This article is licensed under CC-BY 4.

pubs.acs.org/JACS Article

Solar-Driven Paired CO2 Reduction−Alcohol Oxidation Using

Semiartificial Suspension, Photocatalyst Sheet, and
Photoelectrochemical Devices
Motiar Rahaman, Carolina Pulignani, Melanie Miller, Subhajit Bhattacharjee,
Ariffin Bin Mohamad Annuar, Rita R. Manuel, Inês A. C. Pereira, and Erwin Reisner*
Cite This: J. Am. Chem. Soc. 2025, 147, 8168−8177 Read Online
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ACCESS Metrics & More Article Recommendations *

sı Supporting Information

ABSTRACT: Fuel-forming enzymes can display excellent per-

formance, achieving high rates of catalysis with unity selectivity at
Downloaded via 102.185.99.26 on March 23, 2025 at 14:05:47 (UTC).

minimal overpotentials, but they are generally considered to be

fragile and difficult to handle in combination with synthetic
semiconductors in light-driven chemical synthesis. Here, we
demonstrate a biohybrid platform that is assembled from
cyanamide-functionalized carbon nitride (CNX) as a scalable and
inexpensive photosensitizer that selectively photo-oxidizes 4-
methyl benzyl alcohol (MBA) to its aldehyde (MBAld), indium
tin oxide (ITO) nanoparticles as electron conduit and biocatalyst
support material, and the enzyme formate dehydrogenase (FDH)
for selective CO2-to-formate reduction. This integrated semi-
artificial multicomponent system can be assembled and used in several configurations to drive bias-free operation, including (i) a
photocatalytic suspension, (ii) a photocatalyst sheet, and (iii) a photoelectrochemical cell. The unprecedented adaptability and
robustness of the assembled biohybrid systems motivated us to select the best performing and scalable system in practical solar
chemical production by deploying a 50 cm2 CNX−ITO|FDH photosheet device for 3 days under natural sunlight to produce 41
(mmol formate) m−2 and 35 (mmol aldehyde) m−2. We have therefore demonstrated that CNX−ITO|FDH provides a robust and
versatile platform that enables solar chemical synthesis for several days in outdoor operation using natural sunlight.

■ INTRODUCTION
Semiartificial assemblies, consisting of a synthetic light
store to release H2 via dehydrogenation or as direct use in
next-generation fuel cells.13−15 The economical synthesis of
absorber and biocatalyst, have shown their promise for driving formate from the greenhouse gas CO2 using solar energy could
selective catalytic transformations efficiently using sunlight.1 thus contribute to establishing sustainable chemical processes.
The use of isolated enzymes in photocatalytic or photo- To source the electrons and protons required for CO2-to-
electrochemical systems benefits especially from their excellent formate reduction, the oxidation of alcohols (e.g., sourced from
selectivity and high rate of catalysis at minor overpotential.2,3 biomass and plastic waste) is a suitable alternative to water
Metal-dependent formate dehydrogenase (FDH), in particular, oxidation as it provides thermodynamic and economic
W-FDH from Desulfovibrio vulgaris Hildenborough used in this advantages.16,17 4-Methyl benzyl alcohol (MBA) is commonly
study, has been established as a model bioelectrocatalyst as it used as a model substrate to demonstrate alcohol oxidation to
reversibly interconverts CO2, protons, and electrons into 4-methyl benzaldehyde (MBAld).18,19
formate when productively wired to an electrode or semi- Carbon nitride is an inexpensive visible-light absorber that
conductor surface.4,5 The surface reaction of immobilized has been used in many photocatalytic reactions, including
FDH proceeds presumably through direct hydride transfer6,7 solar-fuel synthesis coupled to alcohol oxidations.19,20 This
from the metal−thiol active site to the activated CO2 environmentally friendly and scalable photosensitizer is readily
positioned in the second coordination sphere of the metal
center.8,9 Thus, immobilized FDH enables mediator-free CO2/
formate cycling at the thermodynamic potential, which enables Received: August 1, 2024
optimal utilization of low-potential electrons generated by a Revised: January 25, 2025
photoexcited semiconductor.10−12 Accepted: February 14, 2025
Formate is a useful CO2 reduction product as it can be used Published: February 28, 2025
as a commodity chemical for different applications; for
example, downstream (bio)chemical synthesis, as a liquid
© 2025 The Authors. Published by
American Chemical Society https://doi.org/10.1021/jacs.4c10519
8168 J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

Figure 1. Photocatalytic process, device configurations, and material characterization. (a) Scheme showing the CNX−ITO|FDH biohybrid
composite for solar-driven formate and MBAld production. Schematic of (b) top-down photocatalyst suspension setup, (c) single-compartment
photocatalyst sheet setup, and (d) 2-electrode photoelectrochemical setup. (e) SEM, (f) TEM, and (g) HR-TEM analyses of the CNX−ITO (1:3)
composite. (h−j) STEM−EDX mapping of the CNX−ITO (1:3) composite showing a homogeneous distribution of N and Sn over the material:
combined N and Sn (h), N (i), and Sn (j).

prepared from abundant precursors and has a band gap of procedure.19 The ITO nanoparticles (<50 nm nanopowder,
about ∼2.7 eV with a band alignment that allows for fuel surface area 47 m2 g−1) were mixed with the CNX at different
production, including H2 evolution and CO2 reduction in weight ratios (3:1, 1:1, 1:3, and 1:9 for CNX/ITO, keeping the
combination with a suitable cocatalyst.21−23 Recent advances total mass constant) by sonication or grinding followed by a
demonstrated that photoelectrodes with cyanamide-function- thermal annealing treatment under an Ar atmosphere. The
alized carbon nitride (CNX) can be rationally designed with different compositions were characterized using transmission
improved conductivity and charge-transfer efficiencies.19,24,25 electron microscopy (TEM, Figure S1) and UV−vis spectros-
Although metal nanoparticles,26,27 small-molecule cata- copy (Figure S2) analyses. Scanning electron microscopy
lysts,28,29 and enzymes such as hydrogenase30,31 have been (SEM) and TEM analyses of the CNX−ITO (1:3) photo-
used as a cocatalyst with CNX, their assembly with CO2- catalyst composite show a mixture of pseudospherical, square,
reducing enzymes such as FDH remains elusive due to or octahedral particle morphology (Figures 1e,f and S3). High-
difficulties in designing a functional CNX−enzyme interface.32 resolution TEM (HR-TEM) analysis shows the ITO fringes (d
CNX has been identified as a particularly active photocatalyst = 0.505 nm; Figure 1g). Scanning transmission electron
for selective MBA-to-MBAld oxidation,19 with the surface microscopy−energy dispersive X-ray diffraction spectroscopy
oxidation mechanism going through intramolecular proton (STEM−EDX) elemental mapping analysis reveals the
transfer and cation-radical formation.33,34 Both MBA-to- distribution of N (from CNX) and Sn (from ITO) on the
MBAld and CO2-to-formate conversion are 2e−-transfer composite material (Figure 1h−j), which indicates an efficient
processes. connection between CNX and ITO components, crucial for a
Here, we report a semiartificial assembly consisting of FDH compact composite that facilitates electron transfer processes.
immobilized on a CNX−ITO (ITO = indium tin oxide) STEM−EDX analyses with different compositions show clear
composite, where ITO acts as the support material to bind N and Sn distributions in accordance with their concentration
FDH and collects charge from the photoexcited CNX (Figure (Figure S4). Powder X-ray diffraction (PXRD) shows the
1a). The CNX−ITO|FDH biohybrid photoreduces CO2 to presence of graphite C and ITO peaks in the composite
formate while oxidizing MBA to MBAld. The versatility of this materials (Figure S5).
integrated multicomponent system enabled the assembly of General Considerations for Photocatalysis. The CNX−
three different systems that were directly compared for solar- ITO|FDH biohybrid was employed in three different
driven chemical production: (i) a photocatalytic suspension configurations (Figures 1b−d and S6), which are reported in
(Figure 1b), (ii) a photocatalyst sheet (Figure 1c), and (iii) an detail below. In all systems, the photocatalytic reduction of
unassisted two-electrode, two-compartment photoelectro- CO2 to formate is coupled with the oxidation of MBA into its
chemical cell (Figure 1d). The photocatalyst sheet demon- corresponding aldehyde. The band energies of the CNX
strated excellent performance and was thus selected for the photocatalyst (valence band = 2.2 V; conduction band =
assembly of a medium-scale prototype device that we deployed −0.45 V vs the reversible hydrogen electrode, RHE at pH 7)35
under natural sunlight for 3 days. align wih the thermodynamic reduction potential of CO2

■ RESULTS AND DISCUSSION

Synthesis and Characterization of CNX−ITO Compo-
reduction over FDH (−0.01 V vs RHE at pH 7, which also
reflects the onset potential for FDH catalysis)10,36,37 and thus
ensure the overall thermodynamic feasibility of the process in
site. The synthesis of cyanamide-functionalized polymeric the biohybrid assembly under solar irradiation. The ITO
CNX was carried out following a previously reported nanoparticles function as a support for the biocatalyst as well as
8169 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

Figure 2. Performance comparison of CNX−ITO|FDH biohybrid systems. (a) Performance of the CNX−ITO|FDH photocatalytic suspension with
different compositions, showing areal activity toward formate and MBAld production. (b) Composition-dependent activity of different CNX−ITO|
FDH photocatalyst sheets. (c) Effect of different oxide supports on the activity of photosheet biohybrid systems. (d) Electrochemical−
photoelectrochemical characterization of individual redox process: Overlap linear sweep voltammetry (LSV) scans of electrochemical CO2
reduction and photoelectrochemical MBA oxidation processes (scan rate ν = 10 mVs−1) using three-electrode configurations. The cathodic current
(green trace) is shown inverted to illustrate the current crossover with the photoanodic response (blue trace). (e) Photocurrent density of a two-
electrode CNX−ITO||IO−ITO|FDH system under chopped (50 min on, 10 min off), continuous, and no simulated solar-light irradiation under
zero applied bias voltage. (f) Faradaic efficiency of formate and MBAld production from the photoelectrochemical system. Conditions: CO2-
saturated NaHCO3 solution (100 mM) containing KCl (50 mM) and MBA (7.5 mM) at pH 6.7 was used as the reaction medium for both
suspension and photosheet experiments. For the electrochemical LSV analysis under a CO2 reduction process, a CO2-saturated NaHCO3 solution
(100 mM) with KCl (50 mM) at pH 6.7 was used as a catholyte, and for the photoelectrochemical LSV under the MBA oxidation process, the
same reaction medium (pH 6.7) was used as an anolyte in the presence of 20 mM MBA. For the photoelectrochemical experiments, a CO2-
saturated NaHCO3 solution (100 mM) with KCl (50 mM) at pH 6.7 was used as a catholyte and CO2-saturated NaHCO3 (100 mM), KCl (50
mM), and MBA (7.5 mM) at pH 6.7 was used as an anolyte. All reactions were performed under constant stirring and 1 sun (AM1.5G, 100 mW
cm−2)-simulated solar irradiation at 30 °C.

a conductor to transfer electrons from CNX to FDH (Figure solution (100 mM) with KCl (50 mM) at pH 6.7, where the
S7). same solution was used in the photoanode compartment after
To the best of our knowledge, this study reports for the first addition of MBA (7.5 mM). The photoexperiments were
time the use of the same photocatalytic system to assemble a performed under 1 sun irradiation (AM1.5G, 100 mW cm−2)
photocatalyst suspension, a photocatalyst sheet, and a using a solar-light simulator, and the reaction temperature was
photoelectrochemical cell, which shows the possibility to maintained at 30 °C for all three systems. The prototype
interchangeably use materials developed for one configuration scalable photosheet was studied outdoors under anaerobic
in another design. Ultimately, this has been made possible by conditions, where natural sunlight intensity and the local
the robust and versatile biohybrid assembly used here, which temperature were constantly monitored. Formate was
allows to readily couple and decouple the light absorber, quantified by ion-exchange chromatography and MBAld by
electron relay, and catalytic unit. high-performance liquid chromatography analysis. The head-
Despite the robustness in operation, we note that the space of the reactors was also analyzed by gas chromatography
handling and attachment of FDH as well as the reactor (to identify potential gaseous byproducts such as H2).
assembly were exclusively performed inside a glovebox under Photocatalyst Suspension. The photocatalyst suspension
an inert (N2) atmosphere to avoid damage of the enzyme. All was prepared by grinding CNX and ITO nanoparticles (using
light experiments were conducted under an O2-free atmos- different ratios, see below), followed by annealing of the
phere to prevent O2 reduction and damage of components mixture at 250 °C for 1 h under Ar. FDH (40 pmol) was then
from reactive oxygen species. A CO2-saturated NaHCO3 immobilized on the CNX−ITO (4 mg) composite powder
solution (100 mM) containing KCl (50 mM) and MBA (7.5 inside a top-down photoreactor (Figure 1b) and then 2 mL of
mM) at pH 6.7 was used as the reaction medium for the a CO2-saturated reaction solution (see above for composition
photocatalyst suspension, photocatalyst sheet, and scalable of the reaction medium) was added to the photoreactor to
prototype photosheet experiments. For the photoelectrochem- make a suspension. The headspace of the reactor was further
ical experiments, the cathode compartment of the photo- purged with CO2 before the experiment. The suspension was
electrochemical cell contained a CO2-saturated NaHCO3 irradiated by a simulated solar simulator (AM1.5G, irradiation
8170 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

area of ∼2.5 cm2) for 10 h. The CNX−ITO|FDH biohybrid μmolformate cm−2 and 18 ± 0.1 μmolMBAld cm−2 with a specific
composites were active toward photocatalytic CO2-to-formate activity of 26 ± 3.5 mmolformate (g CNX−ITO)−1 and 22.6 ±
reduction, whereas the exclusion control experiments (CNX| 0.2 mmolMBAld (g CNX−ITO)−1 after 10 h AM1.5G irradiation.
FDH without ITO, ITO|FDH without CNX, and CNX−ITO The FDH loading was then varied from 50 to 300 pmol on the
without FDH, Table S1) showed no photocatalytic activity. optimized CNX/ITO (1:3) composite, resulting in a steady
Thus, all three components of the biohybrid material are increase in activity up to 200 pmol whereupon a constant
essential and serve the proposed function, with the CNX acting product formation rate was observed (Figure S10). Thus, 200
as a light absorber and alcohol oxidation catalyst, ITO allowing pmol FDH was identified as the optimal enzyme loading on
for adsorption of FDH and acting as electron relay avoiding the photocatalyst sheets. The optimized CNX−ITO|FDH
the electrostatic repulsion between FDH and CNX, and FDH biohybrid was also assembled on different glass substrates to
performing CO2 reduction. study its catalytic activity. The biohybrid on the ITO-coated
The maximum photocatalytic activity was reached with a glass showed similar activity compared to the FTO-coated
CNX/ITO ratio of 1:3 (Figure 2a), resulting in 0.70 ± 0.06 glass, whereas a lower performance was achieved on frosted
μmolformate cm−2 (normalized to the irradiation area of the top- glass, which is possibly due to mechanical instability of the
down photoreactor) and 0.64 ± 0.11 μmolMBAld cm−2 after 10 CNX−ITO composite film (Figure S11). The FTO-coated
h AM1.5G irradiation. This activity corresponds to a specific glass was chosen as the primary substrate for biohybrid
activity of 0.42 ± 0.04 mmolformate (g of CNX−ITO)−1 and 0.38 fabrication due to its relative cost and robustness.
± 0.07 mmolMBAld (g of CNX−ITO)−1 (Figure S8). A trace The role of conducting ITO in the CNX−ITO composite
amount of MBAld production (<0.1 μmol cm−2) was observed was further investigated by replacing it with insulating and
on pure CNX (in the absence of ITO) due to its charge-storing semiconducting metal-oxide particles. Photocatalyst sheets
capability under light irradiation (Figure 2a).19 The stoichio- made of CNX−MO2|FDH with M = Zr, Si, or Ti using a
metric ratio of formate and MBAld is ∼1:1, as expected from CNX/MO2 ratio of 1:3 were assembled and assessed for their
the two-electron nature of both half-reactions. photocatalytic activity (Figure 2c). Very low activity was
Significant photo-oxidation of accumulated formate (back observed with the insulators ZrO2 and SiO2 (∼1.5 to 2
reaction) over the CNX−ITO composite was ruled out by μmolformate cm−2) after 10 h AM1.5G irradiation, whereas the
performing control experiments using a CNX−ITO|H2ase wide-band gap semiconductor TiO2 also showed low activity
biohybrid (H2ase = NiFeSe-hydrogenase from Desulfovibrio (∼3 ± 0.7 μmolformate cm−2) for CNX−TiO2|FDH. A 405 nm
vulgaris Hildenborough; see the Experimental Section for UV filter was used to prevent light absorption from the TiO2
assembly details). The functionality of CNX−ITO|H2ase was semiconductor. Thus, using conducting ITO in the composite
confirmed in the presence of MBA as an electron donor, with with CNX enables the highest photocatalytic activity of the
substantial amount of H2 being produced (Figure S9). In the CNX−ITO|FDH system, supporting the importance of ITO to
absence of MBA (water) or in the presence of formate, only act as a solid-state electron conduit with CNX,24 in addition to
trace amounts of H2 were produced after 8 h of AM1.5G its ability to strongly bind FDH.40
irradiation. Thus, water and formate are not suitable as The reusability and stability of the CNX−ITO|FDH
effective electron donors and demonstrate that MBA oxidation photocatalyst sheet were next studied by performing five
is dominating over a potential back reaction in the CNX−ITO| cycles with the same photocatalyst sheet, where the CNX−ITO
FDH biohybrid. (1:3) was washed with water and loaded with fresh FDH after
Photocatalyst Sheet. The photocatalytic performance of each 10 h AM1.5G irradiation cycle (Figure S12). The activity
the CNX−ITO|FDH powder system motivated us to prepare a of the second cycle is comparable to the first cycle, but the
photocatalyst sheet. Photocatalyst sheets are emerging as easily activity started to markedly decrease in the third and fourth
scalable systems for efficient solar fuel synthesis, particularly in cycle (ca. 50% and 25% formate production activity in the
the field of overall water splitting.38,39 We intended to develop third and fourth cycle, respectively, compared to first cycle).
a novel biohybrid photocatalyst sheet for combined CO2 The photocatalyst sheet was then ozone cleaned after the
reduction and alcohol oxidation processes. A mixture of fourth cycle to remove any potential residual deposits from
CNX−ITO powder was dispersed in ethanol and dropcast (20 protein degradation. However, the photocatalytic activity did
μL from a 45 mg mL−1 solution) onto a fluorine-doped tin not recover in the fifth cycle, indicating that residual deposits
oxide (FTO)-coated glass substrate (0.5 cm2 active area with are not the main reason for losing activity. Postexperiment
∼3 μm thickness of the CNX−ITO film) followed by annealing characterization of the photocatalyst sheets (after the first 10 h
at 250 °C for 1 h under Ar.24 The gravimetric analysis light irradiation) by SEM showed cracks on the surface and
confirmed that ∼400 μg of the CNX−ITO composite was TEM/STEM-EDX analyses showed agglomeration of the
deposited on the FTO substrate. FDH (200 pmol) was photocatalyst nanostructures, indicating morphological
dropcasted onto the CNX−ITO photocatalyst sheet (0.5 cm2) changes of CNX−ITO to be a possible explanation for the
before it was dipped into the 10 mL of CO2-saturated solution loss in performance (Figures S13 and S14).
(see above for composition) in a single-compartment solar Exclusion control experiments by systematically removing
reactor (Figure 1c). The FDH-decorated photocatalyst sheets one component of the photocatalyst sheet at a time showed
produced formate and MBAld in similar molar quantities in all that all components were required for photocatalytic activity
experiments, consistent with the selective generation of both (Table S1). The origin of formate was confirmed by isotopic
products. labeling experiments, using 13CO2 and NaH13CO3 in the
The composition of the photocatalyst sheet was optimized experiment. The 13C-labeled formate product (H13CO2−) was
by changing the CNX and ITO ratio in the composite material, obtained with a clear doublet peak (JC−H coupling = 188 Hz)
and a composition-dependent activity was observed (Figure at 8.34 ppm in the 1H-NMR spectrum, whereas only a singlet
2b). The maximum photocatalytic activity was obtained again is observed at 8.34 ppm for H12CO2− when using 12CO2 and
with a CNX/ITO ratio of 1:3, which produced 21 ± 3 NaH12CO3 in the standard experiments (Figure S15).
8171 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

Figure 3. Comparison of different biohybrid systems for solar chemical production. Time-dependent areal activity and activity per g of the
photocatalyst composite for (a,d) suspension photocatalyst, (b,e) photocatalyst sheet, and (c,f) 2-electrode photoelectrochemical system. The
schematic of each process is presented at the top of the corresponding result for better understanding. The experiments were performed under
constant stirring and 1 sun (AM1.5G, 100 mW cm−2)-simulated solar irradiation at 30 °C.

Photoelectrochemical Cell. Following the demonstra- RHE with a good fill factor and a steady-state current density
tions of CNX−ITO|FDH in suspension and photocatalyst of approximately 0.45 mA cm−2 (no photocurrent response
sheet assemblies, we developed a photoelectrochemical under dark conditions; Figure S17). A combined plot with the
platform with those components. A CNX−ITO photoanode LSV scans recorded under electrochemical CO2 conversion
(1:3 w/w %, deposited on an FTO-coated glass substrate, 0.5 over the IO−ITO|FDH cathode and photoelectrochemical
cm2 active area, ∼3 μm film thickness) based on a previous MBA oxidation over the CNX−ITO photoanode shows an
preparation method was employed for MBA oxidation.24 An overlap for current crossing at −0.06 V vs RHE with a current
inverse-opal (IO) ITO cathode41,42 previously optimized for density of around 200 μA cm−2 (Figure 2d). This indicates that
enzyme adsorption was prepared by depositing ITO nano- these two processes (CO2 reduction and alcohol oxidation)
particles with polystyrene beads (750 nm diameter) on an can be simultaneously driven on CNX by sunlight irradiation
ITO-coated glass substrate (0.25 cm2 active area, see Methods without any external energy input. This current overlap also
for the detailed preparation procedure). FDH (100 pmol) was provides a meaningful analysis of photocatalyst powder and
then dropcast on the IO−ITO scaffold to assemble an IO− sheet systems, which effectively represent “short-circuited”
ITO|FDH cathode for the CO2 conversion. A two-compart- photoelectrochemical systems at the nanoscale.
ment cell with a proton exchange Nafion117 membrane LSV analysis (ν = 10 mV s−1) with the final photo-
separating the electrodes was used (Figure 1d). electrochemical assembly in a two-electrode configuration
The feasibility of the overall standalone operation of a two- (CNX−ITO photoanode for MBA oxidation and IO−ITO|
electrode cell was first assessed by characterizing the individual FDH cathode for CO2 reduction) under continuous and
photoanode and cathode using a three-electrode setup with a chopped (5s on, 5s off) AM1.5G irradiation shows a catalytic
Pt mesh counter electrode and a Ag/AgCl (saturated NaCl) onset voltage at −0.2 V and an average photocurrent density of
reference electrode. In agreement with previous reports,10,43 around 140 μA cm−2 at zero applied bias voltage (Figure S18).
the IO−ITO|FDH cathode displays an onset of electrocatalytic This indicates that the simulated solar irradiation allows for
formate production at a marginal overpotential (−0.1 V vs unassisted and bias-free operation, and is consistent with the
RHE) with almost 100% selectivity (Figure S16). The LSV three-electrode measurements of the individual (photo)-
with the CNX−ITO photoanode shows efficient catalytic electrodes (Figure 2d). No photocurrent was observed under
activity for MBA oxidation providing an onset of −0.2 V vs dark conditions.
8172 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

Figure 4. Scaled biohybrid photocatalyst sheet under natural sunlight. (a) Photographs of the 50 cm2 CNX−ITO (1:3) photosheet (fabricated on
the FTO substrate, upper panel), and the solar reactor operating under natural sunlight on the rooftop of Yusuf Hamied Department of Chemistry,
University of Cambridge (lower panel). (b) Variation of natural-sunlight intensity and day temperature with time (upper panel) and accumulative
product formation over time (lower panel).

Photoelectrolysis experiments under bias-free conditions Despite the overall lower performance of the suspension
using CNX−ITO||IO−ITO|FDH were carried out for 10 h system, it is a simple platform to perform multiple experiments
under chopped (50 min light on, 10 min off) and continuous in parallel. As the suspension system provides an overall
AM1.5G irradiation (Figure 2e). An average photocurrent comparable trend with the other two immobilized platform
density of 65 ± 15 μA cm−2 was obtained, with a minor setups, it can be easily employed to optimize the biohybrid
decrease in photocurrent density over time that is attributed to photocatalyst sheet and photoelectrodes by varying conditions
slow dissolution of the photocatalyst composite in the and performing control experiments.
electrolyte as well as FDH deactivation.24 After 10 h The photocatalyst sheet shows high activity and emerges as
continuous light irradiation, 10.6 ± 0.1 μmolformate cm−2 was an excellent platform for scale-up and performing one-pot
obtained as a CO2 conversion product in the catholyte and reactions under natural sunlight due to its facile fabrication, in
11.1 ± 1.0 μmolMBAld cm−2 as an oxidation product in the agreement with the ongoing development of practical photo-
anolyte. The average faradaic efficiencies of formate and catalyst sheets for overall water splitting.44 A practical
MBAld under chopped and continuous light irradiation were disadvantage is that the single-compartment reaction setup
94 ± 3% for both products, confirming the charge balance of suffers from difficulties in product separation and possibilities
the overall redox process (Figure 2f). No photocurrent was of back reactions (note that significant back reactions were not
observed when the experiment was performed in the dark. observed in our systems; see the above and Figure S9). The
Comparison between Different Configurations. After two-compartment photoelectrochemical setup overcomes
demonstrating the use of CNX−ITO|FDH in three different these issues by separating the oxidative from the reductive
configurations, we compared their time-dependent perform- reactions by an ion-exchange membrane. However, the
ance (Figure 3). All three configurations show a gradual separation results in cell resistance (∼10 ± 2 Ω) and mass
increase in product production over time, maintaining an transport limitations, which reduces the performance of the
approximate 1:1 formate/MBAld ratio. After 10 h, the (photo)electrodes.
suspension system produced around 0.7 ± 0.1 μmolformate The amount of FDH also plays a critical role in the activity
cm−2 (Figure 3a), the photocatalyst sheet 21 ± 3 μmolformate of biohybrid assemblies (Figure S10) and different amounts
cm−2 (Figure 3b), and the bias-free photoelectrochemical cell were used in the optimized systems studied here. Importantly,
10.6 ± 0.1 μmolformate cm−2 (Figure 3c). Adjusting for the mass the turnover number (TON) of formate production in all three
of the photocatalyst composite (CNX−ITO) results in specific configurations was in the range of 42,000 to 53,000 after 10 h
activities of 0.42 ± 0.04 mmolformate (g CNX−ITO)−1 for the AM1.5G irradiation (Figure S19). This observation suggests
suspension system (Figure 3d), 26.2 ± 3.5 mmolformate (g that enzyme activity may have a prominent role in controlling
CNX−ITO)−1 for the photocatalyst sheet (Figure 3e), and 13.3 the performance of each system.
± 0.17 mmolformate (g CNX−ITO)−1 for the photoelectro- The versatility and robustness of the CNX−ITO|FDH
chemical cell (Figure 3f). The high specific activity of the biohybrid system were demonstrated by successfully construct-
photocatalyst sheet and photoelectrochemical cell is due to the ing three different solar-driven configurations. All three
small amount of the CNX−ITO composite (0.8 mg cm−2) used systems operated efficiently and resulted in clean catalytic
to prepare the thin photocatalyst layer, whereas a compara- reactions by performing selective CO2 reduction and alcohol
tively high amount (4 mg in 2 mL) of the composite was oxidation processes without any side reactions. The ∼1:1 ratio
required for the suspension system. of oxidation and reduction products demonstrates the charge
8173 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

balance of the overall redox process. Each system has its own of solar chemical devices. While the three device configurations
advantages and challenges that should be considered when are currently employed in the development of solar chemistry,
selecting a specific configuration. Considering the performance they also present promising platforms to develop organic
and practical aspects of each setup, we selected the photocatalysis in the future.
photocatalyst sheet system for scale-up and testing outdoors All three systems perform well and produce their products
under natural sunlight. cleanly, thus demonstrating high activity and stability in
Prototype Scale-Up and Outdoor Testing. To demon- generating the products in the expected stoichiometric ratio.
strate the scale-up potential and operation under natural While each setup has its own merits, photocatalyst sheets
sunlight, we developed a 50 cm2 CNX−ITO|FDH photo- display high areal and specific activity and allow for simple
catalyst sheet using the CNX−ITO (1:3 w/w) composite (∼40 fabrication, which made it the most suitable choice for
mg material) on the FTO-coated glass and dropcast FDH (1.6 assembling a 50 cm2 prototype for testing outdoors, the largest
nmol). To accommodate the 50 cm2 photocatalyst sheet, we integrated semiartificial solar fuel system reported to date. Our
designed an airtight and transparent Perspex (poly(methyl demonstration to operate this biohybrid device for selective
methacrylate)) photoreactor. FDH immobilization and reactor CO2 fixation over continuous 3 days starts to question the
setup were performed under an inert (N2) atmosphere and the common assumption that enzymes cannot be deployed at scale
photocatalyst sheet was submerged in the CO2-saturated under real-world conditions for solar fuel production and
solution (see above for detailed composition). The experiment invites more ambitious exploration of their suitability to
was carried out on the rooftop of the Yusuf Hamied produce practical devices.
Department of Chemistry, University of Cambridge, for 3
days, from 8th to 10th October 2022 (Figure 4a). Liquid
aliquots were taken in regular time intervals for product
quantification, and the light intensity and reaction temperature
■ EXPERIMENTAL SECTION
Materials and Chemicals. MBA (≥98% purity), FTO (∼7 Ω
sq−1)-coated glass, ITO (∼12 Ω sq−1)-coated glass, frosted glass
were measured throughout the experiment (Figure 4b, upper
slides, ITO nanoparticles (<50 nm diameter, surface area 47 m2 g−1),
panel). The recorded average outside temperature and light and melamine were purchased from Sigma-Aldrich. DL-Dithiothreitol
intensity were 20 ± 3 °C and 72 ± 22 mW cm−2 on day 1, 21 (DTT, Sigma-Aldrich), tris(hydroxymethyl)aminomethane (Tris,
± 2 °C and 71 ± 16 mW cm−2 on day 2, and 19 ± 3 °C and 69 Sigma-Aldrich), 13C−NaHCO3 (Sigma-Aldrich), 13CO2 (Sigma-
± 19 mW cm−2 on day 3. Aldrich), H2SO4 (Fisher Chemicals), sodium hydroxide (Fisher
The sunlight intensity was on average relatively constant Chemicals), H2SO4 (Fluka), Na2CO3, and KCl and NaHCO3 (Sigma-
throughout the days except for some short isolated cloudy Aldrich) were used as received without any further purification. Milli-
sessions. The time-dependent product analysis showed a Q-grade H2O was used for all of the experiments. W/Sec-FDH from
gradual increase in formate and MBAld products in a 1:1 ratio Desulfovibrio vulgaris Hildenborough and NiFeSe H2ase from
(Figure 4b, lower panel). The 50 cm2 biohybrid photocatalyst Desulfovibrio vulgaris Hildenborough were used as biocatalysts.
Carbon Nitride Synthesis. Unfunctionalized carbon nitride was
sheet accumulated 25 mmolformate m−2 and 26.5 mmolMBAld m−2 synthesized following a previously reported procedure30 by heating
at the end of day 1, 38.5 mmolformate m−2 and 32 mmolMBAld melamine (5 g) at 550 °C for 3 h (ramp rate 1 °C min−1) under air.
m−2 at the end of day 2, and 41 mmolformate m−2 and 35 The obtained yellow powder was ground using a pestle and mortar
mmolMBAld m−2 at the end of day 3, demonstrating that the (yield 50%). CNX was synthesized19 by mixing the unfunctionalized
biohybrid photosheet was catalytically active over 3 days. carbon nitride powder with potassium thiocyanate, in a 1:2 weight
While FDH production is currently expensive, this cost can in ratio, using a pestle and mortar. The mixture was heated for 1 h at 400
principle be drastically reduced in the future by using °C and then at 500 °C for 30 min (ramp rate 30 °C min−1) under an
engineered enzymes with specific surface linkers directly inert atmosphere (Ar). The powder was allowed to cool to room
from cell lysates without isolation and purification. temperature, washed twice with H2O, once with H2O and ethanol
(1:1), and dried overnight at room temperature.
Inductively coupled plasma optical emission spectroscopy
Synthesis of the Particulate CNX−ITO Composite. The CNX−
(ICP-OES) study reveals an increase in the trace of In inside ITO composite was prepared by grinding cyanamide-functionalized
the electrolyte with time, which suggests minor photocatalyst CNX powder and ITO nanoparticles in a mortar and pestle with
dissolution being one potential reason for the gradual decrease different weight ratios (3:1, 1:1, 1:3, and 1:9) and then annealing the
of photocatalytic activity of the biohybrid composite over time mixture in a tube furnace at 250 °C for 1 h under an Ar atmosphere to
(Figure S20). Postcatalysis SEM also shows some cracks and achieve the photocatalyst composite.
agglomeration in the photocatalyst composite (Figure S21), CNX−ITO Deposition on Glass Substrates. Carbon nitride was
similar to that observed with the small-size photocatalyst sheet. deposited onto an FTO-coated glass substrate following a reported
literature procedure,24 by codeposition with commercial ITO
■ CONCLUSIONS
We have systematically developed and studied three different
nanoparticles (diameter <50 nm, surface area 47 m2 g−1). The CNX
and ITO powders were mixed in different ratios (3:1, 1:1, 1:3, 1:9,
and 25 mg in total), dispersed in ethanol (0.55 mL), and sonicated for
biohybrid systems for solar-driven CO2 fixation, demonstrating 30 min. The FTO side of the glass was covered with a Parafilm mask
the versatility of the CNX−ITO|FDH platform that also leaving 0.5 cm2 active area for deposition. Two layers of 10 μL of
enables direct comparison between the three setups. This solution were dropcast on the template. The first layer was allowed to
study thereby shows the possibility to assemble the three most dry in air before dropcasting the second one. For the 50 cm2
prominent solar fuel configurations, powder suspension, photosheet, the deposition area was masked by a Kapton tape.
After the mask was removed, the photosheet/electrode was annealed
photocatalyst sheet, and photoelectrochemical cell, with the
at 250 °C for 1h 30 min under an inert atmosphere (Ar) and allowed
same material components (light absorber, co-catalyst and to cool down to room temperature. The CNX−ITO (1:3) composite
support/electron relay). This demonstration will encourage the was also dropcast on the ITO-coated glass and frosted glass (active
exploration of classical semiconductor powder photocatalysts area: 0.5 cm2) following the same procedure.
in photoelectrochemical systems and vice versa and thereby Preparation of IO−ITO Electrodes. The IO−ITO electrodes
unite approaches and scientists in a more holistic exploration were prepared following a previously established ITO nanoparticle

8174 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

(<50 nm diameter) and polystyrene bead (750 nm diameter, 2.54% E (V vs RHE) = E (V vs Ag/AgCl) + 0.059 V × pH + 0.197 V
w/v suspension in water) coassembly procedure.41 The IO−ITO
electrode had a macropore diameter of 750 nm, a film thickness of 20 at 298 K
μm, and a geometric surface area of 0.25 cm2.
FDH Activation and Dropcast. FDH activation and dropcasting The chronoamperometry experiments were also performed under
were performed under a N2 atmosphere in an anaerobic glovebox chopped (50 min light on, 10 min off), continuous, and no simulated
(MBRAUN). A 40 μM FDH solution in 20 mM Tris buffer at pH 7.6 solar irradiation without external bias voltage. All photoelectrochem-
was added to a freshly prepared 50 mM DTT solution in a 1:1 (v/v) ical experiments were performed at 30 °C.
ratio and the mixture was incubated for 30 min. The solution was then H2ase-Based Biohybrid Assembly and Photocatalysis. H2ase
diluted to the required FDH amount with 20 mM Tris buffer at pH (100 pmol) was dropcast on the CNX−ITO (5 mg, 1:3 wt ratio)
7.6 before dropcasting to the CNX−ITO photocatalyst composites. photocatalyst under a N2 atmosphere inside a glovebox and added to
Material Characterization. A TESCAN MIRA3 FEG-SEM a CO2-saturated NaHCO3 aqueous solution (100 mM) containing
instrument equipped with an Oxford Instruments Aztec Energy X- KCl (50 mM) and MBA (2 mM) or formate (2 mM) at pH 6.7 in a
maxN 80 EDX system was used for the SEM and EDX analyses. The top-down photoreactor. No MBA or formate was added to the
TEM measurements, STEM mappings, and high-resolution point reaction medium when water was used as the electron donor.
EDX analyses were carried out by a Thermo Scientific Talos F200X Simulated sunlight (1 sun, 100 mW cm−2, AM1.5G) was irradiated
G2 TEM (FEI, operating voltage 200 kV, 300 mesh Ni grids were from the top and the experiments were carried out for 8 h at 30 °C.
used for sample preparation). PXRD measurements were performed Product Quantification. High-performance liquid chromatog-
by a Panalytical X’Pert Pro (K alpha Cu radiation) diffractometer raphy (HPLC) and ion-exchange chromatography (IC) were used to
using a 1° min−1 scan rate. ICP-OES measurements were performed quantify the liquid products. HPLC analyses were performed using a
by using a Thermo Scientific iCAP 7400 ICP-OES DUO Waters Breeze system equipped with refractive index (RID-2414) and
spectrometer. diode array UV−vis (λ = 210 and 254 nm) detectors. Identification
Photocatalysis and Photoelectrocatalysis Experiments. and quantification of the MBA oxidation product were carried out
After immobilizing the activated FDH enzyme (40 pmol), the 4 mg using a C18 column at a column temperature of 40 °C. The samples
powder photocatalyst composites (CNX−ITO ratios 1:0, 3:1, 1:1, 1:3, were analyzed in the isocratic flow mode with a flow rate of 0.5 mL
1:9, and 0:1) were suspended in 2 mL CO2-saturated 100 mM min−1 using a 1:1 (v/v) mixture of H2O and MeCN. Formate was
NaHCO3 buffer containing 50 mM KCl and 7.5 mM MBA (pH 6.7) quantified by ion chromatography. The analysis was performed with
in top-down reactors. Simulated sunlight (1 sun, 100 mW cm−2, an 882 Metrohm Compact Ion Chromatograph Plus using 3.2 mM
AM1.5G) was irradiated from the top and the experiments were Na2CO3 and 1 mM NaHCO3 as an eluent. Calibration was conducted
carried out for 10 h at 30 °C. The reactor setups were performed in an with external standards for both processes. The headspace of the
MBraun glovebox under a N2 atmosphere and the photochemical reactors was injected into a Shimadzu GC-2010 Plus gas chromato-
experiments were performed outside the glovebox with airtight sealed graph equipped with a PLOT-MS 5 A Molsieve column for H2
reactors. detection. Negligible amounts or no H2 was detected after the
Single-compartment reactors were used for the experiments with experiments indicating that the FDH selectively performs the CO2-to-
the photocatalyst sheets. 200 pmol FDH was dropcast on a 0.5 cm2 formate conversion.
surface area of the CNX−ITO photosheets (with different CNX−ITO
ratios 1:0, 3:1, 1:1, 1:3, 1:9, and 0:1). For FDH concentration
optimization, 50, 100, 200, and 300 pmol FDH were dropcast on the
CNX−ITO (1:3) photosheet. A CO2-saturated 100 mM NaHCO3
■ ASSOCIATED CONTENT
Data Availability Statement
buffer containing 50 mM KCl and 7.5 mM MBA (pH 6.7) was used The raw data underpinning the findings of this study can be
as the reaction medium (total volume: 10 mL) and the photosheets accessed through the University of Cambridge data repository:
were irradiated from the front. The experiments were carried out https://doi.org/10.17863/CAM.116141.
under simulated solar irradiation (1 sun, 100 mW cm−2, AM1.5G) at
30 °C. For the prototype scalable photosheet experiment, an 8 cm × 8
*
sı Supporting Information

cm × 5 cm airtight Perspex (poly(methyl methacrylate)) photoreactor The Supporting Information is available free of charge at
was designed. The 50 cm2 photocatalyst sheet experiment was https://pubs.acs.org/doi/10.1021/jacs.4c10519.
performed on the rooftop of the Yusuf Hamied Department of
Chemistry for 3 days (8th, 9th, and 10th October, 2022) under natural Material characterization via TEM, UV−vis, SEM,
sunlight. The sunlight intensity and the reactor temperature were STEM−EDX, and XRD analyses; photographs of
monitored from time to time with a certified Newport 1916 R optical different systems; performance optimization varying
power meter and a laboratory thermometer, respectively. The reactors different parameters; photocatalysis with the H2ase
were assembled inside a glovebox under a N2 atmosphere. biohybrid; recyclability tests; electrochemical and photo-
The photoelectrochemical experiments were performed in a 2- electrochemical characterizations; turnover number
compartment reactor where the photoanode and cathode compart- calculation; time-dependent ICP-OES analysis; and
ments were separated by a proton exchange membrane (Nafion117, postcatalysis analyses (PDF)
Fuel Cell Stores). A CNX−ITO (1:3 ratio) photoelectrode was used
as the photoanode for MBA oxidation. An inverse-opal ITO electrode
(0.25 cm2 surface area, deposited on an FTO-coated glass substrate)
was used as a substrate where 100 pmol FDH was dropcast to prepare
the cathode. CO2-saturated 100 mM NaHCO3 with 50 mM KCl (5
■ AUTHOR INFORMATION
Corresponding Author
mL, pH 6.7) was used as a catholyte and CO2-saturated 100 mM Erwin Reisner − Yusuf Hamied Department of Chemistry,
NaHCO3 with 50 mM KCl containing 7.5 mM MBA (15 mL, pH University of Cambridge, Cambridge CB2 1EW, U.K.;
6.7) was used as an anolyte. The reactor setups were prepared in a
orcid.org/0000-0002-7781-1616; Phone: +44-
glovebox under a N2 atmosphere.
The cyclic voltammograms were measured with a scan rate of 10 1223336323; Email: [email protected]
mV s−1 under chopped, continuous, and no solar (1 sun, AM1.5G,
100 mW cm−2) irradiation. The potentials during the 3-electrode Authors
experiments were measured against a Ag/AgCl (saturated NaCl) Motiar Rahaman − Yusuf Hamied Department of Chemistry,
reference electrode (a Pt mesh was used as a counter electrode) and University of Cambridge, Cambridge CB2 1EW, U.K.;
were converted to the RHE scale using the equation orcid.org/0000-0002-8422-0566
8175 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

Carolina Pulignani − Yusuf Hamied Department of (3) Kornienko, N.; Zhang, J. Z.; Sakimoto, K. K.; Yang, P.; Reisner,
Chemistry, University of Cambridge, Cambridge CB2 1EW, E. Interfacing nature’s catalytic machinery with synthetic materials for
U.K. semi-artificial photosynthesis. Nat. Nanotechnol. 2018, 13 (10), 890−
Melanie Miller − Yusuf Hamied Department of Chemistry, 899.
University of Cambridge, Cambridge CB2 1EW, U.K. (4) Alvarez-Malmagro, J.; Oliveira, A. R.; Gutiérrez-Sánchez, C.;
Villajos, B.; Pereira, I. A. C.; Vélez, M.; Pita, M.; De Lacey, A. L.
Subhajit Bhattacharjee − Yusuf Hamied Department of
Bioelectrocatalytic Activity of W-Formate Dehydrogenase Covalently
Chemistry, University of Cambridge, Cambridge CB2 1EW, Immobilized on Functionalized Gold and Graphite Electrodes. ACS
U.K. Appl. Mater. Interfaces 2021, 13 (10), 11891−11900.
Ariffin Bin Mohamad Annuar − Yusuf Hamied Department (5) Oliveira, A. R.; Mota, C.; Vilela-Alves, G.; Manuel, R. R.;
of Chemistry, University of Cambridge, Cambridge CB2 Pedrosa, N.; Fourmond, V.; Klymanska, K.; Léger, C.; Guigliarelli, B.;
1EW, U.K. Romão, M. J.; et al. An allosteric redox switch involved in oxygen
Rita R. Manuel − Instituto de Tecnologia Química e Biológica protection in a CO2 reductase. Nat. Chem. Biol. 2024, 20 (1), 111−
António Xavier (ITQB NOVA), Universidade NOVA de 119.
Lisboa, Oeiras 2780-157, Portugal; orcid.org/0000- (6) Robinson, W. E.; Bassegoda, A.; Reisner, E.; Hirst, J. Oxidation-
0001-8068-2053 State-Dependent Binding Properties of the Active Site in a Mo-
Inês A. C. Pereira − Instituto de Tecnologia Química e Containing Formate Dehydrogenase. J. Am. Chem. Soc. 2017, 139
Biológica António Xavier (ITQB NOVA), Universidade (29), 9927−9936.
NOVA de Lisboa, Oeiras 2780-157, Portugal; orcid.org/ (7) Dong, G.; Ryde, U. Reaction mechanism of formate
0000-0003-3283-4520 dehydrogenase studied by computational methods. J. Biol. Inorg.
Chem. 2018, 23 (8), 1243−1254.
Complete contact information is available at: (8) Robinson, W. E.; Bassegoda, A.; Blaza, J. N.; Reisner, E.; Hirst, J.
https://pubs.acs.org/10.1021/jacs.4c10519 Understanding How the Rate of C−H Bond Cleavage Affects
Formate Oxidation Catalysis by a Mo-Dependent Formate Dehydro-
Notes genase. J. Am. Chem. Soc. 2020, 142 (28), 12226−12236.
The authors declare no competing financial interest. (9) Meneghello, M.; Uzel, A.; Broc, M.; Manuel, R. R.; Magalon, A.;
Léger, C.; Pereira, I. A. C.; Walburger, A.; Fourmond, V.
■ ACKNOWLEDGMENTS
This work was supported by the European Commission with a
Electrochemical Kinetics Support a Second Coordination Sphere
Mechanism in Metal-Based Formate Dehydrogenase. Angew. Chem.,
Int. Ed. 2023, 62 (6), No. e202212224.
Horizon 2020 Marie Skłodowska-Curie Individual European (10) Miller, M.; Robinson, W. E.; Oliveira, A. R.; Heidary, N.;
Fellowship (SolarFUEL, GAN 839763, M.R.), UK Research & Kornienko, N.; Warnan, J.; Pereira, I. A. C.; Reisner, E. Interfacing
Innovation (UKRI, ERC Advanced Grant, EP/X030563/1 to Formate Dehydrogenase with Metal Oxides for the Reversible
E.R.), UK Department of Science, Innovation & Technology Electrocatalysis and Solar-Driven Reduction of Carbon Dioxide.
and the Royal Academy of Engineering Chair in Emerging Angew. Chem., Int. Ed. 2019, 58 (14), 4601−4605.
Technologies programme (CIET-2324-83 to E.R.), the (11) Edwardes Moore, E.; Andrei, V.; Oliveira, A. R.; Coito, A. M.;
European Union’s Horizon 2020 project SOLAR2CHEM Pereira, I. A. C.; Reisner, E. A Semi-artificial Photoelectrochemical
(Marie Skłodowska-Curie Actions with GAN 861151 to C.P. Tandem Leaf with a CO2-to-Formate Efficiency Approaching 1%.
and E.R.), an ERC Consolidator Grant (MatEnSAP, 682833; Angew. Chem., Int. Ed. 2021, 60 (50), 26303−26307.
(12) Yadav, R. K.; Baeg, J.-O.; Oh, G. H.; Park, N.-J.; Kong, K.-j.;
to M.M. and E.R.), Cambridge Trust (HRH Prince of Wales
Kim, J.; Hwang, D. W.; Biswas, S. K. A Photocatalyst−Enzyme
Commonwealth Scholarship to S.B.), the Petronas Education
Coupled Artificial Photosynthesis System for Solar Energy in
Sponsorship Programme for Postgraduate Studies (to Production of Formic Acid from CO2. J. Am. Chem. Soc. 2012, 134
A.B.M.A.), the Fundação para a Ciência e Tecnologia (28), 11455−11461.
(Portugal) through fellowship DFA/BD/7897/2020 (RRM), (13) Yu, X.; Pickup, P. G. Recent advances in direct formic acid fuel
grant PTDC/BII-BBF/2050/2020, MOSTMICRO-ITQB cells (DFAFC). J. Power Sources 2008, 182 (1), 124−132.
Unit (UIDB/04612/2020 and UIDP/04612/2020), and (14) Grasemann, M.; Laurenczy, G. Formic acid as a hydrogen
LS4FUTURE Associate Laboratory (LA/P/0087/2020). We source − recent developments and future trends. Energy Environ. Sci.
acknowledge funding for TEM (EPSRC Underpinning Multi- 2012, 5 (8), 8171−8181.
User Equipment Call, EP/P030467/1) and thank Dr Heather (15) Singh, A. K.; Singh, S.; Kumar, A. Hydrogen energy future with
Greer for assistance with the TEM analysis. Ana Margarida formic acid: a renewable chemical hydrogen storage system. Catal. Sci.
Coito is acknowledged for the isolation and purification of a Technol. 2016, 6 (1), 12−40.
sample of hydrogenase. We also thank Matthew Woolley from (16) Bhattacharjee, S.; Rahaman, M.; Andrei, V.; Miller, M.;
the Cambridge Chemistry Workshop for helping with the Rodríguez-Jiménez, S.; Lam, E.; Pornrungroj, C.; Reisner, E.
Photoelectrochemical CO2-to-fuel conversion with simultaneous
prototype photoreactor design for outdoor experiments. We
plastic reforming. Nat. Synth. 2023, 2 (2), 182−192.
are grateful to Dr Tessel Bowens and Dr Yongpeng Liu (17) Bhattacharjee, S.; Linley, S.; Reisner, E. Solar reforming as an
(University of Cambridge) for their valuable feedback on the emerging technology for circular chemical industries. Nat. Rev. Chem
manuscript. 2024, 8 (2), 87−105.

■ REFERENCES
(1) Fang, X.; Kalathil, S.; Reisner, E. Semi-biological approaches to
(18) Behrouzi, L.; Bagheri, R.; Mohammadi, M. R.; Song, Z.;
Chernev, P.; Dau, H.; Najafpour, M. M.; Kaboudin, B. Electro-
chemical alcohols oxidation mediated by N-hydroxyphthalimide on
solar-to-chemical conversion. Chem. Soc. Rev. 2020, 49 (14), 4926− nickel foam surface. Sci. Rep. 2020, 10 (1), 19378.
4952. (19) Kasap, H.; Caputo, C. A.; Martindale, B. C. M.; Godin, R.; Lau,
(2) Evans, R. M.; Siritanaratkul, B.; Megarity, C. F.; Pandey, K.; V. W.-h.; Lotsch, B. V.; Durrant, J. R.; Reisner, E. Solar-Driven
Esterle, T. F.; Badiani, S.; Armstrong, F. A. The value of enzymes in Reduction of Aqueous Protons Coupled to Selective Alcohol
solar fuels research−efficient electrocatalysts through evolution. Oxidation with a Carbon Nitride−Molecular Ni Catalyst System. J.
Chem. Soc. Rev. 2019, 48 (7), 2039−2052. Am. Chem. Soc. 2016, 138 (29), 9183−9192.

8176 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177
Journal of the American Chemical Society pubs.acs.org/JACS Article

(20) Kasap, H.; Achilleos, D. S.; Huang, A.; Reisner, E. Containing Formate Dehydrogenase. J. Am. Chem. Soc. 2014, 136
Photoreforming of Lignocellulose into H2 Using Nanoengineered (44), 15473−15476.
Carbon Nitride under Benign Conditions. J. Am. Chem. Soc. 2018, (38) Wang, Q.; Hisatomi, T.; Jia, Q.; Tokudome, H.; Zhong, M.;
140 (37), 11604−11607. Wang, C.; Pan, Z.; Takata, T.; Nakabayashi, M.; Shibata, N.; et al.
(21) Zhao, D.; Wang, Y.; Dong, C.-L.; Huang, Y.-C.; Chen, J.; Xue, Scalable water splitting on particulate photocatalyst sheets with a
F.; Shen, S.; Guo, L. Boron-doped nitrogen-deficient carbon nitride- solar-to-hydrogen energy conversion efficiency exceeding 1%. Nat.
based Z-scheme heterostructures for photocatalytic overall water Mater. 2016, 15 (6), 611−615.
splitting. Nat. Energy 2021, 6 (4), 388−397. (39) Goto, Y.; Hisatomi, T.; Wang, Q.; Higashi, T.; Ishikiriyama, K.;
(22) Wang, G.; Chen, Z.; Wang, T.; Wang, D.; Mao, J. P and Cu Maeda, T.; Sakata, Y.; Okunaka, S.; Tokudome, H.; Katayama, M.;
Dual Sites on Graphitic Carbon Nitride for Photocatalytic CO2 et al. A Particulate Photocatalyst Water-Splitting Panel for Large-Scale
Reduction to Hydrocarbon Fuels with High C2H6 Evolution. Solar Hydrogen Generation. Joule 2018, 2 (3), 509−520.
Angew. Chem., Int. Ed. 2022, 61 (40), No. e202210789. (40) Edwardes Moore, E.; Cobb, S. J.; Coito, A. M.; Oliveira, A. R.;
(23) Teng, Z.; Zhang, Q.; Yang, H.; Kato, K.; Yang, W.; Lu, Y.-R.; Pereira, I. A. C.; Reisner, E. Understanding the local chemical
Liu, S.; Wang, C.; Yamakata, A.; Su, C.; et al. Atomically dispersed environment of bioelectrocatalysis. Proc. Natl. Acad. Sci. U.S.A. 2022,
antimony on carbon nitride for the artificial photosynthesis of 119 (4), No. e2114097119.
hydrogen peroxide. Nat. Catal. 2021, 4 (5), 374−384. (41) Mersch, D.; Lee, C.-Y.; Zhang, J. Z.; Brinkert, K.; Fontecilla-
(24) Pulignani, C.; Mesa, C. A.; Hillman, S. A. J.; Uekert, T.; Camps, J. C.; Rutherford, A. W.; Reisner, E. Wiring of Photosystem II
Giménez, S.; Durrant, J. R.; Reisner, E. Rational Design of Carbon to Hydrogenase for Photoelectrochemical Water Splitting. J. Am.
Nitride Photoelectrodes with High Activity Toward Organic Chem. Soc. 2015, 137 (26), 8541−8549.
(42) Sokol, K. P.; Mersch, D.; Hartmann, V.; Zhang, J. Z.;
Oxidations. Angew. Chem., Int. Ed. 2022, 61 (50), No. e202211587.
Nowaczyk, M. M.; Rögner, M.; Ruff, A.; Schuhmann, W.; Plumeré,
(25) Yang, W.; Godin, R.; Kasap, H.; Moss, B.; Dong, Y.; Hillman, S.
N.; Reisner, E. Rational wiring of photosystem II to hierarchical
A. J.; Steier, L.; Reisner, E.; Durrant, J. R. Electron Accumulation
indium tin oxide electrodes using redox polymers. Energy Environ. Sci.
Induces Efficiency Bottleneck for Hydrogen Production in Carbon
2016, 9 (12), 3698−3709.
Nitride Photocatalysts. J. Am. Chem. Soc. 2019, 141 (28), 11219− (43) Sokol, K. P.; Robinson, W. E.; Oliveira, A. R.; Warnan, J.;
11229. Nowaczyk, M. M.; Ruff, A.; Pereira, I. A. C.; Reisner, E.
(26) Gao, G.; Jiao, Y.; Waclawik, E. R.; Du, A. Single Atom (Pd/Pt) Photoreduction of CO2 with a Formate Dehydrogenase Driven by
Supported on Graphitic Carbon Nitride as an Efficient Photocatalyst Photosystem II Using a Semi-artificial Z-Scheme Architecture. J. Am.
for Visible-Light Reduction of Carbon Dioxide. J. Am. Chem. Soc. Chem. Soc. 2018, 140 (48), 16418−16422.
2016, 138 (19), 6292−6297. (44) Nishiyama, H.; Yamada, T.; Nakabayashi, M.; Maehara, Y.;
(27) Pang, Y.; Li, P.; Ma, X.; Sun, L.; Liu, Y.; Qu, D.; An, L.; Sun, Z. Yamaguchi, M.; Kuromiya, Y.; Nagatsuma, Y.; Tokudome, H.;
Metal-doped carbon nitride: an all-in-one photocatalyst. EES Catal Akiyama, S.; Watanabe, T.; et al. Photocatalytic solar hydrogen
2023, 1 (6), 810−831. production from water on a 100-m2 scale. Nature 2021, 598 (7880),
(28) Roy, S.; Reisner, E. Visible-Light-Driven CO2 Reduction by 304−307.
Mesoporous Carbon Nitride Modified with Polymeric Cobalt
Phthalocyanine. Angew. Chem., Int. Ed. 2019, 58 (35), 12180−12184.
(29) Ma, B.; Chen, G.; Fave, C.; Chen, L.; Kuriki, R.; Maeda, K.;
Ishitani, O.; Lau, T.-C.; Bonin, J.; Robert, M. Efficient Visible-Light-
■ NOTE ADDED AFTER ASAP PUBLICATION
This paper was published ASAP on February 28, 2025, with
Driven CO2 Reduction by a Cobalt Molecular Catalyst Covalently the wrong version of the Supporting Information. The
Linked to Mesoporous Carbon Nitride. J. Am. Chem. Soc. 2020, 142 corrected version was reposted on March 3, 2025.
(13), 6188−6195.
(30) Caputo, C. A.; Wang, L.; Beranek, R.; Reisner, E. Carbon
nitride−TiO2 hybrid modified with hydrogenase for visible light
driven hydrogen production. Chem. Sci. 2015, 6 (10), 5690−5694.
(31) Liu, Y.; Pulignani, C.; Webb, S.; Cobb, S. J.; Rodríguez-
Jiménez, S.; Kim, D.; Milton, R. D.; Reisner, E. Electrostatic [FeFe]-
hydrogenase−carbon nitride assemblies for efficient solar hydrogen
production. Chem. Sci. 2024, 15 (16), 6088−6094.
(32) Badiani, V. M.; Casadevall, C.; Miller, M.; Cobb, S. J.; Manuel,
R. R.; Pereira, I. A. C.; Reisner, E. Engineering Electro- and
Photocatalytic Carbon Materials for CO2 Reduction by Formate
Dehydrogenase. J. Am. Chem. Soc. 2022, 144 (31), 14207−14216.
(33) García-López, E. I.; Abbasi, Z.; Di Franco, F.; Santamaria, M.;
Marcì, G.; Palmisano, L. Selective oxidation of aromatic alcohols in
the presence of C3N4 photocatalysts derived from the polycondensa-
tion of melamine, cyanuric and barbituric acids. Res. Chem.
Intermediat. 2021, 47 (1), 131−156.
(34) Li, Q.; Jiao, Y.; Tang, Y.; Zhou, J.; Wu, B.; Jiang, B.; Fu, H.
Shear Stress Triggers Ultrathin-Nanosheet Carbon Nitride Assembly
for Photocatalytic H2O2 Production Coupled with Selective Alcohol
Oxidation. J. Am. Chem. Soc. 2023, 145 (38), 20837−20848.
(35) Qin, Y.; Martindale, B. C. M.; Sun, R.; Rieth, A. J.; Nocera, D.
G. Solar-driven tandem photoredox nickel-catalysed cross-coupling
using modified carbon nitride. Chem. Sci. 2020, 11 (28), 7456−7461.
(36) Reda, T.; Plugge, C. M.; Abram, N. J.; Hirst, J. Reversible
interconversion of carbon dioxide and formate by an electroactive
enzyme. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (31), 10654−10658.
(37) Bassegoda, A.; Madden, C.; Wakerley, D. W.; Reisner, E.; Hirst,
J. Reversible Interconversion of CO2 and Formate by a Molybdenum-

8177 https://doi.org/10.1021/jacs.4c10519
J. Am. Chem. Soc. 2025, 147, 8168−8177