Molecular Cobalt Complexes with Pendant Amines for Selective Electrocatalytic Reduction of Carbon Dioxide to Formic Acid

Article abstract of DOI:10.1021/jacs.6b11474

We report here on a new series of CO₂-reducing molecular catalysts based on Earth-abundant elements that are very selective for the production of formic acid in dimethylformamide (DMF)/water mixtures (Faradaic efficiency of 90 ± 10%) at moderate overpotentials (500-700 mV in DMF measured at the middle of the catalytic wave). The [CpCo(P^R₂N^R′₂)I]⁺ compounds contain diphosphine ligands, P^R₂N^R′₂, with two pendant amine residues that act as proton relays during CO₂-reduction catalysis and tune their activity. Four different P^R₂N^R′₂ ligands with cyclohexyl or phenyl substituents on phosphorus and benzyl or phenyl substituents on nitrogen were employed, and the compound with the most electron-donating phosphine ligand and the most basic amine functions performs best among the series, with turnover frequency >1000 s^-1. State-of-the-art benchmarking of catalytic performances ranks this new class of cobalt-based complexes among the most promising CO₂-to-formic acid reducing catalysts developed to date; addressing the stability issues would allow further improvement. Mechanistic studies and density functional theory simulations confirmed the role of amine groups for stabilizing key intermediates through hydrogen bonding with water molecules during hydride transfer from the Co center to the CO₂ molecule.

Full text of DOI:10.1021/jacs.6b11474

Subscriber access provided by University of Newcastle, Australia
Article
Molecular cobalt complexes with pendant amines for selective
electrocatalytic reduction of carbon dioxide to formic acid
Souvik Roy, Bhaskar Sharma, Jacques Pecaut, Philippe Simon,
Marc Fontecave, Phong D. Tran, Etienne Derat, and Vincent Artero
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 16 Feb 2017
Downloaded from http://pubs.acs.org on February 16, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Molecular cobalt complexes with pendant amines for selective
electrocatalytic reduction of carbon dioxide to formic acid
9
Souvik Roy,1, 6 Bhaskar Sharma,2 Jacques Pecaut,3 Philippe Simon,4 Marc Fontecave,4 Phong D.
Tran,1, 5 Etienne Derat,2 and Vincent Artero1*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1 Laboratoire de Chimie et Biologie des Métaux; Université Grenoble Alpes, CEA, CNRS, 17 rue
des Martyrs, 38000 Grenoble, France
2 Institut Parisien de Chimie Moléculaire, UMR 8232, Sorbonne Universités, UPMC Univ. Paris
06, CNRS, 75005 Paris, France
3 Reconnaissance Ionique et Chimie de Coordination, DRFꢀINACꢀSyMMES, Univ. Grenoble
Alpes, CEA, 17 rue des Martyrs, 38000 Grenoble, France
4 Laboratoire de Chimie des Processus Biologiques, Collège de France, Université Pierre et Marie
Curie, CNRS UMR 8229, 11 place Marcelin Berthelot, 75005 Paris, France
5 Department of Advanced Materials Science and Nanotechnology, University of Science and
Technology of Hanoi, Vietnam.
Author Information
* Corresponding author: vincent.artero@cea.fr
6 Present Address: Ångström Laboratory, Uppsala University, Lägerhyddsvägen 1, 75120
Uppsala, Sweden
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 36
1
2
3
4
5
Abstract
6
7
8
9
We report here on a new series of CO2ꢀreducing molecular catalysts based on Earthꢀabundant
elements that are very selective for the production of acid formic in DMF/water mixtures
(Faradaic efficiency of 90±10%) at moderate overpotentials (500–700 mV in DMF measured at
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the middle of the catalytic wave). The [CpCo(PR NR’2)I]+compounds contain diphosphine ligands
2
PR NR’2 with two pendent amine residues that act as proton relays during CO2 reduction catalysis
2
and tune their activity. Four different PR NR’ ligands with cyclohexyl or phenyl substituents on
2
2
phosphorus and benzyl or phenyl substituents on nitrogen were employed and the compound with
the most electronꢀdonating phosphine ligand and the most basic amine functions performs best
among the series with turnover frequency >1000 sꢀ1. Stateꢀofꢀtheꢀart benchmarking of catalytic
performances ranks this new class of cobaltꢀbased complexes among the most promising CO2ꢀtoꢀ
formic acid reducing catalysts developed till date, while addressing the stability issues would
allow further improvement. Mechanistic studies and DFT simulations confirmed the role of
amine groups for stabilizing key intermediates through hydrogen bonding with water molecules
during hydride transfer from the Co center to the CO2 molecule.
Introduction
Utilization of CO2 as an economical and renewable C1 feedstock for production of energyꢀdense
carbonꢀbased liquid and gaseous fuels is emerging as a particularly appealing strategy in the
context of developing new energy storage technologies. CO2 is a quite inert molecule that can
only be activated through kinetically constrained multiꢀelectron/multiꢀproton processes.
Therefore, much research has been devoted into developing efficient and scalable catalysts for
CO2 reduction. However, despite promising results, many of these catalysts operate sluggishly
with low energy efficiencies and/or poor product selectivity. Molecular homogeneous catalysts
2
ACS Paragon Plus Environment

Page 3 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
contrast with their heterogeneous counterparts since they usually display high selectivity but the
most effective ones are often based on rareꢀearth metals (ruthenium, rhodium, rhenium, and
iridium).1,2 Consequently, more work must be done on the development on molecular catalysts
using earthꢀabundant elements.3ꢀ6 Among the products formed upon electrochemical CO2
reduction, CO and formic acid are the most economically viable ones.7 Formic acid is a valuable
chemical, derived from 2e– reduction of CO2, which can serve as a suitable energy carrier,8 a
hydrogen storage material,9 a liquid fuel in formic acid fuel cell applications,10 or as a raw
material for bacteria to generate higher alcohols as liquid fuels.11 In nature, COꢀdehydrogenases
(CODH) and formate dehydrogenases (FDH) are the biological catalysts for reversible
CO2 ꢀreduction that contain mono or multimetallic active sites composed of earthꢀabundant
metals ([NiFe]ꢀ and [MoSCu]ꢀCODH, andꢀ, W or MoꢀFDH).12ꢀ16 In addition to the CO2ꢀbinding
site provided by the metal active site, outerꢀsphere protonꢀrelays present in the enzyme often play
an important role in catalysis. For instance, activation of CO2 at the bimetallic active site of NiFeꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CODH is facilitated by the stabilization of the bridged η2ꢀCO2ꢀκ2CNi
,OFe adduct through
hydrogen bonding interaction with closelyꢀspaced histidine and lysine residues. In a similar
fashion, a pendent amine present in the diiron active site of [FeFe]ꢀhydrogenases (Hꢀcluster) has
been shown to facilitate hydrogen evolution by shuttling proton to and from the iron centers.17,18
Inspired by the influence of higher coordinationꢀsphere interactions on enzyme activity, many
synthetic molecular catalysts have been developed as functional models of the enzymes. DuBois
and coworkers at the Pacific Northwest National Laboratory have developed nickelꢀdiphosphine
electrocatalysts for hydrogen evolution and oxidation that feature proton relays in the second
coordination sphere. The high activity of these complexes originates from the cooperative
interaction of H2 with both the nickel center and the multiple pendent amine groups in the
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 36
1
2
3
4
5
6
7
8
9
secondary coordination sphere.19ꢀ23 Similar nickel complexes have also been shown to catalyze
electroꢀoxidation of formate in which the pendent amines of P2N2ꢀligand act as bases and abstract
protons from the Niꢀbound formate.24 The introduction of pendent proton relays also proved
instrumental for the design of efficient molecular catalysts for CO2 reduction. Nickel cyclam is a
wellꢀknown CO2ꢀreducing compound that putatively uses an intramolecular Hꢀbond to stabilize
catalytic intermediate during CO production.25,26 Savéant and coworkers have demonstrated that
CO2ꢀtoꢀCO reduction performance of ironꢀtetraphenylporphyrin catalysts can be enhanced by
increasing local proton concentration via introduction of phenolic groups in the orthoꢀpositions of
the phenyl rings.27,28 More recently, Chapovetsky et al. have reported a cobaltꢀaminopyridine
based electrocatalyst for CO2ꢀtoꢀCO reduction which contains outerꢀsphere secondaryꢀamines.29
Other designs of catalysts containing pendent proton relays include Irꢀ, Rhꢀ and Feꢀcomplexes for
hydrogenation of CO2,9,30ꢀ33 dehydrogenation of formic acid and methanol,9,34,35 and reduction of
O2.36ꢀ38
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Building on similar design principle, we report here a new family of cobalt CO2ꢀreduction
catalysts, [CpCo(PR NR' )I]I, where PR NR'2 denotes a 1,5ꢀdiazaꢀ3,7ꢀdiphosphacyclooctane ligand.
2
2
2
Four different diphosphine ligands were used to prepare a series of cobalt complexes with the
general formula CpCo(PR NR' )I2 (1: PR NR'2 = PCy2NBn2;39 2: PR NR' = PCy2NPh2;40 3: PR NR'
=
2
2
2
2
2
2
2
41
PPh NBn
;
4: PR NR' = PPh2NPh ;42 Cy and Ph denote cyclohexyl and phenyl groups,
2 2 2
2
2
respectively). These compounds exhibit excellent efficiency in terms of both catalytic rate and
low overpotential requirement for reducing CO2 to formic acid. Importantly, the process is highly
selective with little hydrogen and CO produced. These performances rank this series of catalysts
among the most promising Earthꢀabundant molecular catalyst for CO2 to formic acid conversion.
4
ACS Paragon Plus Environment

Page 5 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
Results and Discussion
9
Synthesis and characterization of complexes
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The cobalt complexes [CpCo(PR NR’ )I]I (1: PR NR' = PCy2NBn2; 2: PR NR' = PCy2NPh2; 3:
2
2
2
2
2
2
PR NR'2 = PPh NBn2; 4: PR NR'2 = PPh NPh ), were prepared by adding one equivalent of ligand to a
2
2
2
2
2
solution of CpCo(CO)I2 in dichloromethane at room temperature (Scheme 1). These complexes
were isolated as dark brown airꢀstable solids in excellent yields (75ꢀ88%). To evaluate the
influence of the pendent amine groups, 1,3ꢀbis(diphenylphosphino)propane (dppp) was used as
the ligand to prepare an analogous cobalt compound ([CpCo(dppp)I]I, 5) that did not contain
amines in the outer coordination sphere. The complexes were characterized by 1H and 31P NMR,
mass spectrometry, and optical spectroscopy. Several unsuccessful attempts to crystallize the
iodoꢀcomplexes, [CpCo(PR NR’ )I]+, prompted us to remove both coordinated and nonꢀ
2
2
coordinated iodide ions using silver triflate and generate the [CpCo(PR NR’ )(CH3CN)](TfO)2 (6:
2
2
PR NR' = PCy2NBn2; 7: PR NR' = PCy2NPh ; 8: PR NR' = PPh NBn2; 9: PR NR' = PPh NPh )
2
2
2
2
2
2
2
2
2
2
2
2
compounds (Scheme 1). Bright red crystals of these triflate derivatives were obtained by slow
diffusion of diethyl ether into concentrated acetonitrile (6–8) or dichloromethane (9’, see below)
solutions of the complexes. Xꢀray crystal structures of 6 and 7 are shown in Figure 1 and the
structures of 8 and 9’ are given in the Supporting Information (Figure S1). Molecular geometries
of all four complexes are similar with the cobalt center adopting a threeꢀlegged pianoꢀstool
geometry expected for halfꢀsandwich complexes of this type: the Cpꢀring binds in a η5ꢀmanner
and the Coꢀcoordinated PR NR’2 ligands form two six membered rings, one of which is in a chair
2
conformation and the other in a boat confirmation. An acetonitrile molecule occupies the third
coordination site, except in the structure of 9’, crystallized from 9 in CH2Cl2, which contains a
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 36
1
2
3
4
5
6
7
8
9
triflate anion, instead of acetonitrile, coordinated to the cobalt center (Figure S1, Supporting
Information). With the notable exception of the positioning of the N and P substituents in the
ligands, the four structures are very similar with quasiꢀsuperimposable [CoP2((CH2)2N)2] cores
(Figure S2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1–4 are soluble in common organic solvents, such as dichloromethane, acetonitrile, and
dimethylformamide, yielding brownꢀyellow solutions that are stable in air. They were further
characterized by UVꢀvis spectroscopy and their optical spectra in DMF (Figure S3, Supporting
Information) consist of three absorption maxima at λ = 370–390 nm, λ = 438 nm, and λ = 550–
560 nm. The molar extinction coefficient values for the absorption bands at 370–390 and 438 nm
( ꢀꢁꢂꢃꢄꢁꢅꢃ ꢆꢇ = 1870–2670 Mꢀ1cmꢀ1, and ꢀꢈꢁꢉ ꢆꢇ = 1930–2750 Mꢀ1cmꢀ1, respectively) are
considerably larger than that for the absorption band at 550–560 nm (ꢀꢁꢂꢃꢄꢁꢅꢃ ꢆꢇ= 500–550
Mꢀ1cmꢀ1).
Electrochemical characterization
Electrochemical properties of the complexes were investigated by cyclic voltammetry at a
glassy carbon electrode in DMF (0.1 M NBu4BF4 as supporting electrolyte) under an atmosphere
of argon (commercially available nonꢀanhydrous DMF was used which contains ~0.2% water).
As shown in Figure 2, the voltammograms of 1–4 display two oneꢀelectron electrochemical
features in the –0.85 to –1.35 V vs. Fc+/0 range (unless otherwise mentioned, all potentials are
referenced against the Fc+/0 redox couple). In analogy to a previous report by Bullock and coꢀ
workers, these two couples can be assigned metalꢀcentered CoIII/II and CoII/I redox processes.43
The halfꢀwave potentials ((Epa+Epc)/2) of each redox couple are listed in Table 1. The cathodic
and anodic peak currents (ꢊꢋ) for 1–4 vary linearly with the square root of the scan rate (ν1/2) from
0.05 to 0.25 Vsꢀ1 under Ar, consistent with diffusionꢀcontrolled processes as described by the
6
ACS Paragon Plus Environment

Page 7 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
RandlesꢀSevcik equation (Figure S4). Diffusion coefficients (D) in the range 1.5–3.6 × 10–6 cm2
sꢀ1 were calculated for 1–4 (Table S1). The peakꢀtoꢀpeak separation is consistent with reversible
processes (ꢁꢌꢋ = 62–80 mV, Table 1) for each redox couple, except the CoIII/II process for 1
which appears quasiꢀreversible. The ratios of cathodicꢀ and anodicꢀpeak currents were close to
unity, consistent with the chemical reversibility of those redox couples. The latter suggests that
dissolution of the complexes is likely associated with the loss of the iodide ligand, resulting in the
formation of [CpCo(PR NR’ )]2+cations in solution. This is further supported by the similar
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
2
electrochemical behaviors of the iodoꢀcompounds (1–4) and their triflateꢀanalogs (6–9) (Figures
S5 and S6). The CoIII/II and CoII/I couples shift to more positive potentials as the phosphine
groups of PR NR’ ligand become less electron donating and the amines become less basic. For
2
2
instance, changing the ligand from PCy2NBn to PPh2NBn led to positive shifts of 90 mV and 130
2
2
mV for the CoIII/II and CoII/I couples, respectively (Table 1). Similarly, potentials are shifted
positively by 60 mV (CoIII/II) and 70 mV (CoII/I) from [CpCo(PCy2NBn2)]2+ to [CpCo(PCy2NPh )]2+.
2
The CoIII/II and CoII/I couples for 5, the complex without any pendent amines, appear at less
negative potentials compared to 1–4 (Table 1). These results suggest that the pendent amine
groups, although not directly coordinated to the cobalt center, have a significant influence on its
electronic properties.
Electrochemical CO2 reduction
Complexes 1–4 were evaluated as catalysts for electrochemical CO2 reduction. The DMF
electrolytic solution was saturated with CO2 gas (∼0.20 M) in the presence of various amounts of
water.44 Importantly, hydration of CO2 forms H2CO3 which act as the proton source with a pKa
of 7.37 in DMF.27 Cyclic voltammograms of 1, 2 and 3 recorded under CO2 in DMF, displayed
cathodic current enhancement (Figure 3, S7). The mid wave potential of this process was found at
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 36
1
2
3
4
5
6
7
8
9
–2.08, –2.00, and –1.93 V for 1, 2, and 3, respectively. In contrast, cyclic voltammograms of 4
was unchanged upon addition of CO2 (1 atm.) in DMF. Incremental addition of water to CO2ꢀ
saturated DMF solutions of 1–3 resulted in increased current densities, before reaching a
maximum value and then leveling off with larger concentrations of water (>1.5, 3, and 3.5 M for
1, 2, and 3, respectively). In a control experiment, addition of saturated aqueous CO2 solution to
an Arꢀsaturated DMF solution of 1 led to current enhancement at the same potential as of CO2
(see Figure S29). In the case of 4, a weak cathodic wave was observed with midꢀwave potential
at ca. −1.88 V, upon addition of water (Figure S8). Importantly, no current enhancement was
observed in DMFꢀwater mixtures in the absence of catalyst or CO2 (see Supporting Information).
Notably, very small catalytic current was observed when 5 was employed as the catalyst in CO2
saturated DMF/water mixtures (Figure S9), strongly suggesting a critical role of the pendent
amine groups during electrocatalysis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Controlled potential electrolysis experiments
The observed current enhancement corresponds to the electrocatalytic reduction of CO2 to
formic acid, as verified by controlled potential electrolysis (CPE) carried out at mercuryꢀpool
working electrodes (surface area = 1.77 cm2) with 0.5 mM catalyst (1–4) in CO2ꢀsaturated
DMF/H2O. In general, the CPE experiments were conducted for one hour or up to complete
catalyst deactivation. During the course of electrolysis, the current density was initially quite
stable at ~0.6–1.8 mA cmꢀ2 and then slowly decayed until it reached the level of background
current. In general, the initial period of stability was shorter for higher currentꢀdensities at more
negative potentials suggesting catalyst decomposition is faster under more severe conditions. A
change of color of the solution from bright brownꢀyellow to colorless is observed at the end of
the CPE experiment. The electrocatalytic current could be restored by addition of fresh catalyst
which confirmed that catalyst degradation is the limiting factor (Figure S10). To study the effect
8
ACS Paragon Plus Environment

Page 9 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
of applied potential on product distribution and rate, different potentials were tested for 1 - 3, in
the range –2.00 to –2.25 V vs Fc+/0 (Figure S11). In general, electrolysis at higher overpotentials
led to faster catalysis, however over shorter duration (Figure S11A). The electrolysis results are
summarized in Table 2. Analyses of post electrolysis solutions by ionic chromatography revealed
the generation of large amounts of formic acid with excellent Faradaic yields (FY = 86ꢀ100%)
only in the case of 1–3 (Table 2). Small amounts of CO (FY = 0ꢀ1%) and H2 (FY = 2ꢀ11%) were
also detected as the gaseous products by gas chromatography. Formic acid, CO and H2 were not
detected when the electrolysis of a CO2ꢀsaturated DMF/H2O mixture was carried out in the
absence of catalyst. Total catalytic turnover numbers for formic acid ( ꢍꢎꢏꢐꢑꢒꢒꢐ ) were
determined from the amount of formic acid produced after 1 h electrolysis. The Faradaic yields
for formic acid production remained unaffected, but the TONHCOOH increased upon applying more
negative potentials (Table 2, Entry 1–5 for 1, Entry 6–9 for 2, Figure S12). Catalyst 1 displayed
best activity with TONHCOOH of 23 and 15 in the presence of 1.1 M and 0.56 M water,
respectively, after electrolysis at –2.25 V for 1 h (Table 2, Entry 4 and 5). Only 4 displayed poor
catalytic performance both in terms of product selectivity (FYHCOOH = 38%) and total catalytic
turnover numbers (Table 2, Entry 14), in line with the weak catalytic wave observed in its cyclic
voltammograms. Notably, electrolysis with 4 was carried out in the presence of large excess of
water (5.6 M) as its voltammograms showed very little electrocatalysis at low water
concentration. Consequently, a considerably higher amount of H2 (FYH2 = 67%) was produced
during the electrolysis with 4. Catalytic selectivity of 1 was further tested by performing cyclic
voltammetry and CPE in presence of CO2 and Et3NH+Cl– (Figure S13). Formic acid and
hydrogen were produced by CPE with 48(±5)% and 52(±5)% Faradaic efficiencies, respectively,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 36
1
2
3
4
corresponding to ~11 turnovers for formic acid (TONꢐꢑꢒꢒꢐ) and ~12 turnovers for hydrogen
5
6
(TONꢐ ) after one hour (Table 2, Entry 15).
ꢓ
7
8
9
Notably, similar values for ꢍꢎꢏꢐꢑꢒꢒꢐ was obtained when glassyꢀcarbon or graphite rod
was used as working electrodes, but FYHCOOH was significantly lower (20ꢀ50%) because a larger
amount of H2 was produced (FYH2 = 45–60%) (Table S2 and Figure S14). In the absence of any
catalyst, a small amount of formic acid was produced by direct reduction of CO2 at the graphite
electrode. Deposition of cobalt nanoparticles on carbon electrode through decomposition of the
molecular complexes may account for the color change observed in the course of longꢀterm
electrolysis and it is well known that such deposits catalyze hydrogen evolution, in addition to H2
evolution occurring directly at the surface of the carbon electrode.45,46 On the other hand, use of
mercury electrode ensured that metallic cobalt nanoparticles amalgamated and consequently,
proton reduction was suppressed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Determination of kinetic data and benchmarking of catalytic performances
The catalytic cyclic voltammograms of 1–4 were further analyzed to probe the kinetics of
the electrocatalytic CO2 reduction. For all complexes, catalytic cyclic voltammograms show
typical Sꢀshaped response. The catalytic plateau current (ꢊꢔꢕꢖ) varies linearly with the catalyst
concentration, consistent with a mechanism for CO2 reduction that is firstꢀorder in catalyst
(Figure S15). Electrocatalytic reactions are also firstꢀorder in CO2 which is evidenced by the
II/I
⁄
linear dependence of normalized peak catalytic current (ꢊꢔꢕꢖ ꢊꢋ; ꢊꢋ is determined as the Co
peak current) on the square root of [CO2] as the CO2 partial pressure was varied between 0.2 and
1 atm in N2/CO2 gas mixtures (Figure S16). No saturation of the catalytic current with increasing
CO2 concentration was detected, indicating that CO2 is involved in the rateꢀlimiting step.
Additionally, ꢊꢔꢕꢖ reaches a limiting value (Figure 3, S7) at high water concentration (>1.5, 3, and
10
ACS Paragon Plus Environment

Page 11 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
3.5 M for 1, 2, and 3, respectively) suggesting saturation kinetics in which concentration of water
is sufficiently high that it is not depleted during the course of the experiment. KIE analysis was
carried out within a concentration range of water suitable to get a reaction order of two in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
⁄
protons, i.e. linear dependence of ꢊꢔꢕꢖ ꢊꢋ on water concentration (0–1.5 M). Normal H/D kinetic
isotope effects (KIEs) of 5.0 (±0.4), 10.4 (±0.7), and 5.7 (±0.6) were found for 1, 2, and 3,
respectively (Figure 4 and Figure S17). The catalytic current plateaus in catalytic voltammograms
are scan rateꢀindependent for each complex (Figure S18), indicating catalysis in the pure kinetic
regime.47 Under such conditions, the normalized peak catalytic current (ꢊꢔꢕꢖ ꢊꢋ) is related to the
⁄
maximum turnover frequency (ꢍꢎꢗꢇꢕꢘ) of the electrocatalysis by equation 1,27,48 accounting for
a simplified 2ꢀelectron mechanism with a single rate determining step:
⁄
. ꢍꢎꢗꢇꢕꢘ . νꢄꢢ ꢣ
Equation 1
⁄
ꢝꢞꢟ
ꢊꢙꢚꢛ ꢊꢜ = 4.484
ꢡ
ꢠ
ꢍꢎꢗꢇꢕꢘ values for 1–4 under various water concentrations were calculated (Figure S19 and S20,
and Supporting Information for more details) and summarized in Table 3. Catalyst 1 operates at
highest rate with ꢍꢎꢗꢇꢕꢘ values estimated to be ~1000 sꢀ1 at >1.5 M water. Catalyst 4 showed the
lowest activity and the estimated ꢍꢎꢗꢇꢕꢘ was below 1 sꢀ1 in up to 4 M waterꢀDMF mixtures.49 At
any fixed water concentration, a clear trend for the rate of electrocatalysis rate was observed: 1
>>2 >3 >>4. It follows the evolution of the overpotential value for CO2 to formic acid conversion
at which the plateau value is reached: Eꢙꢚꢛꢋꢤ(1)<Eꢙꢚꢛꢋꢤ(2)<Eꢙꢚꢛꢋꢤ(3)<Eꢙꢚꢛꢋꢤ (4). This interplay
between turnover frequency and applied potential is better illustrated in the catalytic Tafel plots
shown in Figure 5. To build such plots, the standard reduction potential of the CO2/HCOOH
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 36
1
2
couple in DMF/H2O solutions (ꢌꢃ ꢐꢑꢒꢒꢐ (ꢥꢦꢧ) = −1.45 V)50,51, 52 was subtracted from the
3
4
⁄
ꢑꢒ
ꢓ
5
6
7
applied potential to give the overpotentials (η).
8
9
ꢑꢨꢩ
Turnover frequencies ( ꢍꢎꢗ
) could also be obtained from the preparativeꢀscale
ꢐꢑꢒꢒꢐ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
electrolyses at different operating overpotential, using the equations reported by Savéant and
coworkers (see the Supporting Information for details).3,27,28,48,53 The values are listed in Table 2.
It should be noted that the ohmic drop is much more important in the case of CPE experiments
relative to cyclic voltammetry measurements. It has a direct effect on the current measured, and
explains why ꢍꢎꢗꢇꢕꢘ and ꢍꢎꢗꢐꢑꢑꢨꢒꢩꢒꢐ values differ. We nevertheless note that they have the same
ꢑꢨꢩ
order of magnitude. The ꢍꢎꢗ
value for 1, derived from the electrolysis data, is 650 sꢀ1 at 1.1
ꢐꢑꢒꢒꢐ
M water which corresponds to generation of formic acid with a catalytic turnover number of
ꢑꢨꢩ
780,000 within 20 min (TON = ꢍꢎꢗ
× t, where t = 1200 s). TOFCPE values determined for 1–
ꢐꢑꢒꢒꢐ
3 are presented as crosses within the same catalytic Tafel plot. As shown in Figure 5, they
satisfactorily match the logTOF–η plots derived from the cyclic voltammetry measurements. The
logTOFꢀη plots generated from cyclic voltammetry data in DMFꢀwater (1.1 M) mixtures provide
a direct comparison of all four cobaltꢀcatalysts reported here. These catalysts are benchmarked
against an iridiumꢀhydride electrocatalyst (IrꢀPCP; PCP denotes a pincer ligand), previously
reported by Meyer and coꢀworkers, that selectively reduces CO2 to formic acid.54,55 The most
active cobalt catalysts in our series of complexes (1 and 2) display faster rate and greater TOFmax
relative to IrꢀPCP. Importantly, the overpotential required to achieve maximum TOF for 1 is
similar to that for IrꢀPCP. The overpotential requirement for 2 is ca. 150 mV lower than that for
IrꢀPCP while their TOFmax values are similar. We also benchmarked the performances of 1ꢀ4
against that of [Fe(N5)Cl2]2+ (N5
=
(2,13ꢀdimethylꢀ3,6,9,12,18ꢀpentaazabicycloꢀ
[12.3.1]octadecaꢀ1(18),2,12,14,16ꢀpentaene), which displays good selectivity for CO2ꢀtoꢀformic
12
ACS Paragon Plus Environment

Page 13 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
acid conversion with TOF values more than three orders of magnitude lower than 1, and an
overpotential requirement of only ~100 mV lower.50 Berben and coworkers described another
ironꢀbased catalyst [Fe4N(CO)12]– which is selective to formic acid formation and operates at a
more positive potential (–1.6 V vs Fc+/Fc) in CH3CN/H2O (95:5 v/v), corresponding to an
overpotential of ~200 mV, however, with a slower kinetics (TOFCV = 10sꢀ1) compared to the
cobalt catalysts presented herein.6,56ꢀ58
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT calculations and mechanistic considerations
Scheme 2 shows a proposed mechanism for electrocatalytic CO2 reduction that is
consistent with the electrochemical studies, and is supported by DFT calculations performed on
two different models of the acetonitrile derivative of 1 (real system – 1MeCNꢀReal –
[CpCo(PCy2NBn2)(MeCN)]2+; model system – 1MeCNꢀModel – [CpCo(PMe2NMe2)L]2+) and the
acetonitrile derivative of 5: 5MeCN ([CpCo(dppp)(MeCN)]2+) shown in Figure 5. The optimized
structures of key catalytic intermediates of 1MeCNꢀModel, and the transition state for CO2 insertion
reaction (1MeCNꢀReal, 1MeCN-Model and 5MeCN) are shown in Supporting Information (Figure S21
and S22). Indeed, experiments have shown that a CoIIꢀH intermediate generated near –2.0 V vs
Fc+/Fc is competent for catalytic H2 evolution (see the Supporting Information, Figure S23 and
S24), in line with previous reports on comparable complexes.43,59,60 Upon addition of CO2,
enhanced currents are also observed at this potential. This suggests that the same CoIIꢀH
intermediate (V) is involved in both H2 evolution and CO2 reduction. According to DFT
calculations, V can be produced from the initial CoIII species I through three different pathways
as shown in Scheme 2 and detailed below. After two successive electron transfers, a CoI species
(III) is generated. Then, this species is protonated and further reduced to form a cobalt(II)ꢀ
hydride (CoII–H) species (V, Scheme 2). DFT calculations suggest a concerted protonꢀcoupled
electron transfer (PCET) process, as our calculation for the redox potential of a CoIIIꢀH/CoIIꢀH
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 36
1
2
3
4
5
6
7
8
9
couple (–2.5 V vs Fc+/Fc), that would involve a sequential pathway (Scheme S2, top panel), was
systematically more negative than the experimental potential (–2.1 V vs Fc+/Fc) at which both H2
evolution and CO2 reduction reactions are catalyzed. Instead, the computed redox potential (–2.0
V vs Fc+/Fc) of the (CoI, R3NH+)/(CoIIꢀH,R3N) couple (R3N denotes an pendent amine) was
found in line with the experimental observation (Note that H2CO3 has a pKa value similar to
Et3NH+ in DMF). DFT calculations indicated that protonation of the pendent amine either in the
CoII or CoI states (Scheme S2, middle and bottom panels) allows two other pathways through
intermediates III’ and IV, respectively, through intramolecular proton transfer to cobalt
concerted with reduction of the metal center. While no CoIIꢀH species with similar coordination
spheres have been isolated so far, they have been evidenced in studies on CoIIIꢀH derivatives by
Kölle and Paul,59 and Dubois, Bullock and coꢀworkers.43 The CoII–H species V then reacts with
CO2 generating VI. Internal hydride transfer from cobalt to CO2 then yields VII.61 The extrusion
of formic acid from VII regenerates the CoII species II and completes the catalytic cycle. The
CoIIIꢀcomplexes are airꢀstable and easy to handle. They are however better described as preꢀ
catalysts for CO2 reduction as they require an additional electron to generate the catalytically
competent CoIIꢀspecies.62,63
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DFT calculations indicate that addition of water to the system is advantageous, thanks to
the formation of intramolecular hydrogen bonding involving the bound CO2/formate molecules
and one pendent amine residue. First, hydride transfer (Scheme 2, VI to VII) is
thermodynamically more favorable when the amines are present (Figure 5) because the
−
[Co…HCO2 ] species (VII, Scheme 2) is stabilized by Hꢀbonding with the amineꢀbound water
molecule. Second, calculated barriers are relatively small, irrespective of the model employed to
describe the transfer of hydrogen to CO2 (Figure 5). The prominent role of amine residues is
further supported by DFT calculation on 5MeCN that lacks such pendent amine groups. In that
14
ACS Paragon Plus Environment

Page 15 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
case, significantly higher activation energy is observed for the CO2 insertion step (Figure 5 and
Table S3).
7
8
9
Such a mechanistic scheme provides a rationale for the faster electrocatalytic rates
observed for 1 and 3 with the more basic benzylic amine groups compared to the corresponding
complexes 2 and 4 containing less basic aniline residues. In organic solvents, free benzylamine is
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢑꢐ ꢑꢭ
+
+
ꢬ
a stronger base than aniline ( ꢪꢫꢕ
= 16.91 and 10.62 for BnNH3 and PhNH3 ,
respectively)64,65 and a similar trend would be expected for the pendent bases in 1–4. The stronger
benzylamine base (complexes 1 and 3) would provide greater stabilization to the transition state
during hydride transfer (TS VI-VII, Figure 5) through Hꢀbonding.
The mechanism shown in Scheme 2 also agrees with the experimentally determined first
order electrocatalytic rate in the catalyst and CO2 concentration, under the assumption that the
intermediate CO2 adduct exists in rapid equilibrium with the CoII–H species and CO2. This type
of kinetics is at variance with that previously proposed by Ahn et al. for CO2 electroꢀreduction by
pincer iridium catalysts involving formate release as the rate determining step.30 In the proposed
mechanism, CO2 does not interact directly with the metal center (Cꢀcoordination of CO2 usually
leads to CO formation),66 and instead, the whole catalyst acts as a hydride transfer reagent. A
similar mechanism, based on computation, has been recently proposed for putative CO2
hydrogenation by an iron catalyst containing a similar diphosphine ligand with pendent amine
residues.67 Such a pathway is however at variance with that based on CO2 insertion in a metalꢀ
hydride bond reported for other catalysts for CO2 hydrogenation catalysts31,68ꢀ70 or
electrocatalysts for formic acid production.30,55 In this mechanism, as well as in the reverse
mechanism proposed for formate oxidation catalyzed by nickel bisdiphosphine complexes with
similar pendent amine residues,24,41,71 formate binds to the metal center through an oxygen atom
and formate release has sometimes been found as the rate determining step.30 Rather, the catalytic
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 36
1
2
3
4
5
6
7
8
9
pathway proposed here resembles the mechanism followed by catalytic hydride donors based on
dihydropyridine moieties,72,73 in which hydride transfer can be assisted by hydrogenꢀbonded
water molecules. It should also be noted that similar mechanism is operative for several
stoichiometric CO2ꢀreducing organohydride reagents including hydrosilanes,74 hydroboranes,75,76
and ammonia boranes.77 The fairly large (>5) H/D KIE values observed for these catalysts
support hydride transfer from Co to CO2, involving concomitant cleavage of CoꢀH and formation
CꢀH bonds in the rateꢀdetermining step. For comparison, KIE values of ~2 have been reported for
COꢀproducing mechanisms, involving proton transfers only, catalyzed by rhenium and iron
complexes.78,79
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Another prominent feature is the excellent selectivity of the catalysts towards CO2 over
proton reduction. We used DFT to compute the activation barriers for CO2 reduction and H2
evolution from the CoIIꢀH intermediate in the absence and presence of water molecules. In line
with experimental observations, additional DFT calculations suggested that the presence of water
favors CO2 reduction by stabilizing the CoIIꢀH by hydrogen bonding. It leads to an increase in the
activation barrier for H2 evolution to ~10 kcal mol–1 which is considerably higher than the ~3
kcal mol–1 barrier computed for CO2 reduction (Figure 5). In the absence of water, the barriers for
CO2 reduction (Table S4) and H2 evolution (Scheme S3 are ~4.7 and ~5.2 kcal mol–1,
respectively and), thus in line with the lower selectivity experimentally observed when [Et3NH]+
is used as a proton source.
Conclusion
Most molecular CO2ꢀreducing catalysts produce CO. In addition to enzymes, there are few
synthetic catalysts that generate formic acid,30,50,54ꢀ58,80ꢀ90 and among them only two are selective
16
ACS Paragon Plus Environment

Page 17 of 36
Journal of the American Chemical Society
1
2
3
4
5
and based on an Earthꢀabundant metal.50,56,57 Nickel bisꢀdiphosphine complexes bearing pendent
amine groups [Ni(PR NR' ) ]2+ have been shown to catalyze the reverse reaction, i.e. formate
2
2 2
6
7
8
9
oxidation, quite efficiently24,41,71 and this was related to the fact that the corresponding NiIIꢀ
hydride species have lower hydridicities91 than formic acid (ꢁG°H− = 44 kcal mol−1 in CH3CN).57
Such low hydridicities explain also why [Ni(PR NR' ) ]2+ complexes require strong acids to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
2 2
evolve hydrogen in CH3CN under electroꢀassisted conditions. By contrast, cyclopentadienyl
diphosphine cobaltꢀcomplexes have previously been reported as electrocatalysts for hydrogen
evolution that involves reduction of an intermediate hydride, CpCoIII(diphosphine)–H.43,59,60 The
introduction of pendent amines proved beneficial for H2 evolution catalysis although still with
large overvoltage requirement. We reasoned that such a lower activity for proton reduction
probably results in higher hydridicity for the intermediate CpCoII(diphosphine)–H, making
hydride transfer to CO2 favorable. Actually, this class of cobalt based molecular compounds
proved as quite active preꢀcatalysts for electrochemical CO2 reduction reported so far, with two
particularly notable features: (1) their excellent selectivity for reduction of CO2 to formic acid
and (2) the ability to tune their activity via ligand design. Four different PR NR’ ligands with
2
2
cyclohexyl or phenyl substituents on phosphorus (PCy or PPh) and benzyl or phenyl substituents on
nitrogen (NBn or NPh) were indeed employed and all complexes display catalytic activity towards
electrochemical reduction of CO2 to formic acid with excellent Faradaic efficiency (90±10%) and
good turnover numbers at moderate overpotentials (400–600 mV in DMF). Complex 1 that
contains the ligand with the most electronꢀdonating phosphine and most basic amine (PCy2NBn2),
performs best among the series with a remarkable turnover frequency (>1000 sꢀ1). Comparison of
the catalytic activity of the cobalt complexes with PR NR’2 ligands (1–4) and that of the related
2
complex 5, which does not contain the pendent amines, clearly indicates that the nitrogen
substituents of the PR NR ligands play an important role in the catalytic activity for these
2
2
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 36
1
2
3
4
5
6
7
8
9
complexes and suggest that more basic amines favor catalytic CO2 reduction. Mechanistic studies
indicated that such amine groups are not involved in direct proton transfer to CO2 but rather in
the stabilization of key intermediates through hydrogen bonding with water molecules during
hydride transfer. Only three CO2ꢀreducing molecular electrocatalysts (one based on iridium and
the other two based on iron) were known so far to selectively produce formic acid.50,54,56,57 Stateꢀ
ofꢀtheꢀart benchmarking of performances clearly revealed that these new cobaltꢀcatalysts
outcompete all of them in terms of maximum turnover frequency even if the [Fe4N(CO)12]–
cluster reported by Berben and coworkers displays much lower overpotential requirement.56 The
highest overall TONHCOOH of 23 (electrolysis at −2.2 V for 1 h) observed for 1 is comparable to
that obtained for other three formate producing electrocatalysts (IrꢀPCP: TON ~40 after 25 h
electrolysis;54 [Fe(N5)Cl2]2+: TON ~3.6 after 3 h electrolysis;50 [Fe4N(CO)12]–: TON ~15 ± 3
after 1 h electrolysis5). However, in contrast to IrꢀPCP and [Fe4N(CO)12]–, the cobalt catalysts are
limited by their stability under prolonged electrolysis as complete loss of activity is observed
after one hour. Still, the straightforward synthesis of such complexes, combined with the large
possibility offered by diphosphine ligands in terms of ligand tuning and outerꢀsphere control,92
hold promise for optimization of performances and the preparation of molecularꢀengineered
electrode and photoelectrode materials.93
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Reagents and Materials
The ligands PCy2NBn2, PCy2NPh2, PPh NBn2, and PPh2NPh were prepared according to reported
2
2
procedures.39ꢀ41,94 The precursor CpCo(CO)I2 was synthesized by literature method.95 All other
reagents including cyclopentadienylcobalt dicarbonyl (CpCo(CO)2, 95%), tetrabutylammonium
tetrafluoroborate (nBu4NBF4, 99%), are commercially available and were used as received.
Acetonitrile, dichloromethane, diethyl ether, and tetrahydrofuran were dried and distilled using
18
ACS Paragon Plus Environment

Page 19 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
common techniques. All manipulations of phosphineꢀcontaining compounds were carried out
under an atmosphere of purified argon in a glovebox or by standard Schlenk techniques.
7
8
General procedure for the synthesis of CpCo(diphosphine)I2 compounds
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The diphosphine ligand (0.2 mmol) was added to a black brown solution of CpCo(CO)I2 (0.081
g, 0.2 mmol) in anhydrous CH2Cl2 (15 mL) and evolution of CO was observed immediately.
After stirring the solution for 4 h at room temperature, the volatiles were removed under reduced
pressure. The residue was subjected to silica gelꢀcolumn chromatography with 1%
CH3OH/CH2Cl2 to give the pure product as a dark brown solid.
1
CpCo(PCy2NBn2)I2, (1). Yield 0.129 g (74%). HꢀNMR (300 MHz, CDCl3) δppm: 1.1ꢀ1.98 (m,
18H), 2.46ꢀ2.62 (m, 4H), 2.84 (m, 2H), 3.20 (m, 2H), 3.34 (m, 2H), 3.54 (s, 2H), 3.87 (m, 2H),
4.27 (s, 2H), 5.67 (s, 5H), 6.96 (m, 2H), 7.26ꢀ7.47 (m, 8H). 31PꢀNMR (CDCl3): 41.6 (s) ppm.
Analysis Calculated for C35H49CoI2N2P2: C, 48.18; H, 5.66; N, 3.21. Found: C, 49.80; H, 5.97; N,
3.38. SIMꢀMS: 745.2 [MꢀI]+. λmax, nm (ε, Mꢀ1cmꢀ1): 372 (1900), 438 (1900), 560 (400).
CpCo(PCy2NPh2)I2, (2). Yield 0.120 g (70%). 1HꢀNMR (300 MHz, CDCl3) δppm: 1.2ꢀ1.9 (m, 10H),
1.90 (m, 8H) 2.20 (m, 2H), 2.68 (m, 2H), 3.43 (m, 2H), 4.04ꢀ4.28 (m, 6H), 3.56 (s, 2H), 3.91 (m,
31
2H), 4.29 (s, 2H), 5.85 (s, 5H), 6.97 (d, 2H), 7.06 (m, 2H), 7.35 (q, 4H), 7.55 (d, 2H). PꢀNMR
(CDCl3): 42.4 (s) ppm. Analysis Calculated for C33H45CoI2N2P2: C, 46.94; H, 5.37; N, 3.32.
Found: C, 47.60; H, 5.39; N, 3.32. SIMꢀMS: 717.2 [MꢀI]+. λmax, nm (ε, Mꢀ1cmꢀ1): 385 (2400), 438
(2800), 554 (500).
1
CpCo(PPh2NBn2)I2, (3). Yield 0.130 g (76%). HꢀNMR (300 MHz, CDCl3) δppm: 3.25 (m, 4H),
3.66 (s, 2H), 3.88 (d, 2H), 4.27 (m, 2H), 4.64 (s, 2H), 5.36 (s, 5H), 7.03 (m, 2H), 7.26 (2H,
masked by CDCl3 peak), 7.41 (q, 4H) 7.56 (s, 6H), 7.65 (d, 2H), 8.03 (br, 4H). 31PꢀNMR
(CDCl3): 31.7 (s) ppm. Analysis Calculated for C35H37CoI2N2P2: C, 48.86; H, 4.33; N, 3.26.
Found: C, 49.79; H, 4.59; N, 3.37. SIMꢀMS: 733.1 [MꢀI]+. λmax, nm (ε, Mꢀ1cmꢀ1): 373 (2600), 438
(2500), 556 (500).
1
CpCo(PPh2NPh2)I2, (4). Yield 0.133 g (80%). HꢀNMR (300 MHz, CD2Cl2) δppm: 4.12 (m, 2H),
4.30 (s, 2H), 4.55 (d, 2H), 4.72 (m, 2H), 5.51 (s, 5H), 7.05 (m, 2H), 7.05 (m, 4H), 7.28ꢀ7.53 (m,
6H), 7.72 (br, 6H), 8.05 (br, 4H). 31PꢀNMR (CD2Cl2): 33.3 (s) ppm. Analysis Calculated for
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 36
1
2
3
4
5
6
C33H33CoI2N2P2: C, 47.62; H, 4.00; N, 3.37. Found: C, 47.56; H, 4.23; N, 3.42. SIMꢀMS: 705.3
[MꢀI]+. λmax, nm (ε, Mꢀ1cmꢀ1): 380 (2200), 438 (2400), 550 (600).
7
8
CpCo(dppp)I2, (5). Yield 0.14 g (90%). 1HꢀNMR (300 MHz, CDCl3) δppm: 1.65 (br, 2H), 2.93 (br,
9
31
4H), 5.49 (s, 5H), 7.26ꢀ7.77 (m, 20H). PꢀNMR (CDCl3): 27.6 (s) ppm. Analysis Calculated for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C32H31CoI2P2: C, 48.63; H, 3.95. Found: C, 48.28; H, 4.08. SIMꢀMS: 663.2 (20%) [MꢀI]+, 536.5
[Mꢀ2I]+ (45%). λmax, nm (ε, Mꢀ1cmꢀ1): 460 (2100), 580 (600).
General procedure for the synthesis of CpCo(diphosphine)(OTf)2 compounds
A brown solution of CpCo(diphosphine)I2 (0.05 mmol) in anhydrous CH2Cl2 (15 mL) was
cannulaꢀtransferred into a suspension of AgOTf (38.5 mg, 0.15 mmol) in anhydrous CH2Cl2 (10
mL). Pale yellow precipitate of AgI formed within 5 min. After stirring the greenꢀbrown
suspension for 2 h at room temperature, the reaction mixture was filtered to remove AgI. The
filtrate was evaporated under reduced pressure to yield a dark green solid. The residue was
dissolved in minimum volume of dry CH3CN (~5 mL) and filtered via cannula to give a red
solution. This solution was then layered with Et2O (10ꢀ15 mL) and cooled to –20°C to yield the
product as red solid. Xꢀray quality crystals were grown by slow diffusion of Et2O into CH3CN or
CH2Cl2 solution of the complex.
1
CpCo(PCy2NBn2)(OTf)2, (6). Yield 63%. HꢀNMR (300 MHz, CD3CN) δppm: 1.28ꢀ1.60 (m, 11H),
1.83ꢀ1.91 (m, 7H), 2.39ꢀ2.46 (m, 5H), 2.74ꢀ2.84 (m, 4H), 3.14 (m, 2H), 3.31 (m, 2H), 3.72 (s,
2H), 3.91 (s, 2H), 5.71 (s, 5H), 7.12 (m, 2H), 7.35 (m, 5H), 7.48 (m, 3H), 7.71ꢀ7.83 (m, 10H).
31PꢀNMR (CD3CN): 45.7 (s) ppm.
1
CpCo(PCy2NPh2)(OTf)2, (7). Yield 70%. HꢀNMR (300 MHz, CD3CN) δppm: 1.30ꢀ2.15 (m, 19H),
2.46 (m, 3H), 3.01 (m, 2H), 3.26ꢀ3.30 (m, 2H), 3.62 (d, 2H), 4.01ꢀ4.08 (m, 4H), 5.87 (s, 5H),
31
7.12 (t, 2H), 7.25 (t, 4H), 7.56ꢀ7.41 (m, 4H), 7.71ꢀ7.83 (m, 10H). PꢀNMR (CD3CN): 46.9 (s)
ppm.
1
CpCo(PPh2NBn2)(OTf)2, (8). Yield 65%. HꢀNMR (300 MHz, CD3CN) δppm: 3.02 (m, 2H), 3.15
(m, 2H), 3.44 (m, 2H), 3.74 (s, 2H), 3.90 (m, 2H), 4.18 (s, 2H), 5.50 (s, 5H), 7.08 (m, 2H), 7.32
31
(m, 3H), 7.52 (m, 5H), 7.71ꢀ7.83 (m, 10H). PꢀNMR (CD3CN): 35.0 (s) ppm. SIMꢀMS/MeCN:
755.3 (100%) [MꢀTfO]+
20
ACS Paragon Plus Environment

Page 21 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
CpCo(PPh2NPh2)(OTf)2, (9). 9 was precipitated from CH2Cl2/Et2O mixture. Yield 78%. HꢀNMR
(300 MHz, CD2Cl2) δppm: 3.80 (m, 4H), 4.39 (d, 2H), 4.83 (m, 2H), 5.71 (s, 5H), 7.10ꢀ8.16 (m,
20H). 31PꢀNMR (CD3CN): 33.5 (s) ppm. SIMꢀMS: 727.3 (20%) [MꢀTfO]+, 578.4 (45%) [Mꢀ
2TfO]2+.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
CpCo(dppp)(OTf)2, (5'). Yield 84%. HꢀNMR (300 MHz, CD2Cl2) δppm: 2.70ꢀ2.81 (m, 6H), 5.74
(s, 5H), 7.29 (m, 4H), 7.51 (t, 4H), 7.74 (m, 12H). 31PꢀNMR (CD2Cl2: 30.8 (s) ppm. SIMꢀMS:
571.3 (100%) [Mꢀ2TfO+Cl]+
.
Instruments and methods
Electrochemical analysis was performed using a BioꢀLogic science instrument SP300
potentiostat. Electrochemical experiments in DMF (tetrabutylammonium tetrafluoroborate
nBu4NBF4, 0.1 M, as the supporting electrolyte) were carried out in a threeꢀelectrode
electrochemical cell using a glassy carbon working electrode (1.6 mm diameter), a platinum wire
as the auxiliary electrode and a Ag/AgCl/3M KCl reference electrode. For electrochemical
analyses, the DMF was directly used from a freshly opened bottle which contained ~0.1–0.2%
water. Only before H/D KIE studies, the DMF was dried over molecular sieves (3 Å) for 24–48
h. All potentials given in this work are reported with respect to Fc+/0 couple. The voltammograms
were referenced by addition of ferrocene as an internal standard after the final experiment. The
potential of the Fc+/0 couple was found to be 0.53 V vs. Ag/AgCl/3M KCl in DMF. During cyclic
voltammetry, the solutions were degassed using solvent saturated Ar and CO2. Cyclic
voltammograms under different partial pressure of CO2 were recorded by sparging the solution
with N2/CO2 gas mixtures (flow = 20 mL minꢀ1). A mass flow controller was used to mix N2 and
CO2 in various ratios. During the electrocatalysis, aliquots of water was added using a gasꢀtight
Hamilton syringe (1% v/v water/DMF = 0.56 M water in DMF). Bulk electrolysis experiments
and coulometry were carried out using a mercury pool cathode with an active surface area of
~1.77 cm2. The platinumꢀgrid counter electrode was placed in a separate compartment connected
by a glassꢀfrit. Prior to electrolysis, the electrolyte solution was saturated with CO2 and then the
electrochemical cell was kept closed and gasꢀtight during the electrolysis. Typically the volume
of electrolyte working compartment was 8 mL and that in the counter compartment was 2 mL.
Hydrogen produced during electrolysis was quantified with a PerkinꢀElmer Clarus 500 gas
chromatography equipped with a porapack Q 80/100 column (6’ 1/8”) thermostated at 40°C and a
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 36
1
2
3
4
5
6
7
8
9
TCD detector thermostated at 100°C. Carbon monoxide, methane and other volatile
hydrocarbons from the gas phase were analyzed using a flame induction detector (FID). Formic
acid concentrations were determined by ionic exchange chromatography (883 Basic IC,
Metrohm). UVꢀvis absorption spectra of compounds were recorded on an Agilent Technologies
Cary 60 UVꢀVis spectrometer in a cuvette with 1 cm path length.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Computational details. All calculations were carried out using TURBOMOLE (version 6.5)
package.96 Geometry optimizations of all the species have been carried out using the B3LYP
functional,97ꢀ99 complemented by the empirical dispersion scheme D3 developed by Grimme,100
in conjunction with the def2ꢀSV(P) basis set.101 Considering the size of the real system we opted
for a model system where the benzyl and cyclohexyl groups are replaced by methyl groups as
shown in Figure 5 in order to investigate various mechanistic options. Transition states (TS) were
characterized by the presence of one and only one imaginary frequency for the desired reaction
coordinate. Since the experiments were carried out using acetonitrile, the COSMO implicit
solvation scheme was employed to mimic the environment.102 Redox potentials were calculated
using a thermodynamic cycle, as described elsewhere.103 Previous studies have shown that
computed redox potentials depend quite significantly on the DFT functional that was used, and
the solvent model that was employed (±0.35 V).104,105 Multiple studies by Roy et al. and others
have reported similar or even larger discrepancies (±0.5 V) for the calculation of redox potentials
by using both hybrid and nonꢀhybrid functionals.106ꢀ108
Crystal Structure Analysis – X-ray crystallography: Diffraction data (Table S4) were
collected using an Oxford Diffraction XCallibur S Kappa area detector fourꢀcircle diffractometer
(MoꢀKα radiation λ = 0.71073 Å, graphite monochromator), controlled by the Oxford Diffraction
CrysAlis CCD software.109 Unique intensities with I > 10σ (I) detected on all frames using the
Oxford Diffraction RED were used to refine the values of the cell parameters.
The substantial redundancy in data allows analytical absorption corrections to be applied
using crystal shape determination for complex 7, 8 and 9’, and empirical absorption correction
for complex 6.The space group was determined from systematic absences, and it was confirmed
by the successful resolution of the structure. The structure was solved by charge flipping method
using superflip software for complex 7, ShelXT direct method resolution program for complex 6
and 8 and ShelXS for complex 9’, in Olex1.2 environment.109ꢀ113 All the atoms were found by
22
ACS Paragon Plus Environment

Page 23 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
difference Fourier syntheses. All nonꢀhydrogen atoms were anisotropically refined on F2 using
ShelXL program. Hydrogen atoms were fixed in ideal positions for complex 6, 8 and 9’ and
found by fourier transformation and refined isotropically for complex 7.
9
For complex 6; two different orientations were found in the crystal. Twinned crystal (ratio
of 0.87 and 0.13) was used to deconvoluate both contributions during the data reduction, but only
the first one was taken in account for structure resolution.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CCDC 1504837 (9’), 1504838 (7), 1504839 (6), and 1504840 (8) contain the
supplementary crystallographic data for this paper. These data can be obtained free of charge
from the Cambridge Crystallographic Data Center via www.ccdc.cam.ac.uk/datarequest/cif.
Associated Content
Supporting Information. Calculations for TOFs/TONs and overpotentials, additional
electrochemical data, DFT calculated structures and energies, crystallographic data.
Acknowledgements
We acknowledge support from the French National Research Agency (ANR, Carbiored ANRꢀ12ꢀ
BS07ꢀ0024ꢀ03; Labex program ARCANE, ANRꢀ11ꢀLABXꢀ0003ꢀ01 and DYNAMO, ANRꢀ11ꢀ
LABXꢀ0011) and from Fondation de l’Orangerie for individual Philanthropy and its donors.
References
(1)
Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.;
Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A.
R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L.
Chem. Rev. 2013, 113, 6621.
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
Qiao, J. L.; Liu, Y. Y.; Hong, F.; Zhang, J. J. Chem. Soc. Rev. 2014, 43, 631.
Costentin, C.; Robert, M.; Saveant, J. M. Chem. Soc. Rev. 2013, 42, 2423.
Rakowski Dubois, M.; Dubois, D. L. Acc. Chem. Res. 2009, 42, 1974.
Takeda, H.; Cometto, C.; Ishitani, O.; Robert, M. ACS Catal. 2016, 70.
Taheri, A.; Berben, L. A. Chem. Commun. 2016, 52, 1768.
Verma, S.; Kim, B.; Jhong, H. R.; Ma, S.; Kenis, P. J. ChemSusChem 2016, 9, 1972.
Schmidt, I.; Müller, K.; Arlt, W. Energy Fuels 2014, 28, 6540.
Mellmann, D.; Sponholz, P.; Junge, H.; Beller, M. Chem. Soc. Rev. 2016, 45, 3954.
Yu, X.; Pickup, P. G. J. Power Sources 2008, 182, 124.
23
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 36
1
2
3
4
5
6
7
8
9
(11)
Li, H.; Opgenorth, P. H.; Wernick, D. G.; Rogers, S.; Wu, T.ꢀY.; Higashide, W.; Malati,
P.; Huo, Y.ꢀX.; Cho, K. M.; Liao, J. C. Science 2012, 335, 1596.
(12)
(13)
(14)
Mondal, B.; Song, J.; Neese, F.; Ye, S. Curr. Opin. Chem. Biol. 2015, 25, 103.
Reda, T.; Plugge, C. M.; Abram, N. J.; Hirst, J. Proc. Natl. Acad. Sci. 2008, 105, 10654.
Mota, C. S.; Rivas, M. G.; Brondino, C. D.; Moura, I.; Moura, J. J. G.; González, P. J.;
Cerqueira, N. M. F. S. A. J. Biol. Inorg. Chem. 2011, 16, 1255.
(15)
(16)
Dobbek, H.; Svetlitchnyi, V.; Gremer, L.; Huber, R.; Meyer, O. Science 2001, 293, 1281.
Dobbek, H.; Gremer, L.; Kiefersauer, R.; Huber, R.; Meyer, O. Proc. Natl. Acad. Sci.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2002, 99, 15971.
(17)
Lubitz, W.; Ogata, H.; Rüdiger, O.; Reijerse, E. Chem. Rev. 2014, 114, 4081.
(18)
Simmons, T. R.; Berggren, G.; Bacchi, M.; Fontecave, M.; Artero, V. Coord. Chem. Rev.
2014, 270–271, 127.
(19)
(20)
DuBois, D. L. Inorg. Chem. 2014, 53, 3935.
Shaw, W. J.; Helm, M. L.; DuBois, D. L. Biochim. Biophys. Acta Bioenerg. 2013, 1827,
1123.
(21)
Dutta, A.; Lense, S.; Hou, J.; Engelhard, M. H.; Roberts, J. A. S.; Shaw, W. J. J. Am.
Chem. Soc. 2013, 135, 18490.
(22)
Dutta, A.; DuBois, D. L.; Roberts, J. A. S.; Shaw, W. J. Proc. Natl. Acad. Sci. 2014, 111,
16286.
(23)
RodriguezꢀMacia, P.; Dutta, A.; Lubitz, W.; Shaw, W. J.; Rudiger, O. Angew. Chem. Int.
Ed. 2015, 54, 12303.
(24)
2012, 5, 6480.
(25)
Seu, C. S.; Appel, A. M.; Doud, M. D.; DuBois, D. L.; Kubiak, C. P. Energy Environ. Sci.
Schneider, J.; Jia, H.; Kobiro, K.; Cabelli, D. E.; Muckerman, J. T.; Fujita, E. Energy
Environ. Sci. 2012, 5, 9502.
(26)
(27)
(28)
Song, J.; Klein, E. L.; Neese, F.; Ye, S. Inorg. Chem. 2014, 53, 7500.
Costentin, C.; Drouet, S.; Robert, M.; Saveant, J. M. Science 2012, 338, 90.
Costentin, C.; Robert, M.; Savéant, J.ꢀM.; Tatin, A. Proc. Natl. Acad. Sci. 2015, 112,
6882.
(29)
Chapovetsky, A.; Do, T. H.; Haiges, R.; Takase, M. K.; Marinescu, S. C. J. Am. Chem.
Soc. 2016, 138, 5765.
(30)
Ahn, S. T.; Bielinski, E. A.; Lane, E. M.; Chen, Y.; Bernskoetter, W. H.; Hazari, N.;
Palmore, G. T. R. Chem. Commun. 2015, 51, 5947.
(31)
Zhang, Y.; MacIntosh, A. D.; Wong, J. L.; Bielinski, E. A.; Williard, P. G.; Mercado, B.
Q.; Hazari, N.; Bernskoetter, W. H. Chem. Sci. 2015, 6, 4291.
(32)
Lilio, A. M.; Reineke, M. H.; Moore, C. E.; Rheingold, A. L.; Takase, M. K.; Kubiak, C.
P. J. Am. Chem. Soc. 2015, 137, 8251.
(33)
Bays, J. T.; Priyadarshani, N.; Jeletic, M. S.; Hulley, E. B.; Miller, D. L.; Linehan, J. C.;
Shaw, W. J. ACS Catal. 2014, 4, 3663.
(34)
Bielinski, E. A.; Förster, M.; Zhang, Y.; Bernskoetter, W. H.; Hazari, N.; Holthausen, M.
C. ACS Catal. 2015, 5, 2404.
(35)
Bielinski, E. A.; Lagaditis, P. O.; Zhang, Y.; Mercado, B. Q.; Würtele, C.; Bernskoetter,
W. H.; Hazari, N.; Schneider, S. J. Am. Chem. Soc. 2014, 136, 10234.
(36)
(37)
Borovik, A. S. Acc. Chem. Res. 2005, 38, 54.
Shook, R. L.; Peterson, S. M.; Greaves, J.; Moore, C.; Rheingold, A. L.; Borovik, A. S. J.
Am. Chem. Soc. 2011, 133, 5810.
(38)
Rigsby, M. L.; Wasylenko, D. J.; Pegis, M. L.; Mayer, J. M. J. Am. Chem. Soc. 2015, 137,
4296.
(39)
Fraze, K.; Wilson, A. D.; Appel, A. M.; Rakowski DuBois, M.; DuBois, D. L.
Organometallics 2007, 26, 3918.
24
ACS Paragon Plus Environment

Page 25 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(40)
Wilson, A. D.; Newell, R. H.; McNevin, M. J.; Muckerman, J. T.; Rakowski DuBois, M.;
DuBois, D. L. J. Am. Chem. Soc. 2006, 128, 358.
(41)
Galan, B. R.; Schöffel, J.; Linehan, J. C.; Seu, C.; Appel, A. M.; Roberts, J. A. S.; Helm,
M. L.; Kilgore, U. J.; Yang, J. Y.; DuBois, D. L.; Kubiak, C. P. J. Am. Chem. Soc. 2011, 133, 12767.
(42)
Wilson, A. D.; Newell, R. H.; McNevin, M. J.; Muckerman, J. T.; DuBois, M. R.;
DuBois, D. L. J. Am. Chem. Soc. 2006, 128, 358.
(43)
Fang, M.; Wiedner, E. S.; Dougherty, W. G.; Kassel, W. S.; Liu, T.; DuBois, D. L.;
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bullock, R. M. Organometallics 2014, 33, 5820.
(44)
(45)
Gennaro, A.; Isse, A. A.; Vianello, E. J. Electroanal. Chem. 1990, 289, 203.
Kaeffer, N.; Morozan, A.; Fize, J.; Martinez, E.; Guetaz, L.; Artero, V. ACS Catal. 2016,
6, 3727.
(46)
AnxolabéhèreꢀMallart, E.; Costentin, C.; Fournier, M.; Nowak, S.; Robert, M.; Savéant,
J.ꢀM. J. Am. Chem. Soc. 2012, 134, 6104.
(47)
(48)
(49)
(50)
Saveant, J. M. Chem. Rev. 2008, 108, 2348.
Costentin, C.; Drouet, S.; Robert, M.; Savéant, J.ꢀM. J. Am. Chem. Soc. 2012, 134, 11235.
Under such conditions, only 38% of the current corresponds to formic acid production
Chen, L.; Guo, Z.; Wei, X.ꢀG.; Gallenkamp, C.; Bonin, J.; AnxolabéhèreꢀMallart, E.; Lau,
K.ꢀC.; Lau, T.ꢀC.; Robert, M. J. Am. Chem. Soc. 2015, 137, 10918.
(51) Huan, T. N.; Andreiadis, E. S.; Heidkamp, J.; Simon, P.; Derat, E.; Cobo, S.; Royal, G.;
Bergmann, A.; Strasser, P.; Dau, H.; Artero, V.; Fontecave, M. J. Mater. Chem. A 2015, 3, 3901.
(52) The ꢀ1. 35 V vs Fc+/Fc value initially determined in ref 44 was refined to that value by
taking the value for the equilibrium potential in water given in E. Fujita, Coord. Chem. Rev. 1999, 185ꢀ
186, 373.
(53)
Costentin, C.; Passard, G.; Robert, M.; Savéant, J.ꢀM. Proc. Natl. Acad. Sci. 2014, 111,
14990.
(54)
Kang, P.; Cheng, C.; Chen, Z.; Schauer, C. K.; Meyer, T. J.; Brookhart, M. J. Am. Chem.
Soc. 2012, 134, 5500.
(55)
(56)
(57)
(58)
(59)
(60)
(61)
Kang, P.; Meyer, T. J.; Brookhart, M. Chem. Sci. 2013, 4, 3497.
Taheri, A.; Thompson, E. J.; Fettinger, J. C.; Berben, L. A. ACS Catal. 2015, 5, 7140.
Taheri, A.; Berben, L. A. Inorg. Chem. 2016, 55, 378.
Rail, M. D.; Berben, L. A. J. Am. Chem. Soc. 2011, 133, 18577.
Koelle, U.; Paul, S. Inorg. Chem. 1986, 25, 2689.
Wiedner, E. S.; Bullock, R. M. J. Am. Chem. Soc. 2016, 138, 8309.
DFT calculations do not evidence any interaction between a cobalt(III)ꢀhydride species
and CO2, even in the presence of water.
(62)
(63)
Schuster, C. H.; Diao, T.; Pappas, I.; Chirik, P. J. ACS Catal. 2016, 6, 2632.
Obligacion, J. V.; Bezdek, M. J.; Chirik, P. J. J. Am. Chem. Soc. 2017, DOI:
10.1021/jacs.6b13346.
(64)
Muckerman, J. T.; Skone, J. H.; Ning, M.; WasadaꢀTsutsui, Y. Biochim. Biophys. Acta
Bioenerg. 2013, 1827, 882.
(65)
Izutsu, K. AcidꢀBase Dissociation Constants in Dipolar Aprotic Solvents; Blackwell
Scientific: Oxford, U.K. 1990.
(66)
(67)
(68)
Schneider, J.; Jia, H.; Muckerman, J. T.; Fujita, E. Chem. Soc. Rev. 2012, 41, 2036.
Chen, X.; Jing, Y.; Yang, X. Chemistry – A European Journal 2016, 22, 8897.
Jeletic, M. S.; Mock, M. T.; Appel, A. M.; Linehan, J. C. J. Am. Chem. Soc. 2013, 135,
11533.
(69)
(70)
(71)
Zall, C. M.; Linehan, J. C.; Appel, A. M. ACS Catal. 2015, 5, 5301.
Bagherzadeh, S.; Mankad, N. P. J. Am. Chem. Soc. 2015, 137, 10898.
Galan, B. R.; Reback, M. L.; Jain, A.; Appel, A. M.; Shaw, W. J. Eur. J. Inorg. Chem.
2013, 2013, 5366.
25
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 26 of 36
1
2
3
4
5
6
7
8
9
(72)
(73)
Lim, C.ꢀH.; Holder, A. M.; Hynes, J. T.; Musgrave, C. B. J. Am. Chem. Soc. 2014, 136,
Lim, C.ꢀH.; Holder, A. M.; Hynes, J. T.; Musgrave, C. B. J. Phys. Chem. Lett. 2015, 6,
16081.
5078.
(74)
(75)
(76)
Riduan, S. N.; Zhang, Y.; Ying, J. Y. Angew. Chem. Int. Ed. 2009, 48, 3322.
Chakraborty, S.; Zhang, J.; Krause, J. A.; Guan, H. J. Am. Chem. Soc. 2010, 132, 8872.
Courtemanche, M.ꢀA.; Légaré, M.ꢀA.; Maron, L.; Fontaine, F.ꢀG. J. Am. Chem. Soc.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2014, 136, 10708.
(77)
(78)
Ménard, G.; Stephan, D. W. J. Am. Chem. Soc. 2010, 132, 1796.
Smieja, J. M.; Benson, E. E.; Kumar, B.; Grice, K. A.; Seu, C. S.; Miller, A. J. M.; Mayer,
J. M.; Kubiak, C. P. Proc. Natl. Acad. Sci. 2012, 109, 15646.
(79) Costentin, C.; Drouet, S.; Passard, G.; Robert, M.; Savéant, J.ꢀM. J. Am. Chem. Soc. 2013,
135, 9023.
(80)
(81)
Slater, S.; Wagenknecht, J. H. J. Am. Chem. Soc. 1984, 106, 5367.
Bolinger, C. M.; Sullivan, B. P.; Conrad, D.; Gilbert, J. A.; Story, N.; Meyer, T. J. J.
Chem. Soc., Chem. Commun. 1985, 796.
(82)
(83)
(84)
(85)
(86)
(87)
Ishida, H.; Tanaka, H.; Tanaka, K.; Tanaka, T. J. Chem. Soc., Chem. Commun. 1987, 131.
Ishida, H.; Tanaka, K.; Tanaka, T. Organometallics 1987, 6, 181.
Bolinger, C. M.; Story, N.; Sullivan, B. P.; Meyer, T. J. Inorg. Chem. 1988, 27, 4582.
Collin, J. P.; Jouaiti, A.; Sauvage, J. P. Inorg. Chem. 1988, 27, 1986.
Caix, C.; ChardonꢀNoblat, S.; Deronzier, A. J. Electroanal. Chem. 1997, 434, 163.
Pun, S.ꢀN.; Chung, W.ꢀH.; Lam, K.ꢀM.; Guo, P.; Chan, P.ꢀH.; Wong, K.ꢀY.; Che, C.ꢀM.;
Chen, T.ꢀY.; Peng, S.ꢀM. J. Chem. Soc., Dalton Trans. 2002, 575.
(88)
(89)
(90)
Machan, C. W.; Sampson, M. D.; Kubiak, C. P. J. Am. Chem. Soc. 2015, 137, 8564.
Witt, S. E.; White, T. A.; Li, Z.; Dunbar, K. R.; Turro, C. Dalton Trans. 2016, 52, 12175.
Franco, F.; Cometto, C.; Ferrero Vallana, F.; Sordello, F.; Priola, E.; Minero, C.; Nervi,
C.; Gobetto, R. Chem. Commun. 2014, 50, 14670.
(91)
Wiedner, E. S.; Chambers, M. B.; Pitman, C. L.; Bullock, R. M.; Miller, A. J. M.; Appel,
A. M. Chem. Rev. 2016, 116, 8655.
(92)
GinovskaꢀPangovska, B.; Dutta, A.; Reback, M. L.; Linehan, J. C.; Shaw, W. J. Acc.
Chem. Res. 2014, 47, 2621.
(93)
(94)
(95)
(96)
Coutard, N.; Kaeffer, N.; Artero, V. Chem. Commun. 2016, 52, 13728.
G. Märkl, V.; Jin, G. Y.; Schoerner, C. Tetrahedron Lett. 1980, 21, 1409.
King, R. B. Inorg. Chem. 1966, 5, 82.
Furche, F.; Ahlrichs, R.; Hättig, C.; Klopper, W.; Sierka, M.; Weigend, F. WIREs
Comput. Mol. Sci. 2014, 4, 91.
(97)
(98)
(99)
Becke, A. D. Phys. Rev. A 1988, 38, 3098.
Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(100) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104.
(101) Weigend, F.; Ahlrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297.
(102) Klamt, A.; Schuurmann, G. J. Chem. Soc. Perk. Trans. 2 1993, 799.
(103) Roy, L. E.; Jakubikova, E.; Guthrie, M. G.; Batista, E. R. J. Phys. Chem. A 2009, 113,
6745.
(104) Bhattacharjee, A.; Andreiadis, E. S.; ChavarotꢀKerlidou, M.; Fontecave, M.; Field, M. J.;
Artero, V. ChemꢀEur J 2013, 19, 15166.
(105) Canaguier, S.; Fourmond, V.; Perotto, C. U.; Fize, J.; Pecaut, J.; Fontecave, M.; Field, M.
J.; Artero, V. Chem. Commun. 2013, 49, 5004.
(106) Roy, L. E.; Batista, E. R.; Hay, P. J. Inorg. Chem. 2008, 47, 9228.
(107) Roy, L. E.; Jakubikova, E.; Guthrie, M. G.; Batista, E. R. J. Phys. Chem. A 2009, 113,
6745.
26
ACS Paragon Plus Environment

Page 27 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
(108) Reiher, M.; Salomon, O.; Hess, B. A. Theor. Chem. Acc. 2001, 107, 48.
(109) Palatinus, L.; Chapuis, G. J. Appl. Crystallogr. 2007, 40, 786.
(110) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.; Puschmann, H. J. Appl.
Crystallogr. 2009, 42, 339.
(111) Sheldrick, G. Acta Cryst. A 2008, 64, 112.
(112) Sheldrick, G. M. Acta Cryst. 2015, A71, 3.
(113) Sheldrick, G. M. Acta Cryst. 2015, C71, 3.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
27
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 28 of 36
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of cobaltꢀdiphosphine complexes
Figure 1. ORTEP diagrams of cations in 6 (left) and 7 (left) with ellipsoids shown at 50%
probability. Hydrogen atoms are omitted for clarity.
28
ACS Paragon Plus Environment

Page 29 of 36
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Cyclic voltammograms of 1 (red, 0.5 mM), 2 (orange, 0.5 mM), 3 (green, 0.6 mM), and
4 (blue, 0.6 mM) under Ar (ν = 0.1 Vsꢀ1 in DMF, 0.1 M NBu4BF4, glassy carbon electrode). The
Yꢀaxis is offset for clarity.
Figure 3. Cyclic voltammograms of 1 (1 mM) under Ar saturation (black), under CO2 saturation
(red), and under CO2 saturation in the presence of increasing amount of water; 0.33% (v/v):
orange, 0.67% (v/v): green, 1% (v/v): blue. The inset shows the influence of the concentration of
water on the catalytic current. Voltammograms were recorded at 0.1 Vsꢀ1 in 0.1 M NBu4BF4 in
DMF using a glassy carbon electrode.
29
ACS Paragon Plus Environment