Dinuclear Metal Synergistic Catalysis Boosts Photochemical CO2-to-CO Conversion

Article abstract of DOI:10.1002/anie.201811010

The solar-driven CO₂ reduction is a challenge in the field of “artificial photosynthesis”, as most catalysts display low activity and selectivity for CO₂ reduction in water-containing reaction systems as a result of competitive proton reduction. Herein, we report a dinuclear heterometallic complex, [CoZn(OH)L¹](ClO₄)₃ (CoZn), which shows extremely high photocatalytic activity and selectivity for CO₂ reduction in water/acetonitrile solution. It achieves a selectivity of 98 % for CO₂-to-CO conversion, with TON and TOF values of 65000 and 1.8 s^?1, respectively, 4, 19, and 45-fold higher than the values of corresponding dinuclear homometallic [CoCo(OH)L¹](ClO₄)₃ (CoCo), [ZnZn(OH)L¹](ClO₄)₃ (ZnZn), and mononuclear [CoL²(CH₃CN)](ClO₄)₂ (Co), respectively, under the same conditions. The increased photocatalytic performance of CoZn is due to the enhanced dinuclear metal synergistic catalysis (DMSC) effect between Zn^II and Co^II, which dramatically lowers the activation barriers of both transition states of CO₂ reduction.

Full text of DOI:10.1002/anie.201811010

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Dinuclear Metal Synergistic Catalysis Boosts Photochemical
CO2-to-CO Conversion’
Authors: Ting Ouyang, Hong-Juan Wang, Hai-Hua Huang, Jia-Wei
Wang, Song Guo, Wen-Ju Liu, Di-Chang Zhong, and Tong-Bu
Lu
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: Angew. Chem. Int. Ed. 10.1002/anie.201811010
Angew. Chem. 10.1002/ange.201811010
Link to VoR: http://dx.doi.org/10.1002/anie.201811010
http://dx.doi.org/10.1002/ange.201811010

10.1002/anie.201811010
Angewandte Chemie International Edition
DOI: 10.1002/anie.201((will be filled in by the editorial staff))
CO2 Reduction
Dinuclear Metal Synergistic Catalysis Boosts Photochemical CO2-to-
CO Conversion
Ting Ouyang,1,3,† Hong-Juan Wang,1,† Hai-Hua Huang,2 Jia-Wei Wang,2 Song Guo,1 Wen-Ju Liu,2 Di-
Chang Zhong,*,1 and Tong-Bu Lu*,1,2
Abstract: The solar-driven CO2 reduction is a perennial challenge
in the field of ‘artificial photosynthesis’, as most of the reported
catalysts display low activity and selectivity for CO2 reduction in
water-containing reaction systems due to competitive proton
reduction. Herein, we report a dinuclear heterometallic complex,
[CoZn(OH)L1](ClO4)3 (CoZn), which shows extremely high
photocatalytic activity and selectivity for CO2 reduction in
water/acetonitrile solution. It achieves a selectivity of 98% for CO2-
to-CO conversion, with TON and TOF values of 65000 and 1.8 s-1,
respectively, 4, 19 and 45-fold higher than the values of
corresponding dinuclear homometallic [CoCo(OH)L1](ClO4)3
(CoCo), [ZnZn(OH)L1](ClO4)3 (ZnZn) and mononuclear
metalloenzymes undertake diverse physiologic functions in life
system. In previous investigations, we and others have found that
some dinuclear metal complexes with close M···M distance display
synergistic effect for the recognitions and activation/catalysis of
small molecules and anions.[8-14] In addition, it has been found that
the addition of Lewis acidic metal cations to the mononuclear metal
electrocatalysts can lower the overpotentials and subsequently
increase the activities for the electrocatalytic CO2 reduction.[15-17]
The solar-driven reduction of CO2 into chemical fuels/feedstocks
represents a potential strategy for solving the issues of energy crisis
and global warming caused by increasing CO2 emission.[18-25] The
critical bottleneck of this project is to develop cheap catalysts that
[CoL2(CH3CN)](ClO4)2 (Co) respectively under the same conditions. can selectively and effectively reduce CO2 into a single chemical
The increased photocatalytic performance of CoZn is due to the
enhanced dinuclear metal synergistic catalysis (DMSC) effect
between ZnII and CoII, which dramatically lowers the activation
barriers of both transition states of CO2 reduction.
fuel/feedstock.[26-27] Although a large number of mononuclear
molecular photocatalysts for CO2-to-CO conversion have been
developed during the past several decades, most of the catalysts
display relatively low efficiency and selectivity (Table S1).
Particularly in water-containing catalytic systems, the efficiency and
selectivity are even lower due to the competitive reaction of proton
reduction (Table S1).[28-32] Recently, we reported a dinuclear
molecular catalyst of [CoCo(OH)L1](ClO4)3 (CoCo),[18] which
displays much higher photocatalytic activity (TON = 16896) and
selectivity (98%) than the corresponding mononuclear complex of
[CoL2(CH3CN)](ClO4)2 (Co, TON = 1600 with 85% selectivity) for
Dinuclear metal catalyst, where two closely associated metal sites
within one catalyst, usually shows higher catalytic activity than the
mononuclear counterpart due to the dinuclear metal synergistic
catalysis (DMSC) effect between two metal sites.[1-5] Compared
with mononuclear metal catalyst, the dinuclear one bears more
parameters to optimize its catalytic performances, not only by
changing the metal and ligand, but also varying the bimetallic
pairing.[5] In biological systems, many catalytic active centers of
metalloenzymes are dinuclear metallic complexes,[6-7] such as
heterometallic [NiFe4(OH)S4] cluster-based carbon monoxide
dehydrogenase (CODHs), dinuclear-Fe(II)-based hydrogenase,
dinuclear-Cu(II)-based tyrosinase and laccase, dinuclear-Zn(II)-
based alkaline phosphatase, dinuclear-Cu-Zn bovine superoxide
dismutase, and so on. By exploiting metal-metal cooperativity to
accomplish different substrate recognition and transformation, these
the reduction of CO2 to CO in H2O/CH3CN (L1
=
N[(CH2)2NHCH2(m-C6H4)CH2NH (CH2)2]3N, L2 = N,N',N''-tris(2-
benzylaminoethyl)amine, Figure 1), probably due to the DMSC
effect between two CoII ions, that is, one CoII serves as a catalytic
center, and the other CoII acts as an assistant catalytic site to
facilitate the cleavage of C-O bond of the O=C-OH intermediate. As
the coordination environments of both CoII in CoCo are identical, it
is unable to identify the catalytic center and the assistant catalytic
site in CoCo, thus the proposed DMSC effect within dinuclear
CoCo need to be further investigated and confirmed.
To give direct evidence for this significant DMSC, herein, we
designed and synthesized a dinuclear heterometallic cryptate,
[CoZn(OH)L1](ClO4)3 (CoZn), in which one CoII in CoCo was
replaced by a ZnII. The selection of ZnII is based on the following
considerations: (1) The ZnII has little catalytic activity for
photochemical reduction of CO2 to CO. The possible enhanced
photocatalytic activity of CoZn catalyst over Co would give direct
evidence for the existence of DMSC effect within dinuclear metal
cryptates. (2) The ZnII possesses stronger binding affinity to OH-
than CoII, which implies that the cleavage of C-OH bond in O=C-
OH intermediate by CoZn would be easier, and the DMSC effect
within CoZn would be strengthened (Figure S1). As expected,
though ZnII itself has little photocatalytic activity, the stronger
binding affinity of ZnII to OH- greatly strengthens the DMSC effect
between CoII and ZnII, thus significantly enhances the activity of
CoZn for photocatalytic reduction of CO2 to CO. The TON and
TOF values reach 65000 and 1.80 s−1, respectively, 4, 19 and 45-
[]
Dr. T. Ouyang+, Dr. H. J. Wang+, S. Guo, Prof. D. C. Zhong, Prof. T.
B. Lu
Institute for New Energy Materials and Low Carbon Technologies,
School of Materials Science and Engineering, Tianjin University of
Technology, Tianjin 300384, China.
E-mail: zhong_dichang@hotmail.com; lutongbu@tjut.edu.cn
Dr. H. H. Huang, J. W. Wang, W. J. Liu, Prof. T. B. Lu
MOE Key Laboratory of Bioinorganic and Synthetic Chemistry,
School of Chemistry, Sun Yat-Sen University, Guangzhou 510275,
China
Dr. T. Ouyang
School of Chemistry and Chemical Engineering, Guangzhou
University, Guangzhou 510006, China.
†
These authors contributed equally to this work.
Supporting information (SI) for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201xxxxxx.
1
This article is protected by copyright. All rights reserved.

10.1002/anie.201811010
Angewandte Chemie International Edition
fold higher than the corresponding TON and TOF values achieved
by CoCo, ZnZn ([ZnZn(OH)L1](ClO4)3) and Co, respectively.
Further investigation revealed that simply adding ZnII as a Lewis
acid to the reaction system containing mononuclear Co or dinuclear
CoCo catalyst cannot boost the photocatalytic activity for CO2-to-
CO conversion. The strengthened DMSC effect and subsequently
enhanced catalytic activity of CoZn for photochemical reduction of
CO2 to CO were supported by DFT calculations.
Figure 1. Structures of designed molecular catalysts, (a) dinuclear CoZn; (b)
dinuclear CoCo; (c) mononuclear Co.
The CoZn cryptate was synthesized by the reaction of L1 with
ZnII and CoII step by step (see the supporting information, SI).
Electrospray ionization mass spectrometry (ESI-MS) clearly shows
that the mononuclear ZnII complex of [ZnL1]2+ was generated at the
first step (Figure S2a). After further reaction of [ZnL1]2+ with CoII, a
carbonate-bridged complex of [CoZn(μ-CO3)L1](ClO4)2 (CoZn-a)
was formed, which was demonstrated by the result of ESI-MS
measurement. As shown in Figure S2c, the ion peak at m/z = 390.65
corresponds to the species of [CoZn(CO3)L1]2+. The absence of the
ion peaks corresponding to [CoCo(CO3)L1]2+ and [ZnZn(CO3)L1]2+
rules out the possibility for the formation of a mixture of
[CoCo(CO3)L1]2+ and [ZnZn(CO3)L1]2+, demonstrating that the
dinuclear heterometallic CoZn-a has been successfully obtained.
Similar to CoCo,[18] the suitable Co···Zn distance of 5.56 Å in
CoZn cryptate (Figure S3), obtained from the optimized structure of
CoZn-a (see the results of DFT calculations), is favorable for the
formation of a μ-CO32- bridged cryptate of CoZn-a within its cavity.
However, compared to CoCo, CoZn is much easier to absorb
Figure 2. The results of photocatalytic reduction of CO2 to CO by CoZn, CoCo,
ZnZn, Co, and Zn under the same conditions. a) Time-dependent
photocatalytic evolution of CO (blue) and H2 (black) catalyzed by CoZn (★,
0.025 μM), CoCo (●, 0.025 μM), ZnZn (▲, 0.025 μM), Co (▼, 0.025 μM) and
Zn (◆, 0.025 μM), respectively, under the irradiation of a LED light (450 nm,
100 mW·cm-2, irradiation area, 0.8 cm2) in the presence of [Ru(phen)3](PF6)2
(0.4 mM) and TEOA (0.3 M) in 5 mL CO2-saturated H2O/CH3CN (v/v = 1:4)
solution at 25 ºC. To confirm the reliability of the data, each photocatalytic
reaction was repeated at least three times. b) A comparison of photocatalytic
activity of CoZn with CoCo, ZnZn, Co, and Zn.
Table 1. The results of photocatalytic reduction of CO2 to CO by different
catalysts under the same conditions.[a]
Conc
(μM)
CO
H2
TOF for
CO (s-1)
Entry Cat.
CO% TON for CO
(μmol) (μmol)
2-
atmospheric CO2 at room temperature to form μ-CO3 bridged
1
2
CoZn 0.025 8.10
CoCo 0.025 2.11
0.16
98%
98%
65000
17000
1.8
cryptate, thus CoZn can only be obtained in a glove box under
nitrogen atmosphere with CO2-free solvents. Trace amount of
atmospheric CO2 will lead to the formation of CoZn-a (Figure S2e).
The rapid absorption and fixation of atmospheric CO2 suggests that
CoZn bears stronger affinity towards CO2 substrate than CoCo. The
results of XPS and ICP-AES measurements reveal that the valence
of Co and Zn is +2 (Figure S4), and the Co/Zn molar ratio in the
synthesized CoZn-a is 1:1 (Table S2). As CoZn-a is the only
component originated from CoZn by rapid absorption of
atmospheric CO2, and is stable under the general experimental
conditions, thus we used substrate-bonded CoZn-a instead of CoZn
as the starting material for the subsequent investigations. UV-vis
spectra and the cyclic voltamemetry (CV) for CoZn, CoCo, ZnZn,
Co, and Zn are shown in Figures S5 and S6, respectively. As shown
in Figure S6, the redox potential of CoIIZnII/CoIZnII in CoZn (-0.87
0.046
0.47
ZnZn
3
0.025 0.42
0.025 0.18
0.17
71%
3400
0.093
Co
Zn
4
5
0.063
0.050
74%
0
1500
0
0.040
0
0.025
0
[a] Reaction conditions: [Ru(phen)3](PF6)2 (0.4 mM), TEOA (0.3 M), H2O/CH3CN
(v/v = 1:4), LED light (450 nm, 100 mW·cm-2, irradiation area 0.8 cm2), 10 h,
25 °C. TON and TOF values for CO were calculated based on per molecular
formula and averaged over three reactions, with deviations below 5% for Entry
1, and 2, and about 10 % for Entry 3, 4 and 5.
for CoCo and Co.[18] Typically, a 5 mL CO2-saturated H2O/CH3CN
solution (v/v = 1:4) containing certain amount of CoZn-a catalyst,
[Ru(phen)3](PF6)2 photosensitizer (PS) and triethanolamine (TEOA)
sacrificial reductant was irradiated by a LED light (450 nm, 100
mW·cm-2) at 25 oC. The generated gases were analyzed by gas
chromatography (GC). The results show that one single photoredox
cycle produces 8.10 μmol of CO and 0.16 μmol of H2 within 10 h
(Table 1, Entry 1; Figure 2a), along with trace amount of formate
detected in the liquid phase by ion chromatograph (IC). The
I,II
V) is more positive than that of Co2II,II/Co2 in CoCo (-0.94 V)
under CO2 atmosphere, indicating CoII in CoZn is easier to be
reduced than that in CoCo.
The photocatalytic experiments of CO2 reduction by CoZn were
performed under the same conditions as those previously described
2
This article is protected by copyright. All rights reserved.

10.1002/anie.201811010
Angewandte Chemie International Edition
performed to confirm the generated CO coming from the reduction
of CO2. The result clearly illustrates that under a 13CO2 atmosphere,
the generated gas in the photocatalytic system was 13CO (Figure
S9), demonstrating the generated CO originates from the reduction
of CO2 by CoZn, rather than the decomposition of the
catalyst/photosensitizer/sacrificial reductant. Thirdly, a mercury
poison test was carried out to make certain that the catalytic reaction
is homogeneous. The results show that the amount of CO generated
from the system in the presence of mercury is almost equal to that
without mercury (Table 2, Entry 7), illustrating that the production
of CO is catalyzed by CoZn rather than metal Co, and the above
photocatalytic reaction is homogeneous. Fourthly, the kinetics of
this homogeneous photocatalytic reaction was studied. The results
illustrate that the amounts of CO evolution are roughly linearly
dependent on the concentrations of CoZn within 0.00625~0.025 μM
range (Figure S10), suggesting that the photochemical reduction of
CO2 to CO is a first-order reaction and only Co(II) in CoZn acts as
the catalytic center.
amounts of gases generated correspond to the selectivity to CO of
98%, and the TON and TOF values for CO of 65000 and 1.80 s-1,
respectively (Table 1, Entry 1). This TON value is 4-, 19- and 45-
fold higher than those of CoCo (17000), ZnZn (3400), and Co
(1500) under the same conditions (Table 1, Entry 2-4; Figure 2b).
The quantum yield of CoZn to CO was determined to be 0.15 %
(see SI), 4-fold higher than that of CoCo (0.04%).[18] Moreover,
when the solutions containing 0.05 and 0.1 mM of CoZn were
irradiated with LED light for 5 min, much high quantum yields of
4.0% and 4.6% were obtained (see SI). The above results illustrate
that after the replacement of one CoII in CoCo with a ZnII, the
resulted CoZn exhibits significantly enhanced photocatalytic
activity for CO2-to-CO conversion. As the mononuclear zinc
complex of [ZnL2(CH3CN)](ClO4)2 (Zn) shows no catalytic activity
under the same conditions (Table 1, Entry 5), the extremely higher
photocatalytic activity of CoZn over CoCo can evidence directly
that there is a strong synergistic catalysis effect between CoII and
ZnII within CoZn. This result, together with the much higher TON
values of CoCo over Co, and ZnZn over Zn (Table 1, Entry 2-5),
solidly illustrate that there is a DMSC effect between two metals
within dinuclear metal cryptates, that is, CoII acts as a catalytic
centre, and ZnII serves as an assistant catalytic site to facilitate the
cleavage of C-O bond of the O=C-OH intermediate.
Table 2. The Results of Control Experiments for the Photocatalytic Reduction of
CO2 to CO.[a]
Conc
CO
H2
TON
TOF for
CO (s-1)
Entry
Cat.
CO%
(μM)
(μmol)
(μmol)
for CO
To see if the photocatalytic activity for CO2 reduction can be
boosted simply by adding Zn2+ as a Lewis acid to the reaction
system containing dinuclear CoCo or mononuclear Co catalyst, as
observed in the electrocatalytic CO2 reduction by mononuclear
metal catalysts,[15-17] the experiments of photocatalytic CO2
reduction by CoCo and Co in the presence of ZnII were performed.
As shown in Table S3, the addition of Zn2+/Mg2+/Li+ as a Lewis acid
to the reaction systems containing CoCo or Co catalyst cannot boost
and even slightly decrease the CO2 reduction catalytic activity of
CoCo and Co. The above results clearly demonstrate that the
externally added Lewis acid to the CoCo/Co photocatalytic system
cannot generate the DMSC effect as CoZn behaves, indicating the
cryptand ligand (L1) plays a crucial role for the construction of the
highly active CoZn catalyst with DMSC effect, in which L1 can
hold CoII and ZnII in a close proximity (5.56 Å) for strong affinity
towards CO2 substrate, while the externally added metal ions has
little opportunity to form -CO3 bridged dinuclear complexes with
mononuclear Co and Zn catalysts. Therefore, DMSC effect for
photocatalytic CO2-to-CO conversion can only be produced within a
dinuclear metal complexes with suitable M···M distance fixed by an
organic ligand for strong affinity towards CO2 substrate.
A series of control experiments were carried out to thoroughly
investigate this photocatalytic reaction of CO2-to-CO conversion
(Table 2). Firstly, the photocatalytic reactions was performed
without CoZn, [Ru(phen)3](PF6)2, TEOA, or visible-light. The
results show that without any of these reaction factors, no CO was
detected (Table 2, Entry 2-5), suggesting that catalyst,
photosensitizer, sacrificial reductant, and light are all indispensible
to CO2-to-CO conversion. It is worthy to note that water is also
critical for CO2-to-CO conversion. No CO was detected in the
absence of water (Table 2, Entry 6). Slightly increasing the ratio of
H2O/CH3CN to 1:9, significant amount of CO (5.5 μmol) was
generated, demonstrating that water acts as an effective proton
source in the reaction process. When the ratio of H2O/CH3CN
reached to 1:4, a maximum amount of CO (8.1 μmol) was obtained.
Further increase the ratio of H2O/CH3CN to 1:1, the generated CO
was dramatically decreased, and no CO could be detected in pure
water (Table S5, Figure S8), as catalyst and PS can only slightly
soluble in water. Secondly, the 13CO2 isotope trace experiment was
1
2
3
4
5
CoZn
0
0.025
0
8.10
0.16
0.049
0
98%
65000
1.8
0
0
0
0
0
0
0
0
0
0
0
0
0
CoZn
CoZn
CoZn
0.025
0.025
0.025
0
0.11
0
0
0
6
7
CoZn
CoZn
0.025
0.025
0
0.10
0.16
0
0
0
7.98
98%
64000
1.7
[a] Reaction conditions: [Ru(phen)3](PF6)2 (0.4 mM), TEOA (0.3 M), H2O/CH3CN
(v/v = 1:4), LED light (450 nm, 100 mW·cm-2, irradiation area 0.8 cm2), 10 h, 25
C. TON and TOF values are averaged over three reactions, with deviations
below 5%. 2: without CoZn; 3: without [Ru(phen)3](PF6)2; 4: without TEOA; 5:
without visible light; 6: without H2O; 7: adding 0.5 mL of Hg.
The stability of CoZn catalyst during the photocatalytic CO2
reduction reaction was investigated by the durability test and
dynamic light scattering (DLS) measurement. The durability test for
CoZn shows that the addition of [Ru(phen)3](PF6)2 can reactivate
the reaction of photochemical CO2-to-CO conversion reaction that is
nearly ceased (Figure S11), indicating that the cease of CO2
reduction reaction is attributed to the photo-degradation of
[Ru(phen)3](PF6)2 rather than the inactivation of CoZn catalyst. The
dynamic light scattering (DLS) measurement indicates that no
nanoparticle generated in the solution after the CO2 reduction
reaction, revealing that CoZn does not decompose during the
process of photocatalytic CO2 reduction. Therefore, both the
durability test and DLS measurement have demonstrated that CoZn
possesses a good stability and can serve as a stable homogeneous
catalyst for photochemical CO2 reduction to CO.
The fluorescence quenching studies were carried out on the
reaction system containing [Ru(phen)3]2+ photosensitizer (PS) with
various concentrations of CoZn catalyst or TEOA sacrificial
reductant in a degassed solution of H2O/CH3CN (v/v = 1:4). As
shown in Figure S12, the fluorescent intensity of [Ru(phen)3]2+ was
decreased even with the addition of small amount of CoZn catalyst
(Figures S12a and b), while no obvious change was observed with
the addition of even much excessive TEOA (0.3 M) (Figures S12e
and 12f). Moreover, transient absorption spectra of
[Ru(phen)3](PF6)2 in the presence of TEOA and CoZn-a were
3
This article is protected by copyright. All rights reserved.

10.1002/anie.201811010
Angewandte Chemie International Edition
measured (Figure S13), which indicate new species, most probably
[Ru(phen)3]3+ was formed in the presence of CoZn-a, while no
obvious change with the addition of TEOA. The above results
demonstrate the [Ru(phen)3]2+* is quenched through the following
oxidative pathway:[33]
CoIIZnII-a + Ru(phen)3]2+* → CoIZnII-b + [Ru(phen)3]3+
[Ru(phen)3]3+ + TEOA → [Ru(phen)3]2+ + TEOA+
Stern-Volmer analysis from the data in Figure S12c revealed that the
rate constant for electron transfer from [Ru(phen)3]2+* to CoZn-a is
kq = 3.0 × 1010 M-1 s-1, indicating a diffusion-controlled electron
transfer process. Indeed, the redox potential of [Ru(phen)3]3+/2+* (-
0.84 V, see SI) is negative than the Co2+/+Zn onset reduction
potential of CoZn (-0.58 V), and close to its redox potential of -0.87
V (Figure S6a), indicating the electron transfer from [Ru(phen)3]2+*
to CoIIZnII is thermodynamically accessible.
To well understand the dramatically enhanced photocatalytic
activity of CoZn towards CO2-to-CO conversion compared to
CoCo, the catalytic mechanism was investigated at a molecular
level by the DFT calculations (see SI), and compared with the
energy diagram for the reaction pathway of CoCo (Figure S15). As
shown in Figure 3, firstly, CoZn rapidly captures the CO2 substrate
to form a carbonate-bridged complex of CoZn-a, as observed from
the experimental results. Secondly, CoZn-a obtains one electron
from [Ru(phen)3]2+* to generate CoZn-b, during which the CoIIZnII
in CoZn-a was reduced to CoIZnII. Thirdly, CoZn-b is protonated to
yield CoZn-c after elimination of a water molecule, and then the
CO2 in CoZn-c accepts 2e- from CoI to be reduced to CO22- through
the transition state of TS1-CoZn to generate CoZn-d. The total
energy barrier for this transition state is only 6.55 kcal/mol, much
lower than the corresponding values of CoCo (10.41 kcal/mol) and
Co (24.96 kcal/mol) (Figures 4 and S15). Fourthly, CoIII in CoZn-d
is reduced to CoII by [Ru(phen)3]2+* through a PCET oxidative
process to form CoZn-e. Finally, the C-OH bond of COOH- within
CoZn-e is cleaved by the DMSC interaction via the second
transition state TS2-CoZn to generate CoZn-f, where ZnII binds to
OH- and CoII binds to CO. The total energy barrier calculated for
TS2-CoZn is as low as 13.99 kcal/mol, also much lower than the
corresponding values of CoCo (16.35 kcal/mol) and Co (20.73
kcal/mol) (Figures 4 and S15). After the release of CO in CoZn-f,
CoZn is regenerated and the photocatalytic cycle restarts. From
Figure 3 it can be found that the reactions of CO2 reduction contain
several proton-participated reaction process, and the energy barriers
become much higher without the participation of protons, this result
is consistent with the experimental observation that water is critical
for CO2-to-CO conversion (Table 2, Entry 6).
Figure 3. Proposed catalytic mechanism of CoZn for the reduction of CO2 to
CO driven by visible-light.
Figure 4. The energy barriers of two transition states for CoZn, CoCo, and Co.
The above results of DFT calculations strongly support the
experimental observations that the TON value of CoZn is much
higher than those of CoCo and Co under the same conditions (Table
1). As the selected natural bond orbital (NBO) charge of Zn in
CoZn (1.280e) is more positive than that of Co in CoCo (1.115e,
Table S6), the significantly improved photocatalytic performance of
CoZn over CoCo can be ascribed to the stronger binding affinity of
ZnII to OH- over CoII, which enormously strengthens the DMSC
effect between CoII and ZnII in CoZn, thus greatly promotes the
cleavage of the C-OH bond in O=C-OH intermediate, and
significantly accelerates the photochemical reduction reaction of
CO2 to CO.
light irradiation with a low light intensity of 100 mW·cm-2. The
TON and TOF values reach fantastic 65000 and 1.8 s−1, respectively,
and the selectivity to CO reaches as high as 98%. The greatly
enhanced catalytic activity of CoZn can be ascribed to significantly
strengthened DMSC effect between CoII and ZnII, where the
assistant catalytic site of ZnII shows much strong binding affinity to
OH-, greatly promoting the C-OH cleavage of O=C-OH
intermediate, thus significantly enhancing the photocatalytic activity
of CoZn for CO2 reduction to CO. This research has not only
directly evidenced the DMSC effect within dinuclear metal cryptates,
but also developed a homogeneous catalyst with unprecedented high
activity and selectivity for photocatalytic reduction of CO2 to CO in
a water-containing reaction system. The results presented here
provide a new strategy for the design and synthesis of high efficient,
high selective, stable and cheap photocatalysts for CO2-to-CO
conversion, and the concept of evidenced DMSC will greatly boost
In summary, by replacing one CoII in CoCo with a ZnII, we have
successfully obtained a dinuclear heterometallic CoZn catalyst,
which shows greatly enhanced activity for photochemical reduction
of CO2 to CO in a water-containing reaction system under a visible
4
This article is protected by copyright. All rights reserved.

10.1002/anie.201811010
Angewandte Chemie International Edition
[25] F. Studt, I. Sharafutdinov, F. Abild-Pedersen, C. F. Elkjaer, J.
S. Hummelshoj, S. Dahl, I. Chorkendorff, J. K. Norskov, Nat.
Chem. 2014, 6, 320-324.
the developments of ‘artificial photosynthesis’ and other dinuclear
metal catalysts.
[26] V. S. Thoi, N. Kornienko, C. G. Margarit, P. Yang, C. J.
Chang, J. Am. Chem. Soc. 2013, 135, 14413-14424.
[27] L. Chen, Z. Guo, X.-G. Wei, C. Gallenkamp, J. Bonin, E.
Anxolabéhère-Mallart, K.-C. Lau, T.-C. Lau, M. Robert, J.
Am. Chem. Soc. 2015, 137, 10918-10921.
[28] E. Fujita, Coord. Chem. Rev. 1999, 185-186, 373-384.
[29] H. Takeda, K. Ohashi, A. Sekine, O. Ishitani, J. Am. Chem.
Soc. 2016, 138, 4354-4357.
Acknowledgements
This work was supported by National Key R&D Program of China
(2017YFA0700104), the NSFC (21790052 and 21331007), and the
111 Project (D17003).
[30] J. Bonin, M. Robert, M. Routier, J. Am. Chem. Soc. 2014, 136,
16768-16771.
[31] S. Sato, T. Morikawa, T. Kajino, O. Ishitani, Angew. Chem.
Received: ((will be filled in by the editorial staff))
Published online on ((will be filled in by the editorial staff))
Int. Ed. 2013, 52, 988-992.
[32] Z. Guo, F. Yu, Y. Yang, C. F. Leung, S. M. Ng, C. C. Ko, C.
Cometto, T. C. Lau, M. Robert, ChemSusChem 2017, 10,
4009-4013.
Keywords: CO2 reduction • dinuclear metal complex • synergistic
catalysis • homogeneous catalyst • DFT calculations
[33] G. J. Kavarnos, N. J. Turro, Chem. Rev. 1986, 86, 401-449.
[1] Y. Hayashi, S. Santoro, Y. Azuma, F. Himo, T. Ohshima, K.
Mashima, J. Am. Chem. Soc. 2013, 135, 6192-6199.
[2] A. Magnuson, M. Anderlund, O. Johansson, P. Lindblad, R.
Lomoth, T. Polivka, S. Ott, K. Stensjo, S. Styring, V.
Sundstrom, L. Hammarstrom, Acc. Chem. Res. 2009, 42,
1899-1909.
[3] O. Iranzo, A. Y. Kovalevsky, J. R. Morrow, J. P. Richard, J.
Am. Chem. Soc. 2003, 125, 1988-1993.
[4] L. Zong, C. Wang, A. M. Moeljadi, X. Ye, R. Ganguly, Y. Li,
H. Hirao, C. H. Tan, Nat. Commun. 2016, 7, 13455.
[5] N. P. Mankad, Chem. Eur. J. 2016, 22, 5822-5829.
[6] J. B. Vincent, G. L. Olivier-Lilley, B. A. Averill, Chem. Rev.
1990, 90, 1447-1467.
[7] L. Que, Jr., W. B. Tolman, Nature 2008, 455, 333-340.
[8] D. C. Zhong, T. B. Lu, Chem. Commun. 2016, 52, 10322-
10337;
[9] J. M. Stauber, E. D. Bloch, K. D. Vogiatzis, S. L. Zheng, R. G.
Hadt, D. Hayes, L. X. Chen, L. Gagliardi, D. G. Nocera, C. C.
Cummins, J. Am. Chem. Soc. 2015, 137, 15354-15357.
[10] D. Huang, R. H. Holm, J. Am. Chem. Soc. 2010, 132, 4693-
4701.
[11] T. B. Lu, X. M. Zhuang, Y. W. Li, S. Chen, J. Am. Chem. Soc.,
2004, 126, 4760-4761.
[12] J. M. Chen, W. Wei, X. L. Feng, T. B. Lu, Chem. Asian J.,
2007, 2, 710-719.
[13] R. Angamuthu, P. Byers, M. Lutz, A. L. Spek, E. Bouwman,
Science 2010, 327, 313-315.
[14] L. M. Cao, H. H. Huang, J. W. Wang, D. C. Zhong, T. B. Lu,
Green Chem. 2018, 20, 798-803.
[15] I. Azcarate, C. Costentin, M. Robert, J. M. Savéant, J. Am.
Chem. Soc. 2016, 138, 16639−16644.
[16] M. D. Sampson, C. P. Kubiak, J. Am. Chem. Soc. 2016, 138,
1386−1393.
[17] A. Zhanaidarova, H. Steger, M. H. Reineke, C. P. Kubiak,
Dalton Trans. 2017, 46, 12413-12416.
[18] (a) T. Ouyang, H. H. Huang, J. W. Wang, D. C. Zhong, T. B.
Lu, Angew. Chem. Int. Ed. 2017, 56, 738-743; (b) Z. A. Lan,
X. C. Wang, Acta Phys. -Chim. Sin.. 2017, 33, 457-457.
[19] H. Rao, J. Bonin, M. Robert, Chem. Commun. 2017, 53,
2830-2833.
[20] A. Rosas-Hernandez, C. Steinlechner, H. Junge, M. Beller,
Green Chem. 2017, 19, 2356-2360.
[21] R. Schlogl, Angew. Chem. Int. Ed. 2011, 50, 6424-6426.
[22] E. E. Benson, C. P. Kubiak, A. J. Sathrum, J. M. Smieja,
Chem. Soc. Rev. 2009, 38, 89-99.
[23] C. D. Windle, R. N. Perutz, Coord. Chem. Rev. 2012, 256,
2562-2570.
[24] S. Wang, X. Wang, Angew. Chem. Int. Ed. 2016, 55, 2308-
2320.
5
This article is protected by copyright. All rights reserved.

10.1002/anie.201811010
Angewandte Chemie International Edition
CO2 Reduction
A dinuclear heterometallic CoZn catalyst
shows much higher photocatalytic activity
Ting Ouyang, Hong-Juan Wang, Hai-
Hua Huang, Jia-Wei Wang, Song Guo,
Wen-Ju Liu, Di-Chang Zhong*, and
Tong-Bu Lu*
than
the
corresponding
dinuclear
homometallic CoCo and ZnZn catalysts, as
well as mononuclear Co and Zn catalysts
for CO2 reduction under the same
conditions. The increased photocatalytic
performance of CoZn catalyst is due to the
enhanced dinuclear metal synergistic
catalysis (DMSC) effect between ZnII and
CoII, which dramatically lowers the
activation barriers of both transition states
of CO2 reduction.
__________ Page – Page
Dinuclear Metal Synergistic Catalysis
Boosts
Photochemical
CO2-to-CO
Conversion
6
This article is protected by copyright. All rights reserved.