A bioinspired molybdenum-copper molecular catalyst for CO2electroreduction

Article abstract of DOI:10.1039/d0sc01045f

Non-noble metal molecular catalysts mediating the electrocatalytic reduction of carbon dioxide are still scarce. This work reports the electrochemical reduction of CO₂to formate catalyzed by the bimetallic complex [(bdt)Mo^VI(O)S₂Cu^ICN]^2?(bdt = benzenedithiolate), a mimic of the active site of the Mo-Cu carbon monoxide dehydrogenase enzyme (CODH2). Infrared spectroelectrochemical (IR-SEC) studies coupled with density functional theory (DFT) computations revealed that the complex is only a pre-catalyst, the active catalyst being generated upon reduction in the presence of CO₂. We found that the two-electron reduction of [(bdt)Mo^VI(O)S₂Cu^ICN]^2?triggers the transfer of the oxo moiety to CO₂forming CO₃^2?and the complex [(bdt)Mo^IVS₂Cu^ICN]^2?and that a further one-electron reduction is needed to generate the active catalyst. Its protonation yields a reactive Mo^VH hydride intermediate which reacts with CO₂to produce formate. These findings are particularly relevant to the design of catalysts from metal oxo precursors.

Full text of DOI:10.1039/d0sc01045f

Chemical
Science
View Article Online
View Journal
EDGE ARTICLE
A bioinspired molybdenum–copper molecular
catalyst for CO2 electroreduction†
Cite this: DOI: 10.1039/d0sc01045f
a
ab
a
Ahmed Mouchfiq,a Tanya K. Todorova,
Subal Dey,
Marc Fontecave
*
All publication charges for this article
have been paid for by the Royal Society
of Chemistry
b
*
and Victor Mougel
Non-noble metal molecular catalysts mediating the electrocatalytic reduction of carbon dioxide are still
scarce. This work reports the electrochemical reduction of CO2 to formate catalyzed by the bimetallic
complex [(bdt)MoVI(O)S2CuICN]2ꢀ (bdt ¼ benzenedithiolate), a mimic of the active site of the Mo–Cu
carbon monoxide dehydrogenase enzyme (CODH2). Infrared spectroelectrochemical (IR-SEC) studies
coupled with density functional theory (DFT) computations revealed that the complex is only a pre-
catalyst, the active catalyst being generated upon reduction in the presence of CO2. We found that the
two-electron reduction of [(bdt)MoVI(O)S2CuICN]2ꢀ triggers the transfer of the oxo moiety to CO2
Received 21st February 2020
Accepted 8th May 2020
2ꢀ
forming CO3
and the complex [(bdt)MoIVS2CuICN]2ꢀ and that a further one-electron reduction is
needed to generate the active catalyst. Its protonation yields a reactive MoVH hydride intermediate
which reacts with CO2 to produce formate. These findings are particularly relevant to the design of
catalysts from metal oxo precursors.
DOI: 10.1039/d0sc01045f
rsc.li/chemical-science
23
such as CoCp(diphosphine)21 and Fe4N(CO)12
generate formic acid as the major product.
complexes
Introduction
Enzymatic active sites constitute a great source of inspiration
to expand this family of non-noble metal catalysts.24 Metal
complexes replicating features of the active sites of metal-
loenzymes promoting CO2 transformation can be conceived in
a so-called bioinspired approach. In that context, we had
synthesized a series of dithiolene ligands mimicking the
molybdopterin cofactor (MPT, shown in Fig. 1) in formate
dehydrogenases (FDH) enzymes chelating the metal ion (Mo,
W) in a bis(dithiolene) complex with its two sulfur atoms.25–29
One of the corresponding Mo complexes was shown to be the
rst example of a FDH mimic displaying catalytic activity for the
conversion of CO2 into formic acid under photochemical
conditions.25 The Ni analogue with the same dithiolene ligand
also proved active for CO2 electroreduction to formic acid,
showing high faradaic yields.30,31 These complexes, in which the
Ni ion is coordinated by S atoms exclusively, are also reminis-
cent of the Ni active site of the Ni–Fe CO-dehydrogenase
(CODH1), catalyzing the CO2-to-CO interconversion.32
The development of efficient and selective catalysts for the
electrochemical reduction of carbon dioxide is a key step to the
sustainable production of carbon-based chemicals. Molecular
catalysts mediating the two-electron/two-proton reduction of
CO2 to carbon monoxide (CO) or formic acid (HCOOH) with
high activity and selectivity have been identied.1–3 Their
molecular nature enables providing key structural and mecha-
nistic information to rationally design more complex hetero-
geneous catalysts.4–13 However, efficient molecular catalysts for
CO2 reduction based on non-noble metals are still scarce.2,3
Among them, Ni-cyclams,14,15 Co phthalocyanine and Fe-
porphyrins,16,17 have been discovered decades ago and only
a
few new members were recently reported, namely
20
Mn(bpy)(CO)3Br,18 Fe-quaterpyridine,19 Fe4N(CO)12
and Co
diphosphine complexes.21,22 Interestingly, the great majority of
these catalysts perform CO2 reduction to CO, and only very few,
a
`
Laboratoire de Chimie des Processus Biologiques, UMR 8229 CNRS, College de
`
France, Sorbonne Universite, 11 Place Marcelin Berthelot, 75231 Paris Cedex 05,
France. E-mail: marc.fontecave@college-de-france.fr
b
¨
Department of Chemistry and Applied Biosciences, ETH Zurich, Vladimir Prelog Weg
¨
1, CH-8093 Zurich, Switzerland. E-mail: mougel@inorg.chem.ethz.ch
† Electronic supplementary information (ESI) available: UV-Vis, NMR and IR
spectra of 1; electrochemical characterizations of 1, along with catalytic studies;
SEM images and EDS spectra of GC electrodes before and aer CPE of 1;
electrochemical characterizations of [Mo(O)(bdt)2][Et4N]2, along with catalytic
studies; structural parameters, IR frequencies, and frontier molecular orbitals
Fig. 1 Active site of CODH2 from O. carboxidovorans (left), molyb-
dopterin (MPT, centre) and structure of [(bdt)Mo(O)S2CuCN]2ꢀ
,
of
1
and intermediates; xyz coordinates of all intermediates. See DOI:
complex 1 (right) (MoCu–CODH backbone is highlighted in red in
complex 1).
10.1039/d0sc01045f
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Herein, we report the investigation of the electrocatalytic chronoamperometry in a stationary regime at a carbon micro-
properties of a complex mimicking the active site of the second electrode shows that this process corresponds to a two-electron
class of CO-dehydrogenase (CODH2) (Fig. 1). As opposed to reduction of complex 1. We tentatively assigned it to a MoVI/
CODH1 which functions reversibly, CODH2 has so far been MoIV process according to eqn (1).
shown to catalyze only CO oxidation to CO2 and not the reverse
[(bdt)Mo(O)S2CuCN]2ꢀ + 2eꢀ ¼ [(bdt)Mo(O)S2CuCN]4ꢀ (1)
reaction, but reasons for this irreversibility are still unclear.24
Structural characterization of CODH2 from Oligotropha car-
boxidovorans has revealed a heterobimetallic Mo–Cu active site
in which the two ions are bridged by a sulde ion, the Mo ion
also being coordinated by the MPT ligand and an oxo/hydroxo
moiety (Fig. 1).33 This metal cofactor has led to the synthesis
Finally, the peak current displayed a linear relation to the
square root of the scan rate indicating that the active species are
molecular in nature and that they function in a diffusion-
controlled regime (Fig. S8†).
of a number of biomimetic Mo–Cu complexes, but none of them
had been reported to show catalytic activity for CO2
Cyclic voltammetry of complex 1 in the presence of CO2 and
reduction.34–38
proton sources
In this work, we have performed in-depth investigations of
The electroreduction of CO2 by complex 1 in the absence of
a proton source was examined by cyclic voltammetry. Fig. 2
shows the CVs of complex 1 in dry argon- and CO2-saturated
CH3CN on a GC electrode between ꢀ1.52 and ꢀ2.62 V. In the
presence of CO2, the redox process at E1/2 ¼ ꢀ2.07 V became
irreversible and slightly increased in current density, and
a catalytic wave appeared at an onset potential of roughly
ꢀ2.4 V. The irreversibility and current increase of the signal at
ꢀ2.07 V was also observed when cycling only down to ꢀ2.22 V,
before the catalytic wave (inset, Fig. 2).
Further purging the electrochemical cell with argon restored
the reversible redox process associated to 1 at ꢀ2.07 V, indi-
cating that complex 1 did not decompose in these conditions
(Fig. S9†). No products were detected over a 1 hour electrolysis
at ꢀ2.17 V, conrming the absence of catalytic activity of the
complex at that potential. These data suggest a non-catalytic
irreversible reaction of reduced complex 1 with CO2 leading to
the formation of the active catalyst. We assigned the slight
increase of activity mainly to the increase of concentration of
the complex occurring during the CO2 purging process causing
partial evaporation of the solvent.
the CO2 electroreduction catalytic activity of one of these
complexes,38 [(bdt)MoVI(O)S2CuICN]2ꢀ, 1, (bdt ¼ benzenedi-
thiolate), in which Mo and Cu ions are connected by two m2-
sulde ligands and the MPT ligand is mimicked by the bdt
ligand chelating the Mo ion. While mimicking the structure of
CODH2 active site, complex 1 reproduces the function of FDH,
catalyzing CO2 reduction with a high selectivity for formate
production. We demonstrate, using infrared spectroelec-
trochemical (IR-SEC) studies coupled with density functional
theory (DFT) calculations that complex 1 is only a pre-catalyst,
the active catalyst being formed aer reduction and deoxygen-
ation in the presence of CO2.
Results
Synthesis and characterization of [(bdt)MoVI(O)S2CuICN]2ꢀ, (1)
Complex 1 was synthesized following an adapted synthetic
procedure from Tatsumi et al.38 and present identical NMR, UV-
visible and FTIR spectroscopic signatures to previously reported
ones (Fig. S1–S3†). However, the cyclic voltammogram (CV) at
50 mV sꢀ1 of 1 under strict anaerobic conditions in argon-
The effect of proton sources on promoting CO2 reduction
was then investigated. Recording CVs of complex 1 under CO2
saturated
CH3CN
using
tetrabutylammonium
hexa-
uorophosphate (TBAPF6) supporting electrolyte and glassy
carbon (GC) working electrode differed signicantly from that
published in ref. 38. While the latter exhibited two quasi-
reversible waves at E1/2 ¼ ꢀ2.02 V and ꢀ1.23 V vs. Ag/AgNO3
both originally assigned to one-electron Mo-based redox
processes, we only observed a single quasi-reversible process at
E1/2 ¼ ꢀ2.02 V vs. Ag/AgNO3, (or ꢀ2.07 V vs. Fc/Fc+; unless stated
otherwise, all potentials in the text below are vs. Fc/Fc+) (Fig. S4
and S5†). In our hands, we noticed the appearance upon
exposure of the electrochemical cell containing complex 1 to air
of an additional process at ꢀ1.22 V vs. Fc/Fc+ (corresponding to
the wave at ꢀ1.23 V vs. Ag/AgNO3 38) (Fig. S4†). This CV change
proved reversible, as witnessed by the disappearance of this
signal aer 30 min degassing of the solution with Ar, while the
signal at ꢀ2.07 V remained unchanged. Therefore, we assigned
this signal to O2 reduction and argue that the previously re-
ported CV38 reected a contamination of the electrochemical
cell with air (Fig. S4†).
Fig. 2 Cyclic voltammograms of complex 1 under Ar (blue) and CO2
applying potentials from ꢀ1.52 V to ꢀ2.22 V vs. Fc/Fc+ (red, inset) and
from ꢀ1.52 V to ꢀ2.62 V vs. Fc/Fc+ (black). 1 mM solution in dry CH3CN
containing 0.1 M TBAPF6 electrolyte; GC working electrode; The scan
Determination of the number of electrons associated to the
redox process at E1/2 ¼ ꢀ2.07 V (Fig. S6 and S7†) by using rate was 50 mV sꢀ1; the second scan is shown.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
Fig. 4 IR-SEC of complex 1 in dry CH3CN with 0.1 M TBAPF6 at resting
potential under Ar (green), CO2 (blue), and at ꢀ2.24(11) V vs. Fc/Fc+
under Ar (black) and CO2 (red). LabOMak IR-SEC cell equipped with Pt
grid for working and counter electrodes and CE and Ag pseudo-
reference electrode.
Fig. 3 Cyclic voltammograms of 1 mM complex 1 with 0.1 M TBAPF6 in
dry CH3CN on a GC electrode under Ar (blue), CO2 (red), CO2 with
0.1 M TFE (black), CO2 with 0.1 M H2O (green) and CO2 with 0.1 M
PhOH (purple). The scan rate was 50 mV sꢀ1; the second scan is
shown.
slightly increased up to 74%, while FY for CO decreased and FY
for H2 remained almost unaffected (Table 1 and Fig. S11†).
Control experiment without catalyst showed very low current
density and no formate formation (Table 1 and Fig. S12†).
The molecular origin of the catalytically active species was
assessed by several control experiments. First, a rinse test was
carried out. Aer 1 hour CPE in the presence of complex 1 under
standard catalytic conditions (CO2-saturated CH3CN with 0.1 M
TBAPF6 and 0.1 M TFE), the electrode was separated from the
electrolyte and then assayed for a second CPE using a fresh
electrolyte solution in the absence of complex 1. As shown in
Fig. S13,† the current density during the following CPE at
ꢀ2.62 V was very low, as compared to that observed in the rst
CPE run, and analysis of the solution aer one hour did not
show any formation of CO2 reduction products. Second, no
electrodeposited species were observed in electrocatalytic
conditions, as shown by the absence of dissolution processes in
post-electrolysis CV studies (Fig. S14†) and by the unchanged
energy-dispersive X-ray spectra and scanning electron micros-
copy images of the electrode (Fig. S15–S17†) before and aer
electrolysis. This suggests that the catalytic activity originates
from the soluble molecular species.
in the presence of 0.1 M H2O (green line), 0.1 M TFE (tri-
uoroethanol, black line) or 0.1 M PhOH (phenol, purple) as
proton sources allowed identifying TFE as the most effective to
increase the current density of the catalytic process (Fig. 3).The
concentration of 0.1 M TFE was found to be optimal, while
concentrations greater than 0.2 M resulted in a loss of current
density (Fig. S10†).
Catalytic electroreduction of CO2 studied by controlled
potential electrolysis
Controlled potential electrolysis (CPE) was carried out in CO2-
saturated CH3CN on a GC electrode at ꢀ2.62 V in the presence
of 0.1 M of TFE. The current density remained stable over the
1 h experiment (Fig. S11†). Product analysis of the headspace by
gas chromatography revealed the presence of H2 and CO as the
only gaseous products while ion-exchange liquid chromatog-
raphy demonstrated the presence of formic acid. Quantication
of these products (Table 1) indicated that the major reaction
product was formate (FY ¼ 69%), while CO (FY ¼ 8%) and H2
(FY ¼ 19%) accounted for the rest of the passed charge (total FY
¼ 96%) for an overall turnover number (TON) of 4. A control
experiment was carried out under an Ar atmosphere and did not
show any formation of CO2 reduction products, the only
detected product being H2.
The importance of the bimetallic nature of the catalyst was
illustrated in a comparative study with the mononuclear bis-bdt
Mo complex, [MoIVO(bdt)2][Et4N]2, used as a model of the Mo
subunit of complex 1. The CPE of [MoIVO(bdt)2][Et4N]2 at
ꢀ2.62 V in CO2-saturated CH3CN with 0.1 M TBAPF6 and 0.1 M
TFE did not show any CO2 reduction products (Fig. S18†).
The product distribution showed only a minor dependence
on the applied potential. When CPE was carried out at
a signicantly more anodic potential (ꢀ2.37 V), FY for formate
Table 1 Products from controlled potential electrolyses at ꢀ2.62 V and ꢀ2.37 V vs. Fc/Fc+ during 1 hour on GC electrodea
E (V)
Catalyst
FY CO (%)
FY H2 (%)
FY HCOOꢀ (%)
Total FY (%)
TONb
ꢀ2.62
ꢀ2.37
ꢀ2.62
1 (0.5 mM)
1 (0.5 mM)
No catalyst
8
4
4
19
21
11
69
74
0
96
99
14c
2.8 (4.0)d
1.6
—
a
b
c
0.1 M TBAPF6 and 0.1 M TFE in 5 mL CO2-saturated CH3CN. See ESI for details of turnover number (TON) calculation. The low total FY is
a consequence from the very low overall current, rendering background current (faradaic and kinetic/induced background currents) non-
negligible. d TON aer 2 h in brackets.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
According to our calculations (Fig. 5), the two most plausible sites
for CO2 binding on the reduced complex 1red are on the oxo atom
(complex 2a) and one of the m2-S atoms (complex 2b), see also
Fig. S22 and S23.† The energy difference between the two
Infrared spectroelectrochemical studies
In order to gain further insights into the CO2 reduction reaction
mechanism, we investigated the behaviour of complex 1 in the
presence of CO2 via an infrared spectroelectrochemistry (IR-
SEC) technique (Fig. 4).39,40 At resting potential, the IR stretch
corresponding to the nCN mode was observed at 2124 cmꢀ1 in
both Ar and CO2 atmosphere, conrming that complex 1 did not
react with CO2. At ꢀ2.24(11) V, the nCN mode shied to
2081 cmꢀ1, in agreement with an increased p backdonation to
the CNꢀ ligand upon reduction. Under a CO2 atmosphere at
ꢀ2.24(11) V the same nCN stretch was observed, together with
new stretches at 2150 and 2187 cmꢀ1, indicating the formation
of two new complexes presenting a signicant change in the
geometry of the Cu site. In addition, the appearance of new
complexes (B3LYP-D3/def2-TZVPD level) is only 2.8 kcal molꢀ1
.
The observation by IR spectroscopy measurements (see above) of
2ꢀ
CO3 species during reaction of reduced complex 1 with CO2
suggests binding at the Mo]O group and the computed reaction
free energy for the generation of complex 2a is exergonic by
4.5 kcal molꢀ1. Then, aer binding a CO2 molecule 2a generates
the pre-catalyst 3 (DG ¼ ꢀ0.6 kcal molꢀ1) upon an irreversible
2ꢀ
transfer of the oxo group to form CO3 (Fig. 6). Complex 3 has
a triplet ground state (the corresponding singlet state is
18.8 kcal molꢀ1 higher in energy), with Mo ion in a tetrahedral
geometry.
2ꢀ
intense IR bands at 1678 and 1640 cmꢀ1, characteristic of CO3
ions, implied that reduced complex 1 reacted with CO2 via
2ꢀ
DFT mechanistic pathways
oxo-transfer to produce CO3
in solution. The complete
disappearance of the signal for initial complex 1 at 2124 cmꢀ1
highlights the considerably faster kinetics for the complex
reduction in presence of CO2. Finally, the appearance of a new
stretch at 1875 cmꢀ1 suggested the formation of a Mo–H
species.41 While this stretch was observed in the absence of
a proton source, we hypothesize that protons might derive from
the acetonitrile solvent. This was further conrmed by the
disappearance of this stretch when deuterated acetonitrile was
used as a solvent (together with that of the nCN stretch at
2187 cmꢀ1, likely belonging to the same hydride species,
Fig. S19†).
Electrochemical studies suggest that catalysis requires addi-
tional electron(s). Fig. 6 shows the proposed pathways of CO2
reduction toward formate and hydrogen, involving a further
one-electron reduction of 3 leading to 3red complex in a doublet
ground state. The associated computed redox potential
Computational investigations of the CO2 reduction
DFT computations were performed to explore the CO2 reduction
mechanism by 1 and to further elaborate the assignments of the
spectroscopic features discussed above.
Initial reaction of 1 in the presence of CO2. Complex 1 was
fully optimized in CH3CN starting from its reported X-ray crystal
structure.38 Selected bond distances and angles are summarized
in Table S2,† showing excellent agreement with the experimental
˚
parameters (bond lengths are within 0.02 A of the values in the
crystal structure and the bond angles are within 1ꢁ), except for
Mo–Cu/Cu–S bond lengths, which are underestimated by as
˚
much as 0.06 A. The ground state of 1 is a closed-shell singlet,
containing MoVI and CuI centers, which upon 2-electron reduc-
tion generates complex 1red with MoIV and CuI sites (see Fig. S20
and S21† for the corresponding frontier molecular orbitals).
Fig. 6 Proposed reaction mechanisms of CO2 reduction to formate
and hydrogen by complex 3 obtained from the pre-catalyst 1. The
relative Gibbs free energies (DG, kcal molꢀ1) and transition state
Fig. 5 DFT optimized structures (B3LYP-D3/def2-TZVPD) of CO2 barriers (DG#, kcal molꢀ1) are given relative to the preceding inter-
bound to complex 1red on oxo (2a) and on m2-S (2b). Bond distances mediate. The standard one-electron reduction potential (E0, V) is given
are given in A˚. Mo is green, Cu is orange, S is yellow, O is red, N is blue, vs. Fc/Fc+. Note that the added electron in complex 3red resides on the
C is grey and H is white.
S3p orbitals of the ligands.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
(E0 ¼ ꢀ2.20 V) is in good agreement with the experimental onset triggers the irreversible transfer of the oxo group on Mo to a CO2
of ꢀ2.40 V. The added electron resides on the S3p orbitals of the molecule to afford one carbonate equivalent and the tetrahedral
ligands exclusively, thus preserving the oxidation state of MoIV Mo(IV) complex 3. The mechanism for the oxo transfer observed
(Fig. S24b†). Further one-electron reduction of 3red is computed here differs signicantly from oxo-transfer mechanisms
to occur at prohibitive redox potential and, as such, the plau- observed with molybdoenzymes and synthetic analogues,43
sible pathways involving 2-electron reduction of 3 have not been where the oxo transfer is a redox process involving the reduction
considered. The 3red complex is protonated by a TFE molecule of a dioxo Mo(VI) complex to a monooxo Mo(VI) species. In the
most favorably at the Mo center (computed TS barrier of present case, the monooxo Mo(VI) complex 1 has only a weak
10.8 kcal molꢀ1), giving rise to a Mo hydride species, 4-MoVH basic character and the 2-electron reduction of the Mo center
(Fig. S25†). Alternative protonation of the two different S atom allows lowering its Lewis acidity and sufficiently enhancing the
types, i.e., metal bridging S and S from dithiolene, is +4.1 and basicity of the oxo group to induce its transfer to CO2. The
+16.6 kcal molꢀ1, respectively, higher in energy. Formate is resulting complex, 3, can be further reduced by one electron to
formed via a direct CO2 attack on 4-MoVH through a transition generate the active catalyst 3red, which is showing strong
state TS4-3red with a free energy of activation of 9 kcal molꢀ1
.
electron-donating properties at the Mo center provided by the
Direct hydride transfer to form [MoV/HCO2ꢀ] complex is fol- metal-bridging S ligands. This allows for the formation of
lowed by a release of HCOOꢀ and regeneration of 3red upon two- reactive MoVH hydride intermediates, which can react with CO2
electron reduction. Alternatively, the 4-MoVH intermediate can to produce formate. The rather large overpotential observed
accept a second proton from the acid source to form 5-MoVH-SH with this catalyst can be explained by the highly cathodic
(DG# ¼ 12.1 kcal molꢀ1). The protonation occurs on a bridging S potential required to transfer an electron on the S3p, necessary
atom, resulting in a S–H bond syn to the Mo–H bond. The to generate a hard Mo center able to react with TFE and form
computed TS barrier for hydrogen formation amounts to such MoVH hydride intermediates. Furthermore, the higher
16.4 kcal molꢀ1
.
The release of hydrogen (DG
¼
transition state barrier for hydrogen formation than the barrier
ꢀ8.8 kcal molꢀ1) is followed by a two-electron reduction to for formate production of these Mo hydride species is in
regenerate the active species 3red. In contrast, a direct attack of agreement with the experimental nding of formate being the
TFE on the Mo–H bond to form MoH2 adduct has a signicantly major product, while H2 accounts for only 20%.
higher transition state barrier of 28.7 kcal molꢀ1 and is hence
ruled out as an operating H2 pathway.
This work enriches the family of CO2-to-formate electro-
catalysts based on earth-abundant metals. The generation of an
active CO2 reduction catalyst species via the loss of an oxo group
involved in carbonate formation is to our knowledge unprece-
dented. It allows for an in situ generation of a vacant coordi-
nation site leading to the formation of a highly reactive metal
hydride complex. The presence of the CuI site is essential for the
catalytic function of this catalyst, but more efforts are needed to
understand the interplay between the two metal centers, which
will be a subject of future studies.
IR frequencies by DFT
The reaction mechanism in Fig. 6 is further corroborated by
comparison of computed characteristic vibrational frequencies
of complex 1 and some of the key intermediates in the catalytic
cycle to the experimental IR-SEC vibrations (Tables S3 and S4†).
The experimental nCN stretches at 2081 cmꢀ1, 2150 cmꢀ1 and
2187 cmꢀ1 observed upon reduction of complex 1 in a CO2
atmosphere are in good agreement with the computed values at
2059, 2128, and 2135 cmꢀ1 of the reduced complexes 1red, 3 and
4-MoVH. Moreover, the calculated Mo–H stretching mode at
1858 cmꢀ1 strongly suggests that the measured band at
1874 cmꢀ1 can be attributed to the hydride complex 4-MoVH
formed in the presence of a proton source, which is a key
intermediate for the product formation.
Experimental section
General methods
All experiments were performed under dry and oxygen-free
argon atmosphere using standard Schlenk or glovebox tech-
niques. Unless otherwise specied, solvents were dried using
a MBRAUN SPS-800 solvent purication system. Diethyl ether
was vacuum distilled from K/benzophenone. All solvents were
degassed by three freeze–pump–thaw cycles prior to use.
Discussion and conclusions
˚
Acetonitrile was stored over activated 3 A molecular sieves and
˚
The main goal of this work was to evaluate the electrocatalytic diethyl ether and dichloromethane were stored over 4 A
activity in CO2 reduction of the heterobimetallic complex [(bdt) molecular sieves. Water solutions were prepared from distilled
1
MoVI(O)S2CuICN]2ꢀ, 1, which is inspired by the active site of the and degassed MilliQ water. H NMR spectra were recorded on
CODH2 enzyme. We found that complex 1 was catalytically a Bruker Avance-III 300 MHz NMR spectrometer at room
active and shows a good selectivity for formate, formed with an temperature. UV-Vis spectra were recorded using a Cary 100 UV-
overpotential of ca. 800 mV. Such high overpotentials have been Vis spectrophotometer (Agilent). FTIR spectroscopy measure-
oen observed for the few molecular catalysts catalyzing the ments were carried out in a glovebox using a Nicolet iS5
reduction of carbon dioxide to formate.1,42
instrument from Thermo Scientic. SEM measurements were
Combined electrochemical analysis, IR-SEC data, DFT acquired using a SEM-FEG Hitachi SU-70 scanning electron
computations and catalytic activity studies identied that microscope. The complex [MoIVO(bdt)2][Et4N]2 was prepared
complex 1 is only a pre-catalyst. Its two-electron reduction according to literature procedures.44,45
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Synthesis of (Et4N)2[(bdt)MoVI(O)S2CuICN], (1)
experiments were carried out in a LabOMak UF SEC cell with Pt
mesh working and counter electrode. Unless otherwise stated,
FTIR spectra were recorded aer a constant potential electrol-
ysis of 20 s at ꢀ1.9 V vs. a Ag wire pseudo reference electrode
and were refereed to Fc/Fc+ by measuring the reversible poten-
tial of Fc added in the IR-SEC cell in the last cycle of the
measurement sequence. Small shis of potential (in the range
of 200 mV) for the Fc/Fc+ couple were occasionally observed in
our setup and for the rst CV cycle before stabilizing (Fig. S26†).
To account for these small potential shis, all IR-SEC experi-
ments were referenced to the value of +0.34 V, which is the
average E1/2 for Fc/Fc+ determined in our setup for the experi-
ments reported here and provided the associated standard
deviation value of 0.11 V in the text.
In the glovebox, a 5 mL acetonitrile solution of 1,2-benzenedi-
thiol (0.042 g, 0.29 mmol) was added under stirring to an orange
solution of (Et4N)2[O2MoS2CuCN] (0.158 g, 0.29 mmol) in 15 mL
acetonitrile. The solution immediately turned purple-black.
Stirring was pursued for 4 hours. Solvent was removed under
vacuum, and the purple-black solid was washed with diethyl
ether (3 ꢂ 5 mL) to afford a purple powder. Recrystallization
from acetonitrile/diethyl ether yielded ((Et4N)2[1]) as purple
crystals (0.149 g, 76%). IR (cmꢀ1) (intensity): n(CN) 2119
(medium), n(Mo]O) 916 (strong). UV-Vis (lmax, nm, CH3CN):
233, 251, 310, 380, 528. 1H NMR (CD3CN, anion): d 7.30 (dd, 2H),
6.81 (dd, 2H).
Electrochemistry
Product detection
All cyclic voltammetry (CV) and bulk electrolysis experiments
H2 and CO were analyzed by gas chromatography (Multi-Gas
Analyzer #5; SRI Instruments), equipped with HaySep D and
were performed using
a gastight three-electrode single-
compartment cell with a SP300 Bio-Logic potentiostat (Bio-
Logic Science instruments SAS). Solutions (for CV and CPE)
were prepared in the glove box with dry and degassed acetoni-
˚
MoleSieve 5 A columns and thermal conductivity detector and
ame ionization detector equipped with a methanizer using argon
as a carrier gas. The gas chromatograph was calibrated by using
a standard gas mixture containing 2500 ppm of H2 and CO in CO2
(Messer). The typical volume of injected gas was 50 mL.
Formate was quantied using ion exchange chromatography
(883 Basic IC; Metrohm).
trile
(CH3CN),
0.1
M
tetrabutylammonium
hexa-
uorophosphate (TBAPF6) as the supporting electrolyte and
1 mM of the studied complex unless specied. 3 M Ag/AgCl
electrodes, using a guard lled with the electrolyte solution
and separated from the cell by a Vycor frit was used as the
reference electrode unless specied.
Computational details
For CV studies, the electrochemical cell (SVC-2 Voltammetry
cell, ALS Co., Japan) was loaded in glovebox before the experi-
ments. A 0.5 mm diameter platinum wire (5 cm length) was used
as the counter electrode. 1 mm diameter glassy carbon electrode
(BASi, USA) was used as the working electrode. The scan rate was
50 mV sꢀ1 unless specied. In all CV plots reported in this work
the 2nd scan is presented. Experiments were run outside the
glovebox aer further purging the cell with Ar or CO2 for
a minimum of 20 minutes. Ferrocene was used as an internal
standard in each experiment by addition to the cell at the end of
the measurement and running a subsequent CV cycle.
Bulk electrolysis experiments were performed using
a gastight custom-built H-type cell loaded in a glovebox with its
two compartments separated by two Ceramic-PVDF composite
membranes (16 mm thickness, Xuran). A 0.5 mm diameter
platinum wire (10 cm length) was used as the counter electrode
and placed in the anodic compartment. A glassy carbon elec-
trode plate (2 cm2, 1 mm thickness, Alfa Aesar) was used as the
working electrode in the cathodic compartment. The cell was
purged with Ar or CO2 for a minimum of 30 minutes before
controlled potential electrolysis would be carried out. Faradaic
efficiencies were calculated accounting for the two equivalents
of electrons initially required to generate the catalyst from
complex 1.
All geometries were fully optimized at the B3LYP-D3 46–49/def2-
TZVPD50,51 level of density functional theory using the Gaussian
09 soware package52 and the SMD implicit-solvation model (3
¼ 35.688 for acetonitrile).53 A quasi-relativistic SDD pseudopo-
tential was used for Mo.54 Harmonic vibrational frequencies
were computed on the optimized geometries to ensure that all
local minima display real frequencies only, whereas the tran-
sition states were characterized by a single imaginary frequency.
Thermochemical contributions were calculated using the ideal
gas, rigid rotor, and harmonic oscillator approximations at
a temperature of 298.15 K. The activation barriers were
computed using TFE molecule as an explicit proton source. The
reduction potentials were calculated using the relation E0
¼
(ꢀDG0/nF) ꢀ Er0ef, in which n is the number of transferring
electrons, F is Faraday's constant, DG0 is the Gibbs free energy
of reduction, calculated for structures optimized in solution,
and E0ref is the absolute reduction potential of the reference
species (Fc/Fc+ couple) computed at the same level of theory.
Conflicts of interest
There are no conicts to declare.
Acknowledgements
Infrared spectroelectrochemistry (IRSEC)
The authors warmly thank Dr Françoise Pillier for help with
Infrared spectroelectrochemistry (IR-SEC) experiments were SEM measurements, and Dr David Wakerley for valuable
performed using 10 mM solution of the complex in CH3CN assistance for electrochemical measurements and analyses.
containing 0.1 M TBAPF6 as supporting electrolyte. The This work was supported by the ANR JCJC project NitroCOCa
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Edge Article
Chemical Science
(ANR-17-CE05-0021). The calculations were performed using 22 S. Dey, M. E. Ahmed and A. Dey, Inorg. Chem., 2018, 57, 5939–
the HPC resources of GENCI (TGCC) through Grant 2019-
810082.
5947.
23 A. Taheri, E. J. Thompson, J. C. Fettinger and L. A. Berben,
ACS Catal., 2015, 5, 7140–7151.
24 A. M. Appel, J. E. Bercaw, A. B. Bocarsly, H. Dobbek,
D. L. DuBois, M. Dupuis, J. G. Ferry, E. Fujita, R. Hille,
P. J. A. Kenis, C. A. Kerfeld, R. H. Morris, C. H. F. Peden,
A. R. Portis, S. W. Ragsdale, T. B. Rauchfuss, J. N. H. Reek,
L. C. Seefeldt, R. K. Thauer and G. L. Waldrop, Chem. Rev.,
2013, 113, 6621–6658.
25 T. Fogeron, P. Retailleau, L.-M. Chamoreau, Y. Li and
M. Fontecave, Angew. Chem., Int. Ed., 2018, 57, 17033–17037.
26 J.-P. Porcher, T. Fogeron, M. Gomez-Mingot, E. Derat,
L.-M. Chamoreau, Y. Li and M. Fontecave, Angew. Chem.,
Int. Ed., 2015, 54, 14090–14093.
Notes and references
1 R. Francke, B. Schille and M. Roemelt, Chem. Rev., 2018, 118,
4631–4701.
2 N. Elgrishi, M. B. Chambers, X. Wang and M. Fontecave,
Chem. Soc. Rev., 2017, 46, 761–796.
3 C. Costentin, M. Robert and J.-M. Saveant, Chem. Soc. Rev.,
2013, 42, 2423–2436.
´
4 W. W. Kramer and C. C. L. McCrory, Chem. Sci., 2016, 7,
2506–2515.
´
5 X. Wang, I. Thiel, A. Fedorov, C. Coperet, V. Mougel and 27 T. Fogeron, J.-P. Porcher, M. Gomez-Mingot, T. K. Todorova,
M. Fontecave, Chem. Sci., 2017, 8, 8204–8213.
L.-M. Chamoreau, C. Mellot-Draznieks, Y. Li and
6 C. Sun, L. Rotundo, C. Garino, L. Nencini, S. S. Yoon,
M. Fontecave, Dalton Trans., 2016, 45, 14754–14763.
R. Gobetto and C. Nervi, ChemPhysChem, 2017, 18, 3219– 28 J.-P.
Porcher,
T.
Fogeron,
M.
Gomez-Mingot,
3229.
L.-M. Chamoreau, Y. Li and M. Fontecave, Chem.–Eur. J.,
2016, 22, 4447–4453.
29 T. Fogeron, P. Retailleau, L.-M. Chamoreau, M. Fontecave
and Y. Li, Dalton Trans., 2017, 46, 4161–4164.
30 T. Fogeron, T. K. Todorova, J.-P. Porcher, M. Gomez-Mingot,
L.-M. Chamoreau, C. Mellot-Draznieks, Y. Li and
M. Fontecave, ACS Catal., 2018, 8, 2030–2038.
31 T. Fogeron, P. Retailleau, M. Gomez-Mingot, Y. Li and
M. Fontecave, Organometallics, 2019, 38, 1344–1350.
32 J.-H. Jeoung and H. Dobbek, Science, 2007, 318, 1461–1464.
7 B. Reuillard, K. H. Ly, T. E. Rosser, M. F. Kuehnel, I. Zebger
and E. Reisner, J. Am. Chem. Soc., 2017, 139, 14425–14435.
8 A. Zhanaidarova, C. E. Moore, M. Gembicky and
C. P. Kubiak, Chem. Commun., 2018, 54, 4116–4119.
9 M. N. Jackson, S. Oh, C. J. Kaminsky, S. B. Chu, G. Zhang,
J. T. Miller and Y. Surendranath, J. Am. Chem. Soc., 2018,
140, 1004–1010.
10 S. Sato, K. Saita, K. Sekizawa, S. Maeda and T. Morikawa, ACS
Catal., 2018, 8, 4452–4458.
11 A. Zhanaidarova, A. L. Ostericher, C. J. Miller, S. C. Jones and 33 H. Dobbek, L. Gremer, R. Kiefersauer, R. Huber and
C. P. Kubiak, Organometallics, 2019, 38, 1204–1207.
O. Meyer, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 15971–
12 A. Zhanaidarova, S. C. Jones, E. Despagnet-Ayoub,
15976.
B. R. Pimentel and C. P. Kubiak, J. Am. Chem. Soc., 2019, 34 C. Gourlay, D. J. Nielsen, J. M. White, S. Z. Knottenbelt,
141, 17270–17277.
13 D. Karapinar, A. Zitolo, T. N. Huan, S. Zanna, D. Taverna,
M. L. Kirk and C. G. Young, J. Am. Chem. Soc., 2006, 128,
2164–2165.
L. H. G. Tizei, D. Giaume, P. Marcus, V. Mougel and 35 C. Gourlay, D. J. Nielsen, D. J. Evans, J. M. White and
M. Fontecave, ChemSusChem, 2020, 13, 173–179.
C. G. Young, Chem. Sci., 2018, 9, 876–888.
14 M. Beley, J. P. Collin, R. Ruppert and J. P. Sauvage, J. Am. 36 S. Groysman, A. Majumdar, S.-L. Zheng and R. H. Holm,
Chem. Soc., 1986, 108, 7461–7467. Inorg. Chem., 2010, 49, 1082–1089.
15 M. Beley, J.-P. Collin, R. Ruppert and J.-P. Sauvage, J. Chem. 37 T. S. Hollingsworth, R. L. Hollingsworth, R. L. Lord and
Soc., Chem. Commun., 1984, 1315–1316. S. Groysman, Dalton Trans., 2018, 47, 10017–10024.
16 M. Wang, K. Torbensen, D. Salvatore, S. Ren, D. Joulie, 38 M. Takuma, Y. Ohki and K. Tatsumi, Inorg. Chem., 2005, 44,
F. Dumoulin, D. Mendoza, B. Lassalle-Kaiser, U. I¸sci,
6034–6043.
C. P. Berlinguette and M. Robert, Nat. Commun., 2019, 10, 1–8. 39 C. W. Machan, M. D. Sampson, S. A. Chabolla, T. Dang and
´
´
17 M. Hammouche, D. Lexa, J. M. Saveant and M. Momenteau,
J. Electroanal. Chem. Interfacial Electrochem., 1988, 249, 347– 40 Q. Zeng, J. Tory and F. Hartl, Organometallics, 2014, 33,
351. 5002–5008.
18 M. Bourrez, F. Molton, S. Chardon-Noblat and A. Deronzier, 41 H. D. Kaesz and R. B. Saillant, Chem. Rev., 1972, 72, 231–281.
C. P. Kubiak, Organometallics, 2014, 33, 4550–4559.
Angew. Chem., Int. Ed., 2011, 50, 9903–9906.
19 C. Cometto, L. Chen, P.-K. Lo, Z. Guo, K.-C. Lau,
42 H. Takeda, C. Cometto, O. Ishitani and M. Robert, ACS
Catal., 2017, 7, 70–88.
´ `
E. Anxolabehere-Mallart, C. Fave, T.-C. Lau and M. Robert, 43 J. H. Enemark, J. J. A. Cooney, J.-J. Wang and R. H. Holm,
ACS Catal., 2018, 8, 3411–3417.
Chem. Rev., 2004, 104, 1175–1200.
20 M. D. Rail and L. A. Berben, J. Am. Chem. Soc., 2011, 133, 44 A. Roodt, S. S. Basson and J. G. Leipoldt, Polyhedron, 1994,
18577–18579. 13, 599–607.
21 S. Roy, B. Sharma, J. Pecaut, P. Simon, M. Fontecave, 45 S. Boyde, S. R. Ellis, C. D. Garner and W. Clegg, J. Chem. Soc.,
´
P. D. Tran, E. Derat and V. Artero, J. Am. Chem. Soc., 2017,
139, 3685–3696.
Chem. Commun., 1986, 1541–1543.
46 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
This journal is © The Royal Society of Chemistry 2020
Chem. Sci.

View Article Online
Chemical Science
Edge Article
47 P. J. Stephens, F. J. Devlin, C. F. Chabalowski and 52 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
48 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys.,
2010, 132, 154104.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, et al., Gaussian 09, Gaussian,
Inc., Wallington CT., 2013.
49 S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem., 53 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem.
2011, 32, 1456–1465. B, 2009, 113, 6378–6396.
50 F. Weigend and R. Ahlrichs, Phys. Chem. Chem. Phys., 2005, 54 D. Andrae, U. Haußermann, M. Dolg, H. Stoll and H. Preuß,
¨
7, 3297–3305.
Theor. Chim. Acta, 1990, 77, 123–141.
51 D. Rappoport and F. Furche, J. Chem. Phys., 2010, 133,
134105.
Chem. Sci.
This journal is © The Royal Society of Chemistry 2020