectroscopy. Journal of the American Chemical 339

Society, 2019, 141 (40), 16064–16070. 340

(30) Senger, M.; Eichmann, V.; Laun, K.; Duan, J.; Wittkamp, F.; Knör, G.; Apfel, U. P.; 341

Happe, T.; Winkler, M.; Heberle, J.; Stripp, S. T. How [FeFe]-Hydrogenase Facilitates 342

Bidirectional Proton Transfer. Journal of the American Chemical Society, 2019, 141 343

(43), 17394–17403. 344

(31) Qiu, Y.; Ren, H.; Edwards, M. A.; Gao, R.; Barman, K.; White, H. S. Electrochemical 345

Generation of Individual Nanobubbles Comprising H2, D2, and HD. Langmuir, 2020, 346

36 (22), 6073–6078. 347

348

download fileview on ChemRxiv170121 MS file.pdf (1.13 MiB)

download fileview on ChemRxivTOC.png (212.99 KiB)

Following electroenzymatic hydrogen production by rotating ring disk 1

electrochemistry and mass spectrometry 2

Jaloliddin Khushvakov,a† Robin Nussbaum,a† Cécile Cadoux,a Jifu Duan,b Sven T. Strippc and 3

Ross D. Miltona* 4

a Department of Inorganic and Analytical Chemistry, University of Geneva, Sciences II, Quai 5

Ernest-Ansermet 30, 1211 Geneva 4, Switzerland. 6

b Faculty of Biology and Biotechnology, Photobiotechnology, Ruhr-Universität Bochum, 7

Universitätsstrasse 150, 44801 Bochum, Germany. 8

c Department of Physics, Freie Universität Berlin, 10623 Berlin, Germany. 9

* Email: ross.milton@unige.ch 10

11

† These authors contributed equally to this work. 12

13

14

Supporting Information 15

Materials and methods 16

General 17

Unless stated otherwise, all chemicals were purchased from Sigma Aldrich, Inc. (Switzerland) 18

and used as received. 1-(2,5-dioxopyrrolidinylcarboxy)-cobaltoceniumhexafluorophosphate 19

(Cc-NHS) was purchased from MCAT (Germany) and used as received. Hydrogen (H2, 20

product = 769088), deuterium hydride (HD, product = 488690) and deuterium (D2, product = 21

361860) were purchased from Sigma Aldrich, Inc. and used for calibration. [FeFe]-22

hydrogenase purification was performed inside of a glovebox filled with ~97:3% N2/H2, at 23

ambient temperature (Coy Laboratory Products Inc., USA). All electrochemical and mass 24

spectrometry experiments were performed in an Ar-filled glovebox (Jacomex, France). All gas 25

and liquid transfers, when not inside of a glovebox, were performed using gas-tight syringes 26

and septum-sealed vials. 27

28

[FeFe]-hydrogenase “CpI” preparation 29

Strep-tagged [FeFe]-hydrogenase “CpI” from Clostridium pasteurianum was recombinantly 30

expressed in Escherichia coli and purified as previously detailed, although a “StrepTactin XT” 31

column was used (Sigma, 1 mL column volume) and CpI was eluted using 50 mM biotin 32

(Figure S1).1,2 Purified enzyme was flash-frozen and stored in liquid nitrogen until use. The 33

specific H2-oxidation activity of CpI was determined to be 17.1 ± 3.3 µmol min−1 mg−1 protein, 34

in pH 9 glycine-NaOH buffer (50 mM) at 28 ºC and under 100% H2 using methylviologen 35

(1 mM) as the electron acceptor (e = 12,700 cm−1 M−1 at 606 nm). 36

37

38

Electrochemistry and mass spectrometry 39

All electrochemical and mass spectrometry experiments were performed inside of an Ar-filled 40

glovebox (Jacomex, France) at ambient temperature. Rotating ring-disk electrochemistry was 41

performed using a Metrohm-Autolab system (product = AUT.RRDE, Switzerland) equipped 42

with a 5 mm glassy carbon disk and a Pt ring electrode (375 µm gap, 24.9% collection 43

efficiency). A Metrohm-Dropsense µStat 400 bipotentiostat/galvanostat was used for the cyclic 44

voltammetry and amperometric experiments. For RRDE experiments, a cell equipped with a 45

Luggin capillary and a fritted counter electrode compartment was purchased from Adams & 46

Chittenden Scientific Glass (product = 957219, USA). A saturated Ag/AgCl electrode was used 47

as the reference electrode and platinum wire was used as the counter electrode. Potentials were 48

applied vs. the saturated Ag/AgCl(satd.) reference electrode and converted to the standard 49

hydrogen electrode (SHE) by: ESHE = EAg/AgCl + 0.197 (V).3 All buffer electrolytes were 50

prepared aerobically and transferred into the anoxic chamber at least 2 days prior to 51

experiments (to allow for O2 exchange with Ar). The GC/Pt RRDE with a 0.05 µm alumina 52

slurry prior to use or functionalization with the hydrogenase/redox polymer mix (below). 53

Figure S1. SDS-PAGE (4-12% acrylamide) of strep-tagged CpI. Lane 1 = protein markers (kDa), lane 2 = purified CpI, ~1 µg of protein.

Immediately before use, the Pt ring electrode was electrochemically cleaned by cyclic 54

voltammetry (in phosphate buffer electrolyte, pH 6.5, 0.5 M) between −0.4 to 1 V vs. 55

Ag/AgCl(satd.) (scan rate = 250 mV s−1, 5 scans, rotation rate = 1000 rpm). Phosphate buffer 56

electrolyte (pH 6.5, 0.5 M) was used as the electrolyte for the electrochemical experiments. For 57

kinetic isotope effect (KIE) experiments requiring deuterated phosphate buffer electrolyte, 58

dibasic potassium phosphate was dissolved in D2O and the pD was adjusted with the addition 59

of sodium deuteroxide (NaOD, Sigma Aldrich). The pD was determined using a standard pH 60

glass electrode and the following correction for the activity of D: pD = “pH” + 0.41.4 Non-61

PTFE-treated Toray carbon paper was purchased from Fisher Scientific (Switzerland). Carbon 62

paper electrodes were prepared by cutting a strip of 0.5 cm × 3 cm Toray carbon paper and 63

wax-coating the center of the strip to yield two exposed ends; one to clip to the potentiostat 64

lead and the other (with an exposed geometric area of ~0.25 cm2) for CpI/Cc-BPEI 65

functionalization (below). While mass transport was controlled by electrode rotation in the 66

RRDE experiments, a stirrer bar and magnetic stirrer plate were used in the case of experiments 67

employing carbon paper electrodes. 68

The production of H2, HD and D2 was followed using a residual gas analyzer (HPR-20 69

R&D, HIDEN Analytical, UK) equipped with an unheated capillary inlet. The inlet was passed 70

through an air-tight feedthrough into the Ar-filled glovebox for product analysis. 71

Electrochemical MS experiments were performed using a custom-built electrochemical cell 72

(Adams & Chittenden, USA) with a “pear-shape” design and a total volume of 58 mL. 73

Phosphate buffer electrolyte (10 mL) was added to the cell and electrodes were connected via 74

gas-tight fittings; butyl stoppers (Chemglass Life Sciences, CLS420914) were cored to house 75

the reference electrode, counter electrode (coiled platinum wire in a Vycor-tipped glass tube, 76

OD = 8 mm) and working electrode (connected to a titanium wire). The inlet tubing of the MS 77

was also fed through a butyl stopper in addition to a vent tube (to counter the vacuum pulled 78

by the MS). The probabilities (𝑝) of producing H2, HD (pathways 2 and 3 combined) or D2 can 79

be expressed as:5 80

𝑝𝐇𝟐 =𝐊𝐈𝐄)+,)

((𝐊𝐈𝐄)+,/+0)) and 𝐊𝐈𝐄 =

+012𝐇𝟐+,3+,12𝐇𝟐

81

𝑝𝐇𝐃 =5(678)+0+,((678)+,/+0))

and 𝐊𝐈𝐄 =93+0)+,)(52𝐇𝐃3:)32𝐇𝐃+0+,/+0+,

+,)2𝐇𝐃 82

𝑝𝐃𝟐 =+0)

((678)+,/+0)) and 𝐊𝐈𝐄 =

+0;<

9=𝐃𝟐3:>

+, 83

where 𝑓@ represents the mole fractions of H and D in the reaction (0.5 for both, when performed 84

in 50% deuterated buffer). 85

86

CpI [FeFe]-hydrogenase electrode preparation 87

Cobaltocene-functionalized branched poly(ethylenimine) (Cc-BPEI) was prepared by 88

dissolving 60 mg of BPEI (average Mn ~60000, 50 wt. % in H2, Sigma product = 181978) 89

alongside 6.3 µmol of Cc-NHS in 200 µL of dimethylsulfoxide, with 19 µmol of N,N,N’,N’-90

tetramethyl-O-(1H-benzotriazol-1-yl)uranium hexafluorophosphate (HBTU) and 11 µmol of 91

N,N-diethyethanamine (triethylamine) and stirring the reaction mixture for 48h at room 92

temperature. After, the 7.5 mL of H2O was added to the reaction mixture and the polymer was 93

concentrated over a 10,000 kDa molecular weight cut-off centrifugal filtration membrane to a 94

volume of ~2 mL. The polymer was dried by lyophilization and strong vacuum, and stored at 95

a concentration of 30 mg mL−1 in milliQ water. 96

CpI/Cc-BPEI electrodes were prepared by mixing 21 µL of Cc-BPEI (stock 97

concentration of 2 mg mL−1) with 6 µL of CpI (stock concentration of 2 mg mL−1) and 3 µL of 98

a fresh 10% v/v solution of poly(ethyleneglycol) diglycidyl ether (Mn = 500, Sigma product = 99

475696) prepared in milliQ water. After mixing, 6 µL of this mixture was applied to a freshly 100

polished GC RRDE electrode (0.196 cm2) or to a carbon paper electrode (geometric area of 101

0.25 cm2) and left to dry for 2 h. Prior to testing, each RRDE electrode was equilibrated by 102

performing cyclic voltammetry in phosphate buffer electrolyte (pH 6.5, 0.5 M) between −0.5 103

to −1.0 V vs. Ag/AgCl(satd.) for 5 cycles with the electrode rotating at 1000 rpm. All 104

electrochemical analyses were conducted within an Ar-filled glovebox. 105

Control electrodes were prepared using deactivated CpI. Since CpI is sensitive to 106

deactivation by O2, control electrodes for Figure S10 were prepared in the same manner as 107

above except the electrodes were prepared outside of the glovebox and left to dry on the bench, 108

before being cycled into the glovebox for analysis. In Figure S5 we also demonstrated that 109

active CpI/Cc-BPEI electrodes can be rapidly deactivated within the Ar-filled glovebox by 110

rotating a prepared RRDE in an oxic phosphate buffer electrolyte for 10 minutes. Subsequent 111

analysis by cyclic voltammetry in an anoxic equivalent phosphate buffer electrolyte revealed a 112

significant decrease in the magnitude of the reductive current confirming CpI deactivation by 113

O2. 114

115

ATR FTIR spectroscopy 116

Equipped with a mercury cadmium telluride (MCT) detector and a triple-reflection ZnSe/Si 117

crystal ATR cell (Smith Detection, USA), the FTIR spectrometer (Tensor27, Bruker, 118

Germany) was placed in an anaerobic chamber (Coy Laboratory Products). Infrared spectra 119

were recorded with 80 kHz scanning velocity at a spectral resolution of 2 cm−1. Under these 120

conditions, the time-resolution of data acquisition is in the range of seconds (i.e., five 121

interferometer scans in forward/backward direction). ATR FTIR measurements were 122

performed at 25°C and on sample films (0.5 mM CpI or 0.5 mM CpI + 2 mg mL−1 Cc-BPEI) 123

derived by controlled dehydration and rehydration as reported earlier.6 A constant gas stream 124

of 1.5 L min-1 was adjusted with digital mass flow controllers (SmartTrak, Sierra, USA) 125

passing through a wash bottle containing 150 mL buffer solution (100 mM Tris/HCl, pH 8). 126

The resulting aerosol was fed to the sample film on the ATR crystal. The oxidized H-cluster 127

state Hox was enriched in the films under a constant stream of N2 aerosol for 30 minutes before 128

10% H2 was added to the N2 stream via a separate flow controller. After 90 s, H2 was removed 129

from the gas stream. 130

131

References 132

(1) Kuchenreuther, J. M.; Grady-Smith, C. S.; Bingham, A. S.; George, S. J.; Cramer, S. 133

P.; Swartz, J. R. High-Yield Expression of Heterologous [FeFe] Hydrogenases in 134

Escherichia Coli. PLoS ONE, 2010, 5 (11), e15491. 135

(2) Ruth, J. C.; Milton, R. D.; Gu, W.; Spormann, A. M. Enhanced Electrosynthetic 136

Hydrogen Evolution by Hydrogenases Embedded in a Redox-Active Hydrogel. 137

Chemistry - A European Journal, 2020, 26 (32), 7323–7329. 138

(3) Bard, A. J.; Faulkner, L. Introduction and Overview to Electrode Processes. In 139

Electrochemical Methods: Fundamentals and Applications; Swain, E., Ed.; John 140

Wiley & Sons: New Jersey, 2001; pp 3–4. 141

(4) Glasoe, P. K.; Long, F. A. Use of Glass Electrodes to Measure Acidities in Deuterium 142

Oxide. Journal of Physical Chemistry. American Chemical Society 1960, pp 188–190. 143

(5) Qiu, Y.; Ren, H.; Edwards, M. A.; Gao, R.; Barman, K.; White, H. S. Electrochemical 144

Generation of Individual Nanobubbles Comprising H2, D2, and HD. Langmuir, 2020, 145

36 (22), 6073–6078. 146

(6) Senger, M.; Mebs, S.; Duan, J.; Wittkamp, F.; Apfel, U. P.; Heberle, J.; Haumann, M.; 147

Stripp, S. T. Stepwise Isotope Editing of [FeFe]-Hydrogenases Exposes Cofactor 148

Dynamics. Proceedings of the National Academy of Sciences of the United States of 149

America, 2016, 113 (30), 8454–8459. 150

151

Supporting Figures

152

Figure S2. Absolute spectra. Overlay of the FTIR spectra of (a) CpI and (b) CpI + Cc-BPEI under N2 (black) and 10% H2 (red). The insets shows the energy regime of the CO/CN ligands of the H-cluster. Blue traces depict H2 – N2 difference spectra (unscaled, compare Figure S3). The hydration level is the same in both samples, but CpI + Cc-BPEI forms a more homogenous film, which causer the slightly higher signal intensity in panel (b).

153

Figure S3. Direct comparison. (a) Overlay of the H2 - N2 difference spectrum of CpI + Cc-BPEI (black/dotted, as discussed in the main script) and pure CpI (red/solid). The spectra are virtually identical and differ only for a slightly stronger decrease of Hox, as depicted in the double difference spectrum (inset, blue). (b) Overlay of the temoral evolution of Hox under N2 and H2 for CpI + Cc-BPEI (black/dotted, as discussed in the main script) and pure CpI (red/solid). Neither reduction (decrease of Hox under H2) nor auto-oxidation (slow increase of Hox under N2) is signifcantly affected.

Figure S4. Cyclic voltammograms shwoing H2 detection on the Pt ring of a RRDE (red lines), in the absence (red dashed line) or presence (red solid line) of electrocatalytic H2 generation at the GC disk electrode. The scan rate at the GC disk electrode was 50 mV s−1 and the RRDE was rotated at 3000 rpm. Electrolyte = anoxic phosphate buffer electrode, 0.5 M, pH 6.5. The black lines report the current of the GC disk electrode when poised at 0 V vs. SHE (black dashed line, minimal H+ reduction) or −1 V vs. SHE (black solid line, electrocatalytic H+ reduction). The inset presents a zoom where the oxidation of electrogenerated H2 at the Pt ring can be observed above potentials of ~ −0.3 V vs. SHE. H+ reduction on the Pt ring is observed with an approximate onset potential of −0.26 V vs. SHE under these conditions.

Figure S5. RRDE Cyclic voltammograms of electroenzymatic H2 production by CpI/Cc-BPEI at a GC disk and coupled H2 detection at a Pt ring electrode (amperometric i-t). The solid lines present the experiments performed with active CpI. After performing this scan, the electrode was rotated in oxic (O2-containing) phosphate buffer electrolyte for 10 minutes (1000 rpm) prior to rinsing and subsequent analysis in the original anoxic phosphate buffer electrode (dashed lines). The decrease in the magnitude of the reductive current at the disk electrode (and the concurent decrease in the current for H2 oxidation at the Pt ring electrode) are consistent with loss of CpI activity upon exposure to O2. GC scan rate = 10 mV s−1. RRDE rotated at 1000 rpm. The Pt ring electrode was poised at 0 V vs. SHE. The experiment was performed inside an Ar-filled glovebox, in phosphate buffer electrolyte (0.5 M, pH 6.5).

Figure S6. (A) RRDE Cyclic voltammogram of ferrocyanide oxidation (GC disk, black line) and subsequent ferricyanide reduction (Pt ring, red line, amperometric i-t, poised at +0.1 V vs. SHE). The scan rate of the GC disk electrode was 50 mV s−1 and the RRDE was rotated at 1000 rpm. (B) Double amperometric i-t traces of ferrocyanide oxidation (GC disk, black trace) and subsequent ferricyanide reduction (Pt ring, red trace), used to determine collection efficiencies of the RRDE. Ferrocyanide oxidation (ferricyanide generation) was performed by applying +0.8 V vs. SHE at the GC disk electrode, whereas ferricyanide detection (reduction) was perfomed by applying +0.1 V vs. SHE at the Pt ring electrode. Electrolyte solution = 0.1 M KCl containing 2 mM ferrocyanide.

Figure S7. Cyclic voltammogram for electroenzymatic H2 production (black line) and its subsequent detection by the Pt ring of a RRDE (red line), as reported in Figure 3 of the main artice. The calculated collection efficiency is overlaid (black dashed line, right y axis) which can be observed to stabilize at around 22% in the H2 generation/detection region of the cyclic voltammogram (black line). The scan rate of the hydrogenase-functionalized glassy carbon working electrode was 10 mV s−1. The potential of the Pt ring electrode was fixed at 0 V vs. SHE during the experiment. The RRDE was rotated at 1000 rpm in an argon-filled glovebox. Electrolyte = 0.5 M phosphate buffer, pH 6.5.

154

Figure S8. Cyclic voltammogram of electrochemical cleaning of the Pt ring RRDE electrode; the first (black line) and second (red line) cycles are shown, where a large oxidative current can be observed in the first scan at around +0.55 V vs. SHE after performing RRDE experiments for H2 detection at the Pt ring (poised at 0 V vs. SHE). The electrode was rotated at 1000 rpm in phosphate buffer electrolyte (0.5 M, pH 6.5). Scan rate = 250 mV s−1.

Figure S9. Cyclic voltammograms indicating the impact of high phosphate buffer electrolyte concentration on O2 electrocatalytic reduction on a Pt electrode. In 0.1 M phosphate buffer electrolyte (black lines), the purging of a stirred N2-saturated solution (black dashed line) with air results in a reductive catalytic wave at potentials more negative than approximately 0.5 V vs. SHE (black solid line). In contrast, performing the same experiment in 1.5 M phosphate buffer electroyte (red lines) significantly decreases the magnitude of the reductive catalytic current upon introducing air purging to a stirred N2-saturated solution (dashed to solid red lines). Scan rate = 50 mV s−1, electrode = 3 mm Pt disk electrode, pH = 7.

Figure S10. Amperometric i-t trace for electroenzymatic H2 generation by a CpI/Cc-BPEI carbon paper electrode. The electrode was poised at −0.6 V vs. SHE and the solution was agitated by gentle stirring with a magnetic stirrer bar. Performed in phosphate buffer electrolyte (0.5 M, pH 6.5). At 750 s, an equal volume of deuterated phopshat buffer electrolyte was introduced, bringing the H:D ratio to 1:1 (pH/pD = 6.5). Increased stirring at 1050 s did not significantly increase the magnitude of the reductive current, indicating that bulk mass transport is not rate-limiting.

Figure S11. (A) Amperometric i-t trace for electroenzymatic H2 generation by active (solid line) or deactivated (dashed line) CpI/Cc-BPEI carbon paper electrodes, during the mass spectrometry experiment performed in Figure 5 of the main article. The working electrode was poised at −0.6 V vs. SHE. The control electrode with deactivated CpI was prepared outside of the Ar-filled glovebox and cycled inside after drying. (B) Calibrated H2 (black), HD (red) and D2 (blue) quantities produced during bulk electrolysis (Figure 5b of the main article), showing the quantities of H2, HD and D2 produced by the deactivated CpI control electrodes (dashed lines). The theoretical quantities of the sum of the gases is shown as black dotted lines, alongside the experimentally determined sums of the gases from each experiment (purple solid and dashed lines). All experiments were performed in phosphate buffer electrolyte (0.5 M, pH 6.5) and the solution was agitated by stirring.

download fileview on ChemRxiv170121 Supporting information.pdf (1.82 MiB)