Cobalt Complex with Redox-Active Imino Bipyridyl Ligand for Electrocatalytic Reduction of Carbon Dioxide to Formate

Article abstract of DOI:10.1002/cssc.201800136

An imino bipyridine cobalt(II) complex was developed for the electrocatalytic reduction of CO₂ to formate in acetonitrile with a faradaic efficiency of approximately 80 %. For comparison, a symmetric bis-imino pyridine complex showed lower catalytic activity because of less conjugation in the system. Cyclic voltammetry, electron paramagnetic resonance and IR spectroscopy studies provided mechanistic details and the structures of the key intermediates. DFT calculations confirmed the role of large π-conjugated groups for stabilizing key intermediates through electronic conjugation.

Full text of DOI:10.1002/cssc.201800136

Accepted Article
Title: Cobalt Complex with Redox Active Imino Bipyridyl Ligand for
Electrocatalytic Reduction of Carbon Dioxide to Formate
Authors: Fangwei Liu, Jiaojiao Bi, Yuanyuan Sun, Shuping Luo, and
Peng Kang
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201800136
Link to VoR: http://dx.doi.org/10.1002/cssc.201800136
A Journal of
www.chemsuschem.org

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
Cobalt Complex with Redox Active Imino Bipyridyl Ligand for
Electrocatalytic Reduction of Carbon Dioxide to Formate
Fang-Wei Liu,[a,c] Jiaojiao Bi,[a,c] Yuanyuan Sun, Shuping Luo, Peng Kang*
[d]
[d]
[a,b,c]
Abstract: An imino bipyridine cobalt (II) complex was developed for
electrocatalytic reduction of CO to formate in acetonitrile with ca. 80%
for CO
with phosphorus ligands. Little is reported for cobalt complexes
-to-formate conversion are Ir,[3] Co,[2d] and Ni[6] complexes
2
2
faradaic efficiency. For comparison, a symmetric bis-imino pyridine
complex showed decreased catalytic activity due to less conjugated
system. CV, EPR and IR spectroscopic studies provided mechanistic
details and structures of key intermediates. DFT computation
confirmed the role of redox active ligand for stabilizing key
intermediates through electronic conjugation.
that can selectively reduce CO
ligands.
2
to formate using redox active
Introduction
Anthropogenic CO
changes. Electrochemical approach could transform industrial
CO
stream into value-added chemicals.[1] Recently,
electrocatalytic reduction of CO to carbon monoxide (CO) and
2
emission has led to observable climate
2
2
formate/formic acid have been investigated,[2] and formic
acid/formate could serve as hydrogen storage material, deicing
agent, drilling completion fluid and precursor to methanol, etc.
Symmetric bis(imino)pyridine iron and cobalt complexes
ArPDI)MCl
2
were discovered by Brookhart, and Gibson,[8] and
[
7]
(
Efficient and selective electocatalysts for conversion of CO
formate are desirable.[3] Yet, a major challenge is that the
reduction of CO is kinetically restrained with multiple electron
2
to
[
9]
other researchers for ethylene polymerization, and recently
regain focus in base-metal catalysis.[ However, the above π-
conjugated system is not sufficiently large for electrochemical
purpose. Herein, a new cobalt complex with large π-conjugated
5]
2
transfer progresses, and often accompanied with highly
competitive hydrogen evolution reaction.
Transition metal complexes are unique catalysts because they
can store and transfer multiple electrons, thus circumventing high
2
system was developed for CO electroreduction to formate.
Compared to bis-terpyridine and bis(imino)pyridine iron and
cobalt complexes[5, 10], single tridentate ligand in this study allows
more open sites, and large sterics of the diisopropyl moieties can
prevent dimerization of the active hydride intermediate,[11] which
could contribute to increased catalytic reactivity. The
electrocatalytic reactivity and mechanism was also explored in
detail.
energy CO
2
radical intermediate. Molecular catalysts using Ru, Ir,
Pd and Re have been investigated in converting CO
2
to CO and
formate. Catalysts based on earth abundant transition metals,[
such as Fe, Co and Ni, could provide inexpensive materials for
large scale use.[5] However, these catalysts are quite labile and
4]
2 2
prone to generate H , causing low selectivity for CO reduction.
Redox active ligands with large π-π conjugation can facilitate
electron transfer and storage, thus accelerating electrocatalytic
kinetics. Known cobalt complexes with redox active ligands prefer
Results and Discussion
to reduce CO
2
to CO.[2a-c, 2e, 2f] Reported transition metal catalysts
Synthesis and Physical Properties
[
a]
F.-W. Liu, J. Bi, Prof. P.Kang
Key Laboratory of Photochemical Conversion and Optoelectronic
Materials,Technical Institute of Physics and Chemistry, Chinese
Academy of Sciences, 29 Zhongguancun East Rd, Beijing 100190,
China
E-mail: pkang@mail.ipc.ac.cn
Phone: +86-10-82543570 Fax: +86-10-62554670
Prof. P.Kang
School of Chemical Engineering and Technology, Tianjin University,
[b]
[c]
[d]
135 Yaguan Rd, Tianjin 300350, China
F.-W. Liu, J. Bi, Prof. P.Kang
University of Chinese Academy of Sciences, 19A Yuquan Rd,
Beijing 100049, China
Y. Sun, Prof. S.Luo
State Key Laboratory Breeding Base of Green Chemistry-Synthesis
Technology, Zhejiang University of Technology, Chaowang Road
18, Hangzhou 310014, China
Supporting information for this article is given via a link at the end of
the document
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
Scheme 1. Synthetic route for complexes 1MeCN and 1OAc
.
O1-Co-O2 86.69(15), O1-Co-N1 102.91(15), O1-Co-N2 177.94(16), O1-Co-N3
105.99(15), O1-Co-N4 87.09(16), O2-Co-N1 89.02(15), O2-Co-N3 89.14(15),
N1-Co-N3 150.89(15), N2-Co-O2 91.86(15), N2-Co-N1 75.59(16), N2-Co-N3
7
5.43(15), N2-Co-N4 94.24(17), N4-Co-O2 172.21(16), N4-Co-N1 87.79(17),
The ligand L1 was synthesized as shown in Scheme 1. 4,4’-
dimethyl-2,2’-bipyridine was oxidized by H to afford N-oxidized
N4-Co-N3 97.09(16). 1OAc: Co1-O1 2.800, Co1-O2 1.984(3), Co1-O3 3.191,
2
O
2
Co1-O4 1.964(3), Co1-N1 2.165(3), Co1-N2 2.046(3), Co1-N3 2.316(3).; Bond
angles: O2-Co1-N1 100.64(13), O2-Co1-N2 142.29(13), O2-Co1-N3
compound 3. 3 was cyanated at the 6-position to yield 4. 4 was
then functionalized by Grignard addition with MeMgBr, and
subsequent hydrolysis gave the corresponding ketone 5.[12] The
bipyridine ligand L1 and symmetrical bis(imino) pyridine ligand L2
was prepared in 85% yield by acid-catalyzed condensation of 5
101.42(13), O4-Co1-O2 104.35(13), O4-Co1-N1 103.73(12), O4-Co1-N2
112.88(13), O4-Co1-N3 89.61(12), N1-Co1-N3 150.41(12), N2-Co1-N1
76.69(12), N2-Co1-N3 73.78(12). Complete crystallography data are collected
in SI.
with 2,6-diisopropylaniline.[13]
MeCN
and 2MeCN were synthesized by treating
CN) ](CF SO in 1:1 ratio in
tetrahydrofuran at room temperature. Complex 1
was similarly synthesized using Co(OAc) as metal source. Pale-
Complexes 1
Electrochemistry under Ar
ligands L1 and L2 with [Co(CH
3
4
3
3 2
)
OAc
(Scheme 1)
2
red crystal of 1MeCN suitable for X-ray crystallography was
obtained by slow diffusion of diethyl ether into acetonitrile solution
at room temperature.[14] The crystal structure of 1MeCN (Figure 1a)
shows that the cobalt atom was coordinated by three nitrogen
atoms of L1 in meridial mode, with one acetonitrile axially
coordinated, and two water molecules axially and equatorially
coordinated. The water molecules were found in the crystal
2
probably due to adventitious amounts of H O in MeCN. The
OAc
crystal of 1
in Figure 1b shows that two acetate anions
coordinated to the cobalt center together with ligand L1. Acetate
is a more coordinating ligand and it substituted MeCN in 1OAc. It
is interesting that an acetate anion assumed κ-2 coordination,
which could serve as a faithful analog for formate coordination.
1MeCN is a paramagnetic species as observed with broadened
NMR resonances. Magnetic susceptibility was measured for
1MeCN using a SQUID magnetometer from 100 to 300 K. Figure S1
Figure 2. CVs of 1MeCN (a) and 2MeCN (b) under Ar (black) and CO
(red, 1 atm)
2
4 6
in MeCN. Conditions: 2 mM complex in 0.1 M nBu NPF /MeCN; glassy carbon
working electrode, Pt wire counter electrode, Ag/AgNO
scan rate 50 mV/s.
3
reference electrode,
Cyclic voltammograms (CVs) of 1MeCN and 2MeCN were
shows the magnetic susceptibility χ
temperature. Between 100 and 215 K, the χ
M
and the product χ
M
T vs
2
obtained under Ar or 1 atm CO in acetonitrile solutions at room
temperature (Figure 2). 1MeCN displayed two reversible redox
M
vs 1/T plot (Figure
S1; curve for 1MeCN) remained almost constant at 1.53, which
corresponds to the spin-only value for μ = 6.21; from 215 to 300
K, the spin-only value changed to μ = 7.36. The above results are
waves at E1/2 = -0.33V (wave I) and -1.08V (wave II) vs. NHE
under Ar (Figure 2a).[16-17] Wave
I was electrochemically
II
I
reversible corresponding to Co /Co redox couple, and wave II
was attributed to L1 ligand-centered redox process. 2MeCN also
showed two redox waves similar to 1MeCN, but both waves shifted
more positively in potential, at E1/2= 0.08V for wave I and -0.80V
for wave II (Figure 2b). The positive shift in redox potentials was
presumably due to decreased electron-donating ability of L2 than
L1. The CV of 1OAc under Ar showed two waves at E1/2= -0.81V
II
similar to reported hexacoordinate high-spin Co complexes with
N-donor ligands (μeff > 4.9μ
coordinate Co complexes (μeff = 1.79–2.13 μ
the equation, μ =g√s (s+1) , and the g value of 2.18 for 1
e
),[15] other than low-spin six-
II
).[16] According to
B
MeCN
obtained from the solution EPR spectrum (Figure 8a), the electron
II
spin s value was calculated to be 3/2 for Co , consistent with high-
spin cobalt complexes.
for wave I and -1.26V for wave II (Figure S2), and both shifted
MeCN
more negatively compared to 1
, due to the coordination of
-
OAc to the Co center.
CVs at varied scan rates (10-500 mV/s) under Ar were used to
probe the nature of wave I and II (Figure 3a). Both waves were
reversible at all scan rates. When normalized for the scan rate
(i
cat/ʋ1/2)(Figure 3b), the cathodic peak currents (ip,c) of wave I and
II varied linearly with the square root of the scan rate (ʋ1/2) from
1
0 to 500 mV s−1 under Ar (Figure 3c,3d), consistent with
MeCN
diffusional reduction of 1
as shown by the Randles–Sevcik
equation (1) and thus these waves were not catalytic under Ar.
1/2
1/2
i = 0.4463(F /RT ) np3/2ADCo [Co]υ1/2…………………. (1)
3
d
In equation (1), F is Faraday's constant, R is the universal gas
constant (J K-1 mol ), n
-1
(=1 for both wave I and II) represents the
p
Figure 1. Crystal structures of 1MeCN (a) and 1OAc (b). Selected bond lengths [Å]
and angles [°] for 1MeCN and 1OAc. 1MeCN: Co-O1 2.030(3), Co-O2 2.129(4), Co-
N1 2.179(4), Co-N2 2.052(4), Co-N3 2.191(4), Co-N4 2.113(5). Bond angles:
number of electrons transferred, T is the temperature (K), A is the
2
electrode area (cm ), D represents the diffusion coefficient of
Co
2
-1
-1 [18]
the complex (cm s ), and ʋ is the scan rate (V s ).
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
a new wave III appeared at Ep,c = −1.23 V vs NHE with current of
ca. 2.5-fold of that of wave I. The above observations suggested
MeCN
that catalytic processes occur at waves II and III. 2
displayed current enhancement at wave II under CO
also
2
, although
the catalytic current enhancement was comparably lower, for ca.
MeCN
only 1.2-fold (Figure 2b), suggesting that 2
was kinetically
MeCN
much less efficient than 1
Under CO
remained diffusional in nature, however, wave II became
electrocatalytic. When normalized for the scan rate (icat/ʋ1/2
Figure 5b), the current of wave II became higher with decreasing
.
2
with CV scan rates varied (Figure 5a), wave I
)
(
−1
scan rate from 500 to 10 mV s , which pointed to an
electrocatalytic process. This was also the case for wave III.
Another observation was that the ratio of peak currents of wave
II/wave III changed upon variation of the scan rates. At slower
scan rates, the current of the wave II increased relative to that of
wave III, and vice versa for faster scan rates. We postulated that
waves II and III were two separate catalytic processes instead of
sequential processes, and it was possible that wave III arised from
coordination of generated formate to the Co center as the peak
potential was consistent with the wave II of 1OAc with acetate
coordination.
Figure 3. Electrochemical study of 1MeCN at varied scan rates (10–500 mV/s) in
MeCN
MeCN under Ar (0.1 M nBu
4
NPF
6
, glassy carbon). (a) CVs of 2 mM 1
in
MeCN under Ar; (b) scan rate normalized (i/ʋ /2) CVs from (a); (c) plot of peak
1
current i
d
of wave I under Ar vs the square root of the scan rate (ʋ 1/2, υ in V/s);
d) plot of peak current i of wave II under Ar vs the square root of the scan rate
, ʋ in V/s). ▽: Ferrocene standard.
(
(
d
1/2
ʋ
Figure 5. Electrochemical study of 1MeCN at varied scan rates (10–500 mV/s) in
MeCN under 1 atm CO (0.1 M nBu NPF , glassy carbon). (a) CVs of 2 mM
2 4 6
MeCN
1
at various scan rates (10–500 mV/s) in MeCN under 1 atm CO
2
; (b) Scan
rate normalized (i/ʋ1 ) CVs of 2 mM 1MeCN at various scan rates. ▽: Ferrocene
/2
standard.
Figure 4. CVs of 1MeCN under Ar starting from 1.05 V to various cathodic limits.
To -1.45 V vs. NHE (blue, c); to -0.95V (black, b); to 0.05V (red, a). Conditions:
2
mM 1MeCN in 0.1 M nBu
counter electrode, Ag/AgNO
4
NPF
6
/MeCN; glassy carbon working electrode, Pt wire
The catalytic kinetics were calculated from CVs at scan rates of
3
reference electrode, scan rate 50 mV/s.
0-50 mV s-1 using foot-of-wave method reported by Savéant et
1
[
20]
al. (Figure S3).
Foot of the wave analysis has been used to
catalytic efficiency of homogeneous
evaluate
the
CV of 1MeCN in the presence of 0.1% v/v water showed a new
[20]
electrocatalysts, and was applied to our study, especially when
wave at Ep,a = 0.74V vs NHE (Figure 4, trace c), suggesting rapid
protonation of the Co intermediate to generate cobalt hydride
species [L1Co H] (1 ), which was reactive toward CO
hydride wave appeared after the cathodic scans from 1.05V to
0
a catalytic current plateau cannot be obtained in the CVs. In
equations (2)-(5), icat is given by eqn (3) with ncat = 2 for 2e-
I
II
+
H
.[19] No
2
2
p
reduction of CO to form formate and n = 1 for wave II.
[CO2]
dt
Rate = –d
=kcat[Co]=kCO2[Co][CO2]
(2)
.05V or to -0.95V vs. NHE (Figure 4, trace a and b) that did not
-
MeCN
icat=(ncatFA)[Co](kcatDCo)1/2= (ncatFA)[Co](kCO2DCo[CO ])
1/2
passed wave II, indicating only upon the 2e reduction of 1
could 1 be formed.
2
H
(
3)
icat
i d
= 2.242(kcatRTncat2/np3Fυ)1/2
Electrochemistry under CO
2
(4)
(5)
When the solution of 1MeCN was saturated with CO
, wave I kept
2
icat
i d
= 2.242(kCO2RT[CO ]ncat2/np3Fυ)1/2
essentially the same, but the peak current of wave II was
catalytically enhanced for ca. 2.2-fold (Figure 2a)[4d]. Additionally,
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
2
The observed turnover frequency kcat under 1 atm CO was
calculated as 11.0(1.5) s-1 using CVs at the 10-50 mV s scan
rates (Figure S3). 1MeCN is the second efficient Co catalyst in
generating formate, compared to reported P2N2 Co complexes
-1
with pendant amines[2d]
.
Water was used as an exogenous proton source to generate
formate. Figure 6a showed CVs of adding water of 0-3% v/v under
Ar in MeCN. The wave I of 1MeCN shifted more negatively
compared to under formally dry conditions, presumably due to
water coordination to the Co center to form aqua complexes.[3a]
Wave II kept essentially unchanged upon adding water,
suggesting that 1MeCN was not a catalyst toward water reduction
to form hydrogen. Still, under CO
increased for 26% and 12% when 3% v/v water was added
Figure 6b), suggesting that moderate water addition can
accelerate CO reduction catalysis. It is possible that water
2
, both wave II and wave III
(
2
substitutes the coordinated formate increasing the overall
catalytic efficiency.
Figure 7. Controlled potential electrolyses and product distribution. (a) CVs of
1MeCN under 1 atm CO
(20 successive scan cycles). Scan rate = 100 mV/s. (b)
Time course of CPE and formate efficiencies with 1% added water under CO
in 0.1 M nBu NPF /CH CN. Applied potential, −1.25 V vs. NHE. (c) Faradaic
2
2
4
6
3
efficiencies and product distribution at various applied potentials. (d) Faradaic
efficiencies and product distribution at -1.05 V vs. NHE with different amount of
MeCN
2
added water. 2 mM 1
, glassy carbon working electrodes, 0.071 cm for CV
, 0.1 M nBu NPF /CH CN, room temperature.
and 0.5 cm2 for CPE, 1 atm CO
2
4
6
3
Under formally dry conditions, as the applied potential was
more negative from -1.05 to -1.35V, formate selectivity decreased,
and CO selectivity increased from 20% to 30% (Figure 7c).
Adding water from 0-4% reduced CO selectivity and increased
Figure 6. CVs of 2 mM 1MeCN with 0–3% of water (v/v) added in MeCN under Ar
2 4 6
(a) and CO (b). (0.1 M nBu NPF /MeCN, glassy carbon, 50 mV/s)
formate selectivity, but also generated more H
Figure 7d). As much as 80% formate was obtained at −1.15 V in
MeCN upon CPE of 5.6 h. Control experiments using
Co(CH CN) (CF SO and < 2%
instead of 1MeCN yielded 96 % H
CO, and no formate was found. The average current density was
2
from 0 to 18%
(
Controlled Potential Electrolysis
3
4
3
3
)
2
2
Complex 1MeCN showed decent stability in eleclctrocatalysis. In
Figure 7a, peak currents of wave II and wave III kept nearly
constant with 20 cycles of scans.[21] Controlled potential
electrolysis (CPE) showed fairly stable current density at ca. 1.5
mA/cm2 in average for 9 h at -1.25V vs NHE (Figure 7b). To
-2
MeCN
significantly less (ca. 0.01 mA cm ) than using 1
(0.36 mA
-2
cm ). Those evidences suggested that an intact L1 coordinated
cobalt complex was involved in the electrocatalytic process.
investigate the CO
2
reduction products, the headspace of the
Catalytically Relevant Intermediates
electrochemical cell was analyzed by gas chromatography and
liquid phase by NMR. The H-NMR spectrum of post electrolysis
Low-valent Co species were synthesized by preparative
electrochemical reduction of 1MeCN. By applying constant potential
1
I
I
0
solutions revealed formate product. The new emerging wave I’ of
at -0.85V and -1.45V respectively, formal Co species 1 and Co
1MeCN (Figure 7a) under CO
was similar to the wave I of 1OAc
0
species 1 were generated in MeCN under Ar. EPR spectra were
measured for the above Co species. The 90 K EPR spectrum of
1MeCN in frozen MeCN solution (Figure 8a) showed a rhombic
signal at g=2.18 with hyperfine coupling, typical for high-spin
2
(
Figure S2), which suggested that formate product could
coordinate to the Co center to form an adduct. With 1% added
water, the formate yield was stable at >70% over time (Figure 7b).
The XPS spectrum of the working electrode after CPE using 1MeCN
II
7
II
square planar Co complexes (d , s = 3/2). Low-spin (s = 1/2) Co
(
Figure S4) showed negative response of cobalt element,
complexes could be expected only in the presence of a sufficiently
strong ligand field (Δ≥15 000 cm ), which is required for 2E
ground state,[12b],[22] and neither acetonitrile or aqua ligand in 1MeCN
is strong enough.
suggesting that 1MeCN did not decompose onto the electrode
during the CPE. The NMR spectrum of post CPE solution upon
acidic workup showed that the ligand was intact with no
observable degradation (Figure S5).
−1
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
moiety, suggesting that the second reduction was ligand based
(
Figure S6). In 1CO2 species, the spin density was redistributed
from the reduced ligand to the CO
formation of carboxylate species.[4d]
2
moiety, suggesting the
The catalytic mechanism was proposed as shown in Scheme
0
2
. 1MeCN was electrochemically reduced twice to form 1 which
0
II
was catalytically active. 1 can be protonated by water to form Co
hydride species 1 (wave II). 1 was not sufficiently reactive with
H
H
water to form H
2
as shown in CVs, yet sufficiently hydridic to insert
F
CO
2
to form formate adduct 1 . The formate dissociated from the
MeCN
Co center with the aid of water to regenerate 1
, completing
the formate pathway
Figure 8. Spectra of catalytically relevant Co species. (a) EPR spectra of 10
Alternatively, under formally dry conditions, 10 could
mM 1MeCN (black), 1I (red) and 1 (blue) in 0.1 M nBu
0
4 6
NPF /CH
3
CN. Microwave
CO2
frequency 9.43 GHz, 10 mW, modulation frequency 100 KHz, T = 90 K. (b) IR-
nucleophilically attack CO
pathway in Scheme 2. The formation of 1
reduction at the metal center or the ligand at the same potential
followed by proton transfer to form a transient [LCo (COOH)]
1COOH species (Scheme 2). The decomposition of 1COOH
2
to form 1 , introducing the CO
SEC spectra of 1MeCN (black), 1 (dash) and 1 (red) in 0.1 M nBu
0
CO2
CO2
4 6
NPF /CH
3
CN.
enabled a second
2
Electrode, glassy carbon (0.071 cm ); for reference, an IR spectrum without
complex 1MeCN under CO
is shown in dash dot.
2
I
0
presumably yielded the CO product. However, in CPE
experiments with water added, this became a comparably minor
pathway as 1 was rapidly protonated to form 1 .
I
EPR spectrum of 1 species (Figure 8a) was basically silent with
MeCN
I
a few residual signals (<1%) of 1
, suggesting that 1 is low
0
H
8
I
[2e]
0
spin d Co species. The 1 species showed a free radical signal
In the proposed mechanism, the CO and formate pathways
[
23]
at g=2.00, consistent with an organic radical,
electronic structure of 1 should be (L
and thus the
0
0
share the same Co species 1 , and redox active imino bipyridial
ligand is effective in delocalizing electron density from the Co
center to the ligand, thus making the Co center less electron rich.
0
·-
I
1
)Co species instead of
0
true L
1
Co species.
The 10 species reacted with CO
species 1CO2 under anhydrous conditions. In the infrared
to yield a new intermediate
2
H
Compared to phosphine based Co complexes, 1 species is not
sufficiently hydridic, which reduces the chance for undesired H
2
MeCN
spectroscopy (IR) spectra for 1
at 1634 cm for imine group.
, ν(C=N) bands were located
evolution in the presence of added water and enhances the
formate selectivity.[19]
-1
[13a]
The IR spectrum of 1CO2 showed
feature at 1728 cm−1 (Figure 8b), consistent with C=O vibration of
a Co carboxylate [LCo (COO )] species (Scheme 2).[2e, 4b, 24]
II
− 0
Conclusions
DFT Calculations and Proposed Mechanism
The Co complex of redox active imino bipyridial ligand is efficient
in electrochemical reduction of CO
2
to formate in acetonitrile with
80% selectivity. The Co complex does not catalyze water
2
reduction to generate H . Adding water can promote the formation
of formate and reduce CO formation, although leading to slightly
more H evolution. Catalytically relevant Co intermediate species
2
were synthesized, and based on these intermediates the catalytic
mechanism was proposed. The function of redox active ligand is
critical in contributing the CO reduction efficiency and selectivity,
2
and translation of the catalytic reactivity into aqueous phase is
currently underway.
Experimental Section
Materials and Methods
2
Scheme 2. Proposed mechanism for electrocatalytic reduction of CO .
by 1MeCN
All chemicals were purchased from commercial sources if not
mentioned otherwise. 4,4'-dimethyl-2,2'-bipyridine was purchased
[
25]
[12a]
[12]
[26]
Density Functional Theory (DFT) calculations were performed
to suggest the electronic configuration of critical intermediates.
from Acros. Compounds
3
,
4
,
5
,
L2
were synthesized according to
Particularly, the synthesis of
is similar to Fe(CH CN) (OTf) in the
and
[27]
Co(CH
published procedures.
Co(CH CN) (CF SO
3 4 3 3 2
CN) (CF SO )
The optimized structures of 1MeCN, 1, 1 and 1
I
0
CO2
species were
3
4
27]
3
3
)
2
3
4
2
I
shown in Figure S6. The HOMO of 1 mainly consisted of Co dz2
orbitals, indicating that the first reduction of 1MeCN occurred at the
reference.[ Acetonitrile was of HPLC grade and further purified
by a solvent purification system. Deionized water was further
purified by using a Master-S15 UV Water Purification system.
0
metal center. The spin density of 1 mainly distributed over the L
1
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
NMR spectra were recorded with a Bruker Advance 400
evaporation, further purification was carried out by column
spectrophotometer.
High-resolution
mass
spectrometry
chromatography using CH
afford 308 mg of product L1. Yield: 80 %. H NMR (400 MHz,
CDCl ): δ 8.57 (d, J = 4.9 Hz, 1H, Py-CH), 8.38 (d, J = 6.0 Hz, 2H,
Py-CH), 8.24 (s, 1H, Py-CH), 7.20 (d, J = 7.6 Hz, 2H, Ar-CH), 7.18
– 7.10 (t, 1H, Ar-CH), 7.15 (d, J = 4.0 Hz, 1H, Py-CH) 2.91 – 2.74
2 2 3
Cl / CH OH (100: 3, v / v) as eluent to
1
experiments were performed with a Bruker Daltonics Apex IV
spectrometer. Infrared-spectrum was recorded by Varian 3100
FT-IR. Electron paramagnetic resonance spectra were obtained
on a Varian-E-500 EPR spectrometer. The spectra were recorded
3
for solutions of the complexes in CH
3
CN solvent at liquid nitrogen
(m, 2H,CH
CH C=N), 1.24 – 1.12 (m, 12H, CH
CDCl ): δ 167.5 (Cq, s, C=N-Ar), 156.1 (Cq, s, Py-C), 155.7 (Cq,
s, Py-C), 155.1 (Cq, s, Py-C), 149.1 (CH, s, Py-C), 148.7 (Cq, s,
Py-C), 148.1 (Cq, s, Py-C), 146.7 (Cq, s, Ar-C-N), 136.1 (Cq, s,
Ar-C), 124.9 (CH, s, Py-C), 123.7 (Cq, s, Ar-C), 123.2 (CH, s, Py-
C), 123.1 (CH, s, Ar-C), 122.1 (CH, s, Ar-C), 121.9 (CH, s, Py-C),
3 3 3
CHCH ), 2.48-2.55 (d, 6H, CH -Py), 2.36 (s, 3H,
13
temperature (77 K). Crystal structures were solved with direct
methods and refined with a full-matrix least-squares technique,
using the SHELXS software package. Mercury (CSD software)
was used for crystal structure visualization. The Co elements on
the glassy carbon electrode were investigated using X-ray
photoelectron spectroscopy (XPS; ESCALab250Xi, Thermo
Scientific).
3
3 3
CHCH ). C NMR (400 MHz,
3
28.4 (CH, s, CH
CH, s, CH -Py), 17.6(CH, s, CH
: 386.259074; found: 386.259151, error: 0.2
3
CHCH
3
), 23.2-23.4 (CH, d, CH
3 3
CHCH ), 21.5
(
3
3
-C=N). HR-ESI-MS: m/z calcd
+
for [M]
ppm; calcd for [M+Na]
08.240803, error: 0.5 ppm.
26 32 3
C H N
Electrochemical experiments were performed using a CHI 660E
potentiostat (CH Instruments, Inc.). The three-electrode system
consisted of a glassy carbon working electrode, a coiled Pt wire
+
26 31 3
C H N Na: 408.241019; found:
4
counter electrode, and a Ag/AgNO
3
reference electrode (CHI, 10
mM AgNO , 0.1 M nBu NPF in acetonitrile, 0.55 V νs NHE) in an
3
4
6
Preparation of 1MeCN: To 20 mL dry THF solution of L1 (77 mg,
0.20 mmol) was added Co(CH CN) (CF SO (104 mg, 0.20
mmol) and stirred at room temperature overnight. The solvent was
airtight, glass frit-separated two compartment cell. Prior to each
measurement, glassy carbon electrode (CHI, 7.1 mm ) was
3
4
3
3 2
)
2
polished with 0.05-µm alumina slurry to obtain a mirror surface
and it was then sonicated and thoroughly rinsed with ultrapure
water and acetone. For cyclic voltammogram experiments,
working and counter electrodes were separated from the
reference electrode. For controlled potential electrolyses,
reference and counter electrodes were separated from the
working electrode. Ferrocene was added at the end of the
experiment and the potential was converted relative to NHE by
adding 0.55V following the literature.[28] Gaseous product was
analyzed using an SRI 8610C GC with a molecular sieve column
removed with reduced pressure to afford 92 mg of yellow product.
1
Yield: 95 %. H NMR (400 MHz, CD
3
CN): δ 8.51 (s, 3H), 5.97 (s,
1H), 5.73 (s, 3H), -1.33 (s, 6H), -2.13 (s, 3H), -8.96 (s, 6H). HR-
2+
ESI-MS: m/z calcd for 1/2[M-CH
3
CN-2H
2
O]
26 3
C H31CoN :
222.091948; found: 222.091992, error: 0.2 ppm; calcd for 1/2[M -
2
+
2H
ppm; calcd for 1/2[M+2H]
261.613047, error: -0.0 ppm.
2
O]
28 4
C H34CoN : 242.605222; found: 242.605218, error: 0.0
2
+
28 4 2
C H38CoN O : 261.613048; found:
Preparation of 2MeCN: 2MeCN were synthesized according to the
and a HID detector. The concentrations of H
2
and CO were
MeCN
synthesis of 1
.
obtained from GC calibrated using external gas standard. The
1
concentration of formate was obtained by H-NMR using DMF as
Computational Details. All calculations were carried out with
internal standard.
[30]
the unrestricted B3LYP hybrid density functional method.
Geometry optimizations have been carried out using B3LYP/6-
Syntheses
31G*/gas phase in Gaussian 09 program package (Revision
g09w). Frequency calculations on the optimized structures
confirm the absence of imaginary frequencies. The Cartesian
coordinates of the optimized structures are available in SI.
The synthesis of 1MeCN and 2MeCN is outlined in Scheme 1. The
details are as follows.
Preparation of 5[17]: To a solution of 4 (1.05 g, 5.0 mmol) in dry
2
THF (50 mL) was added dropwise MeMgBr (3.0 M in Et O, 0.85
mL, 2.5 equiv, 12.5 mmol) at -15 °C. The reaction mixture was
further stirred for 1 h at -15 °C and then for 2 h at rt to give an
Acknowledgements
This work was financially supported by National Key Research
and Development Program of China (2016YFB0600901), and
Key Research Program of the Chinese Academy of Sciences
(ZDRW-ZS-2016-3). Thanks for the computing service of CNGrid
and ScGrid based on the SCE software.[31]
orange-red solution. Slow addition of a saturated NH
30 mL) was followed by phase separation. The aqueous phase
was extracted with THF (50 mL) and then CH Cl (50 mL). The
combined organic phases were washed with saturated NaCl
solution (50 mL) and then dried over MgSO . Evaporation of the
4
Cl solution
(
2
2
4
filtrate in vacuo left a reddish oil. Extraction with hexane (3 × 100
mL) and removal of the solvent in vacuo yielded 5 as an off-white
1
Keywords: Cobalt Complexes, Electrocatalytic Reduction of
solid (735 mg, 3.25 mmol, 65.0%). H NMR (400 MHz, CDCl
3
):δ
8.55 (d, J = 5.0 Hz, 1H), 8.44 (s, 1H), 8.33 (s, 1H), 7.88 (s, 1H),
Carbon Dioxide, Formate, Redox Active Ligand
7.17 (d, J = 4.9 Hz, 1H), 2.83 (s, 3H), 2.49 (d, J = 4.3 Hz, 6H).
Preparation of L1[29]: 5 (226 mg, 1.00 mmol) was dissolved in
ethanol (10 mL), and ten drops of acetic acid was added to the
solution, and then 2,6-diisopropylaniline (0.38 mL, 2.00 mmol)
was introduced. The solution was heated to reflux and stirred for
[1]
a) B. Kumar, M. Llorente, J. Froehlich, T. Dang, A. Sathrum,
C. P. Kubiak, Annu. Rev. Phys. Chem. 2012, 63, 541-569;
b) H. Takeda, C. Cometto, O. Ishitani, M. Robert, ACS Catal.
2016, 7, 70-88.
[
2]
a) Z. Guo, S. Cheng, C. Cometto, E. Anxolabéhère-Mallart,
S.-M. Ng, C.-C. Ko, G. Liu, L. Chen, M. Robert, T.-C. Lau,
J. Am. Chem. Soc 2016, 138, 9413-9416; b) A.
Chapovetsky, T. H. Do, R. Haiges, M. K. Takase, S. C.
Marinescu, J. Am. Chem. Soc 2016, 138, 5765-5768; c) D.
7
2 h. After cooling to room temperature, the solvent was removed
with reduced pressure. The precipitate was resolved in chloroform
and washed with saturated aqueous NaHCO , brine and dried
over Na SO After the solvent was removed by rotary
3
2
4
.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
C. Lacy, C. C. McCrory, J. C. Peters, Inorg. Chem. 2014,
3, 4980-4988; d) S. Roy, B. Sharma, J. Pecaut, P. Simon,
M. Zolnhofer, S. Mossin, F. W. Heinemann, J. Sutter, K.
Meyer, Inorg. Chem. 2014, 53, 2460-2470.
5
M. Fontecave, P. D. Tran, E. Derat, V. Artero, J. Am. Chem.
Soc 2017, 139, 3685-3696; e) M. Zhang, M. El-Roz, H. Frei,
J. Mendoza-Cortes, M. Head-Gordon, D. C. Lacy, J. C.
Peters, The Journal of Physical Chemistry C 2015, 119,
[15]
D. D. S. G. S. Wilson, and R. S. Glass, Inorg. Chem. 1985,
25, 3827-3829.
A. K. Singh, R. Mukherjee, Dalton Trans. 2008, 260-270.
a) S. Seka, O. Buriez, J. Périchon, Chem. Eur. J. 2003, 9,
3597-3603; b) S. Heider, H. Petzold, J. M. Speck, T. Rüffer,
D. Schaarschmidt, Zeitschrift für anorganische und
allgemeine Chemie 2014, 640, 1360-1367.
P. Kang, T. J. Meyer, M. Brookhart, Chem. Sci. 2013, 4,
3497.
a) E. S. Wiedner, M. B. Chambers, C. L. Pitman, R. M.
Bullock, A. J. M. Miller, A. M. Appel, Chem. Rev. 2016, 116,
8655-8692; b) Z. M. Heiden, A. P. Lathem, Organometallics
2015, 34, 1818-1827.
a) E. S. Rountree, B. D. McCarthy, T. T. Eisenhart, J. L.
Dempsey, Inorg. Chem. 2014, 53, 9983-10002; b) C.
Costentin, J.-M. Savéant, ChemElectroChem 2014, 1,
1226-1236.
a) S. Zhang, P. Kang, M. Bakir, A. M. Lapides, C. J. Dares,
T. J. Meyer, Proc. Natl. Acad. Sci. USA 2015, 112, 15809-
15814; b) Z. Chen, J. J. Concepcion, M. K. Brennaman, P.
Kang, M. R. Norris, P. G. Hoertz, T. J. Meyer, Proc. Natl.
Acad. Sci. USA 2012, 109, 15606-15611.
[16]
[17]
4645-4654; f) H. Sheng, H. Frei, J. Am. Chem. Soc 2016,
138, 9959-9967.
[
[
3]
4]
a) P. Kang, C. Cheng, Z. Chen, C. K. Schauer, T. J. Meyer,
M. Brookhart, J. Am. Chem. Soc 2012, 134, 5500-5503; b)
P. Kang, S. Zhang, T. J. Meyer, M. Brookhart, Angew. Chem.
Int. Ed. 2014, 53, 8709-8713.
a) P. L. Cheung, C. W. Machan, A. Y. Malkhasian, J.
Agarwal, C. P. Kubiak, Inorg. Chem. 2016, 55, 3192-3198;
b) C. W. Machan, J. Yin, S. A. Chabolla, M. K. Gilson, C. P.
Kubiak, J. Am. Chem. Soc 2016, 138, 8184-8193; c) M. D.
Sampson, C. P. Kubiak, J. Am. Chem. Soc 2016, 138, 1386-
[18]
[19]
[20]
[21]
1
393; d) B. A. Johnson, S. Maji, H. Agarwala, T. A. White,
E. Mijangos, S. Ott, Angew. Chem. Int. Ed. 2016, 128,
857-1861.
1
[
[
5]
6]
P. J. Chirik, Acc. Chem. Res. 2015, 48, 1687-1695.
C. S. Seu, A. M. Appel, M. D. Doud, D. L. DuBois, C. P.
Kubiak, Energy Environ. Sci. 2012, 5, 6480-6490.
[
[
7]
8]
S. D. Ittel, L. K. Johnson, M. Brookhart, Chem. Rev. 2000,
1
00, 1169-1204.
[22]
[23]
C. C. HojillaAtienza, C. Milsmann, E. Lobkovsky, P. J. Chirik,
Angew. Chem. Int. Ed. 2011, 50, 8143-8147.
a) G. J. Britovsek, S. P. Baugh, O. Hoarau, V. C. Gibson, D.
F. Wass, A. J. White, D. J. Williams, Inorg. Chim. Acta 2003,
a) S. Pramanik, S. Dutta, S. Roy, S. Dinda, T. Ghorui, A.
Kumar Mitra, K. Pramanik, S. Ganguly, New J. Chem. 2017,
41, 4157-4164; b) P. Sarkar, M. K. Mondal, A. Sarmah, S.
Maity, C. Mukherjee, Inorg. Chem. 2017, 56, 8068-8077.
a) J. Hawecker, J. M. Lehn, R. Ziessel, Helve. chim. acta
1986, 69, 1990-2012; b) F. P. Johnson, M. W. George, F.
Hartl, J. J. Turner, Organometallics 1996, 15, 3374-3387; c)
F. Franco, C. Cometto, L. Nencini, C. Barolo, F. Sordello, C.
Minero, J. Fiedler, M. Robert, R. Gobetto, C. Nervi, Chem.
Eur. J. 2017, 23, 4782-4793; d) S. C. Cheng, C. A. Blaine,
M. G. Hill, K. R. Mann, Inorg. Chem. 1996, 35, 7704-7708;
e) F. Blasco, L. Perelló, J. Latorre, J. Borrás, S. Garciá-
Granda, Inorg. Biochem. 1996, 61, 143-154.
M. Mettry, M. P. Moehlig, R. J. Hooley, Org. Lett. 2015, 17,
1497-1500.
A. P. Singh, H. W. Roesky, E. Carl, D. Stalke, J.-P. Demers,
A. Lange, J. Am. Chem. Soc 2012, 134, 4998-5003.
K. S. Hagen, Inorg. Chem. 2000, 39, 5867-5869.
V. V. Pavlishchuk, A. W. Addison, Inorg. Chim. Acta 2000,
298, 97-102.
345, 279-291; b) G. J. Britovsek, M. Bruce, V. C. Gibson, B.
S. Kimberley, P. J. Maddox, S. Mastroianni, S. J. McTavish,
C. Redshaw, G. A. Solan, S. Strömberg, J. Am. Chem. Soc
1
999, 121, 8728-8740.
H. Kazumasa, K. Kouji, N. Hiroshi, Bull. Chem. Soc. Jpn.
016, 89, 394-404.
[24]
[
[
9]
2
10]
a) S. Aroua, T. K. Todorova, P. Hommes, L.-M. Chamoreau,
H.-U. Reissig, V. Mougel, M. Fontecave, Inorg. Chem. 2017,
5
6, 5930-5940; b) S. M. Fatur, S. G. Shepard, R. F. Higgins,
M. P. Shores, N. H. Damrauer, J. Am. Chem. Soc 2017, 139,
493-4505.
D. Zhu, I. Korobkov, P. H. M. Budzelaar, Organometallics
012, 31, 3958-3971.
4
[
[
11]
12]
2
[25]
[26]
a) J. I. van der Vlugt, S. Demeshko, S. Dechert, F. Meyer,
Inorg. Chem. 2008, 47, 1576-1585; b) B. Schneider, S.
Demeshko, S. Neudeck, S. Dechert, F. Meyer, Inorg. Chem.
2013, 52, 13230-13237.
[27]
[28]
[
13]
a) Y. D. M. Champouret, J.-D. Marechal, R. K. Chaggar, J.
Fawcett, K. Singh, F. Maseras, G. A. Solan, New J. Chem.
2
007, 31, 75-85; b) V. Radkov, T. Roisnel, A. Trifonov, J.-F.
Carpentier, E. Kirillov, J. Am. Chem. Soc 2016, 138, 4350-
353; c) C. Bianchini, D. Gatteschi, G. Giambastiani, I.
[29]
J. D. Pelletier, Y. D. Champouret, J. Cadarso, L. Clowes, M.
Gañete, K. Singh, V. Thanarajasingham, G. A. Solan,
Organometallics 2006, 691, 4114-4123.
4
Guerrero Rios, A. Ienco, F. Laschi, C. Mealli, A. Meli, L.
Sorace, A. Toti, F. Vizza, Organometallics 2007, 26, 726-
[30]
[31]
C. Lee, W. Yang, R. G. Parr, Physical Review B 1988, 37,
785-789.
a) H. Xiao, H. Wu, X. Chi, Springer Berlin Heidelberg, Berlin,
Heidelberg, 2009, pp. 35-42; b) P. Vicat-Blanc Primet, T.
Kudoh, J. Mambretti, Springer Berlin Heidelberg, Berlin,
Heidelberg, 2009, pp. E1-E1.
739.
[
14]
a) W.-H. Sun, S. Jie, S. Zhang, W. Zhang, Y. Song, H. Ma,
J. Chen, K. Wedeking, R. Fröhlich, Organometallics 2006,
25, 666-677; b) M. Käß, J. Hohenberger, M. Adelhardt, E.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201800136
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
FULL PAPER
An imino bipyridine cobalt (II)
complex was developed for
Fang-Wei Liu, Jiaojiao Bi,
Yuanyuan Sun, Shuping Luo,
Peng Kang*
electrocatalytic reduction of CO
to formate in acetonitrile with
0% efficiency. CV, EPR and IR
2
8
spectroscopic studies have
provided mechanistic details
Page No. – Page No.
and
structures
of
key
Cobalt Complex with Redox
Active Imino Bipyridyl Ligand
for Electrocatalytic Reduction
of Carbon Dioxide to Formate
intermediates. DFT computation
confirmed the beneficial role of
large π-π conjugate groups.
This article is protected by copyright. All rights reserved.