Lattice-Hydride Mechanism in Electrocatalytic CO2 Reduction by Structurally Precise Copper-Hydride Nanoclusters

Article abstract of DOI:10.1021/jacs.7b05591

Copper electrocatalysts can reduce CO₂ to hydrocarbons at high overpotentials. However, a mechanistic understanding of CO₂ reduction on nanostructured Cu catalysts has been lacking. Herein we show that the structurally precise ligand-protected Cu-hydride nanoclusters, such as Cu₃₂H₂₀L₁₂ (L is a dithiophosphate ligand), offer unique selectivity for electrocatalytic CO₂ reduction at low overpotentials. Our density functional theory (DFT) calculations predict that the presence of the negatively charged hydrides in the copper cluster plays a critical role in determining the selectivity of the reduction product, yielding HCOOH over CO with a lower overpotential. The HCOOH formation proceeds via the lattice-hydride mechanism: first, surface hydrides reduce CO₂ to HCOOH product, and then the hydride vacancies are readily regenerated by the electrochemical proton reduction. DFT calculations further predict that hydrogen evolution is less competitive than HCOOH formation at the low overpotential. Confirming the predictions, electrochemical tests of CO₂ reduction on the Cu₃₂H₂₀L₁₂ cluster demonstrate that HCOOH is indeed the main product at low overpotential, while H₂ production dominates at higher overpotential. The unique selectivity afforded by the lattice-hydride mechanism opens the door for further fundamental and applied studies of electrocatalytic CO₂ reduction by copper-hydride nanoclusters and other metal nanoclusters that contain hydrides.

Full text of DOI:10.1021/jacs.7b05591

Subscriber access provided by EAST TENNESSEE STATE UNIV
Article
Lattice Hydride Mechanism in Electrocatalytic CO2 Reduction
by Structurally Precise Copper-Hydride Nanoclusters
Qing Tang, Yongjin Lee, Dai-Ying Li, Woojun Choi, C.W. Liu, Dongil Lee, and De-en Jiang
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 22 Jun 2017
Downloaded from http://pubs.acs.org on June 22, 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 23
Journal of the American Chemical Society
1
2
3
4
5
6
Lattice Hydride Mechanism in Electrocatalytic CO2 Reduction by Structurally
Precise Copper-Hydride Nanoclusters
7
8
9
Qing Tang,† Yongjin Lee,‡ DaiꢀYing Li,§ Woojun Choi,‡ C. W. Liu,*,§
Dongil Lee,*,‡ and Deꢀen Jiang*,†
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
†Department of Chemistry, University of California, Riverside, CA 92521, USA
‡Department of Chemistry, Yonsei University, Seoul 03722, South Korea
§Department of Chemistry, National Dong Hwa University, Hualien 97401, Taiwan, R.O.C.
*To whom correspondence should be addressed. Eꢀmail: chenwei@mail.ndhu.edu.tw,
dongil@yonsei.ac.kr, djiang@ucr.edu.
ABSTRACT: Copper electrocatalysts can reduce CO2 to hydrocarbons at high
overpotentials. However, mechanistic understanding of CO2 reduction on nanostructured
Cu catalysts has been lacking. Herein we show that the structurally precise ligandꢀ
protected Cuꢀhydride nanoclusters, such as Cu32H20L12 (L is a dithiophosphate ligand),
offer unique selectivity for electrocatalytic CO2 reduction at low overpotentials. Our
density functional theory (DFT) calculations predict that the presence of the negatively
charged hydrides in the copper cluster plays a critical role in determining the selectivity
of the reduction product, yielding HCOOH over CO with a lower overpotential. The
HCOOH formation proceeds via the lattice hydride mechanism: first, surface hydrides
reduce CO2 to HCOOH product, and then the hydride vacancies are readily regenerated
by the electrochemical proton reduction. DFT calculations further predict that hydrogen
evolution is less competitive that HCOOH formation at the low overpotential.
Confirming the predictions, electrochemical tests of CO2 reduction on the Cu32H20L12
cluster demonstrate that HCOOH is indeed the main product at low overpotential while
H2 production dominates at higher overpotential. The unique selectivity afforded by the
latticeꢀhydride mechanism opens the door for further fundamental and applied studies of
electrocatalytic CO2 reduction by copperꢀhydride nanoclusters and other metal
nanoclusters that contain hydrides.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 23
1
2
3
4
1. INTRODUCTION
5
6
7
8
The increasing emission of CO2 into atmosphere due to the burning of fossil fuels
and other human activities has been shown to be the main contributor to global warming
and climate change.1,2 A promising strategy to alleviate this concern is to convert CO2
into useful fuels and chemicals.3,4 CO2 itself is, however, very stable and relatively inert,
making its chemical conversion challenging due to the need of large energy input and
efficient catalysts. Among the different methods developed for CO2 reduction,
electrochemical reduction has attracted great attention since it is undertaken at ambient
conditions and can be powered by renewable energy.5,6
Numerous research efforts have focused on transitionꢀmetal electrocatalysts;7 in
particular, copper has stood out due to its unique ability to produce significant amounts of
hydrocarbons from CO2 electroreduction. The product distribution is found to be
potentialꢀdependent on Cu:8ꢀ13 at a less negative potential or small overpotential,
formation of H2, HCOOH and CO is predominant, while at a sufficiently higher
overpotential, formation of CH4 and C2H4 becomes major. A main drawback for CO2
electroreduction on the Cu surface is that the hydrocarbon formation takes place only at a
high overpotential of about 1 eV.
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
Understanding how CO2 is reduced on the Cu surface can shed light on the design of
new catalysts that decrease the onset overpotential and increase the product selectivity.
As such, many recent experimental14ꢀ25 and theoretical26ꢀ32 studies have examined the
catalytic mechanism of CO2 reduction on the metallic copper surface. It is generally
accepted that CO is the key intermediate to formation of CH4 and C2H4. Although the
highꢀquality singleꢀcrystal Cu surfaces can serve as useful benchmark models to elucidate
how Cu catalyzes the electroreduction of CO2, these model systems usually have low
specific surface areas. Instead, Cu nanoparticles have been recently explored to tune the
electrocatalytic activity; the morphology and particle size of the copper
crystals/nanoparticles have been shown to have a dramatic influence on the catalytic
activity and product selectivity of CO2 electroreduction.33ꢀ35
Although nanostructured Cu catalysts have been widely investigated for the
electroreduction of CO2, their surface structures are not well defined to serve as a model
to reveal the mechanism of CO2 reduction. Different from the bulk metal surfaces and
2
ACS Paragon Plus Environment

Page 3 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
larger nanoparticles, ligandꢀprotected atomically precise metal clusters exhibit not only
wellꢀdefined atomic structures, but also sizeꢀdependent electronic, chemical and catalytic
properties. Many atomically precise Au36,37 and Ag38ꢀ41 clusters have been synthesized
recently, with thiolate,42ꢀ44 phosphine,45 or alkynyl ligands.46ꢀ49 More excitingly, several
ligandꢀprotected copperꢀhydride nanoclusters,50 such as [Cu14H12(phen)6(PPh3)4][X]2,51
[Cu18H7{1,2ꢀS(C6H4)PPh2}10(I)],52 [Cu20H11{S2P(OiPr)2}9],53,54 [Cu25H22(PPh3)12]Cl,55
[Cu28H15(S2CNR)12]PF6,56 and Cu32H20{S2P(OiPr)2}12,57 have been crystallized. A
particularly interesting feature in these copper clusters is the presence of hydrides (Hꢀ)
which can be found at the capping (or surface) sites and/or the interstitial sites. These
atomically and structurally precise Cu clusters with their wellꢀdefined structures provide
a unique opportunity to understand the mechanism of electrocatalytic CO2 reduction and
to reveal the role of the hydride ligands that could lead to new mechanistic insights and
product selectivity. To the best of our knowledge, these structurally precise CuꢀH clusters
have not been tested for electrochemical CO2 reduction.
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
Among the experimentally realized, structurally precise CuꢀH clusters, the neutral
Cu32H20L12 cluster (L= S2P(OiPr)2) offers a diverse set of hydrides of different
coordination environments that could offer synergy in CO2 reduction. In this work, we
use the Cu32H20L12 cluster as the prototypical CuꢀH nanosystem to investigate its activity
and selectivity for electrochemical reduction of CO2: first, we reveal the role of hydrides
and predict the product selectivity by density functional theory (DFT) calculations; then.
we test the electrocatalytic activity of the synthesized Cu32H20L12 cluster in an aqueous
solution.
2. METHODS
Computational. Spinꢀpolarized DFT calculations were performed to investigate the
electrocatalytic activity of Cu32H20L12 nanocluster for CO2 reduction by using the Vienna
ab initio simulation package (VASP).58 To save the computational cost, we simplified L
from S2P(OiPr)2 to S2PH2 as done by others.54,57 The ionꢀelectron interaction is described
with the projector augmented wave (PAW) method.59 Electron exchangeꢀcorrelation is
represented by the functional of Perdew, Burke and Ernzerhof (PBE) of generalized
gradient approximation (GGA).60 A cutoff energy of 400 eV was used for the planeꢀwave
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 23
1
2
3
4
5
6
7
8
9
basis set. The van der Waals interactions were considered and described using the
empirical correction in Grimme’s scheme (DFTꢀD3).61 The Cu32H20L12 nanocluster was
placed in a cubic box with dimension of 25×25×25 Å3. The rodꢀlike cluster is about 1.2
nm in diameter (in the xy plane) and 1.35 nm in length (along the z direction), and the
vacuum space distance along the x, y and z directions is 12 Å, 11 Å, and 14 Å,
respectively. Only the Γꢀpoint was used to sample the Brillouin zone. The climbingꢀ
image nudged elastic band (CIꢀNEB) method62 implemented in VASP was used to
determine the minimum energy paths for the nonꢀelectrochemical elementary step. The
convergence threshold for structural optimization was set to be 10ꢀ4 eV in energy and
0.03 eV/Å in force. The transition states were obtained by relaxing the force below 0.05
eV/Å.
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 reaction free energies of the electrochemical steps with the involvement of
protonꢀelectron pair were calculated using the computational hydrogen electrode
proposed by Nørskov et al.63 The change in Gibbs free energy (ꢁG) for all the
electrochemical or nonꢀelectrochemical steps was defined as ꢁG = ꢁE + ꢁEzpe − TꢁS,
where ꢁE is the different in the DFT total energy, ꢁEzpe is the zeroꢀpoint energy
difference calculated from the vibrational frequencies, and ꢁS is the entropy difference
between the products and the reactants. The entropies of the free molecules at 298K and 1
atm were taken from the NIST database, while the vibrational entropy was considered for
the adsorbed species. In this work, we considered the reduction of CO2 to HCOOH and
CO products. To improve the accuracy of PBE functional in predicting the free energy of
reaction species, we considered the following three types of correction: DFT energy
correction of free CO2 and CO molecules; the solvation effect of COOH*, OH* and CO*
species; liquidꢀphase correction of H2O and HCOOH products. Details of the corrections
are provided in Supporting Information, including Tables S1 and S2.
Synthesis of [Cu32(H)20{S2P(OiPr)2}12]. The Cu32 cluster was prepared via the previously
reported method57 and details are provided in Supporting Information. Briefly,
[Cu(CH3CN)4](PF6) (0.500 g, 1.340 mmol) dissolved in 10 mL CH3CN was transferred
to the cooled mixture of NH4[S2P(OiPr2)2] and [NaBH4] suspended in 30 mL CH3CN.
After stirring at 10 oC for 4 h, 0.2 mL of MeOH was added to reaction mixture over 1 h,
and then the solvent was evaporated under vacuum. The residue was dissolved in 20 mL
4
ACS Paragon Plus Environment

Page 5 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
dichloromethane (DCM) and washed with deionized water (2 × 30 mL). After the filtrate
from the DCM layer was evaporated under vacuum, 30 mL methanol was added to the
dry residue, leading to a turbid solution that was placed in a freezer for several hours for
the precipitate to settle down. The obtained brown precipitate was further washed with
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
acetone and then dried under vacuum, yielding
a
brown precipitate of
[Cu32(H)20{S2P(OiPr)2}12] (0.079 g, 41%, based on Cu), which was confirmed by melting
point, elemental analysis, ESIꢀMS, 1HꢀNMR, and 31PꢀNMR.
Electrocatalysis experiments. Controlled potential electrolysis (CPE) experiments were
carried out for 90 min under vigorous stirring with a ZIVE MP1 potentiostat (WonATech,
Korea) in an Hꢀtype cell that was equipped with a composite working electrode, and a
Ag/AgCl (3 M NaCl) reference electrode in one compartment and a platinum plate (1.68
cm2) counter electrode in the other compartment. The working electrode and the counter
electrode were separated by a protonꢀexchange membrane (Nafion 117; SigmaꢀAldrich).
Both compartments were filled with 60 mL of 0.1 M KHCO3 and 0.4 M KCl solution.
The working electrode compartment was saturated with CO2 gas. The composite working
electrode was fabricated by spreading a catalyst ink, prepared by mixing 80 ꢂg of the
Cu32(H)20{S2P(OiPr)2}12 cluster catalyst, 200 ꢂg of carbon black (Vulcan XCꢀ72) and 3.5
ꢂL of Nafion solution (5 wt%, SigmaꢀAldrich) in 50 ꢂL of tetrahydrofuran, on a gas
diffusion layer (GDL, model N1S1007, CeTech Co., Taiwan) electrode (2 cm2). The
amounts of CO and H2 produced were quantified from an analysis of the headspace using
an Agilent 7890B gas chromatography equipped with a flame ionization detector and
1
thermal conductivity detector. HCOOH production was quantified by calibrating the H
NMR spectrum (Bruker Avance II 400 MHz NMR spectrometer) of the product solution
against 100 ppm acetic acid used as an internal standard.
3. RESULTS AND DISCUSSION
3.1. Hydrides in the Cu32H20L12 nanocluster. Figure 1 shows the optimized structure of
Cu32H20L12 nanocluster. The Cu32 skeleton can be viewed as a distorted hexacapped
rhombohedral core of 14 Cu atoms sandwiched between two nestꢀlike triangular cupola
fragments of Cu atoms (each cupola comprises of 9 Cu atoms). The 20 hydrides can be
grouped into 12 triꢀcoordinated H (µ3ꢀH, capping mode), six tetraꢀcoordinated H (µ4ꢀH,
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 23
1
2
3
4
5
6
7
8
9
interstitial mode), and two pentaꢀcoordinated H (µ5ꢀH, interstitial mode). The 12 µ3ꢀH
hydrides can be further divided into three µ3ꢀH1, two µ3ꢀH2, four µ3ꢀH3, and two µ3ꢀH4.
Both µ3ꢀH1 and µ3ꢀH2 are coordinated to the Cu3 triangle of the periphery cupola
fragments: µ3ꢀH1 neighbors to µ4ꢀH1 and µ4ꢀH2, while µ3ꢀH2 neighbors to µ4ꢀH2 and µ5ꢀ
H (as clearly seen in Figure 1). Both µ3ꢀH3 and µ3ꢀH4 (at the central part) bridge Cu3
triangles of the hexacapped rhombohedron: the difference is that one of µ3ꢀH3’s three
coordinated Cu atoms is a central Cu without any bonding to S. The six µ4ꢀH hydrides
can be divided into four µ4ꢀH1 and two µ4ꢀH2: µ4ꢀH1 is located in a tetrahedral interstitial
site at the interface between the central hexacapped rhombohedron and the periphery
cupola, while µ4ꢀH2 is located at a tetrahedral interstitial site of the periphery cupola. The
calculated Bader charges (Table 1) show that these hydrides have a negative charge of ꢀ
0.25 ~ ꢀ0.34|e|, while the Cu atoms bonded directly with the hydrides and/or the L ligands
carry a positive charge (+0.20 ~ +0.45 |e|). The many lattice hydrides in the Cu clusters
can provide the hydrogen source needed for CO2 reduction, against the commonly
assumed protonꢀreduction channel whereby hydrogen is sourced from the proton in
solution.
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 1. Atomic structure of Cu32H20L12 nanocluster (L=S2PH2). Color code: orange, Cu; green,
hydride; yellow, S; purple, P; white, H on the dithiophosphate ligands. Different types of hydrides are
indicated by the arrows.
6
ACS Paragon Plus Environment

Page 7 of 23
Journal of the American Chemical Society
1
2
3
4
5
Table 1. Average Bader charges for different types and numbers of hydrides and Cu atoms in the
Cu32H20L12 cluster (L=S2PH2). Charge variation (minimum and maximum) are for shown in brackets.
6
7
8
9
Atom type
Number
Bader charge (|e|)
ꢀ0.289 [ꢀ0.294, ꢀ0.281]
ꢀ0.318 [ꢀ0.342, ꢀ0.293]
ꢀ0.267 [ꢀ0.270, ꢀ0.263]
ꢀ0.291 [ꢀ0.313, ꢀ0.269]
ꢀ0.300 [ꢀ0.317, ꢀ0.294]
ꢀ0.307 [ꢀ0.307, ꢀ0.307]
ꢀ0.278 [ꢀ0.281, ꢀ0.274]
+0.380 [+0.319, +0.434]
+0.237 [+0.210, +0.278]
µ3-H1
4
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
µ3-H2
µ3-H3
4
µ3-H4
µ4-H1
2
4
µ4-H2
2
µ5-H
2
26
6
Cu (bonded to H and S)
Cu (bonded to H only)
3.2. CO2 reduction to HCOOH on the Cu32H20L12 nanocluster. DFT calculations
showed that the exposed capping µ3ꢀH1 hydride has the most favorable binding for the
CO2 molecule, with a binding strength of ꢀ0.24 eV (the adsorption geometry of CO2 is
shown in Supporting Information, Figure S1). The favorable interaction is attributed to
the attraction between the negatively charged hydride and the positively charged C of
CO2 reactant. After CO2 adsorption, the next key initial step of CO2 reduction is where
the first hydrogen is added (Table 2): C or O of CO2. If H is added to C, then the reaction
will proceed to form HCOOH; if H is added to O, then the reaction will proceed to form
CO.64 We first examined the addition of H to C. Table 2 shows that there are two possible
channels to form HCOO* (an intermediate adsorbed on the catalyst surface): (1) the nonꢀ
electrochemical step of CO2 reaction with one of the lattice hydrides (lattice hydride
channel); (2) the electrochemical step of reacting CO2 with the proton from solution and
electron from the electrode (protonꢀreduction channel). Considering the many hydrides
and different Cu sites in the Cu32 system, we exhaustively explored the HCOO* binding
at different hydride and Cu sites. The relative energies are summarized in Table S3, and
the structures are provided in Figures S2 (for the hydride channel) and S3 (for the protonꢀ
reduction channel). One can see that for both channels, µ3ꢀH1 and its associated Cu site
are the most active for HCOO* binding, while the energy of HCOO* varies up to 1.25 eV
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 23
1
2
3
4
5
6
for the hydride channel and 0.76 eV for the protonꢀreduction channel. So in the following
we focus on µ3ꢀH1 and its associated Cu site for CO2 reduction.
7
8
9
From the computed free energy change, we found that the lattice hydride channel
(Figure 2 right) is more facile, whereby the µ3ꢀH1 hydride in the Cu32H20L12 cluster is
transferred to the C atom to form adsorbed HCOO* (Cu32H19L12ꢀHCOO) with free
energy of 0.32 eV. In comparison, the protonꢀreduction mechanism leading to formation
of Cu32H20L12ꢀHCOO is highly unfavorable with a high overpotential of about 1.08 eV
(Figure 2 left). In the lattice hydride channel of the HCOO* adsorption configuration
(Figure S2a), the two O atoms of HCOO* are strongly bonded to the two Cu atoms on the
Cu3 triangle facet of µ3ꢀH1. Analysis of the electronic structure and the binding from the
local density of states (LDOS) for this adsorption configuration shows that the 2pꢀorbitals
of the O atom are highly hybridized with the 3dꢀorbitals of the bonded Cu atoms (Figure
S7a).
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
Table 2. The two different channels for HCOOH and CO formation from CO2 reduction. * represents
an active site on the catalyst.
Step
Lattice hydride channel
CO2 + H*↔ HCOO*
Proton-reduction channel
CO2 + * + H+ + eꢀ ↔ HCOO*
HCOO* + H+ + eꢀ ↔ HCOOH
CO2 + * + H+ + eꢀ ↔ COOH*
COOH* + H+ + eꢀ ↔ CO* + H2O
HCOOH formation
via 1st H addition to C
CO formation
HCOO* + H* ↔ HCOOH
CO2 + H* ↔ COOH*
via 1st H addition to O
COOH* + H* ↔ CO* + H2O
After forming the Cu32H19L12ꢀHCOO species, the second step is the addition of
another hydride (µ4ꢀH1) to HCOO* to yield the final product, HCOOH (ꢁG=0.07 eV;
Figure 2 right). The loss of two lattice hydrides leads to two hydrogen vacancies in the
resulting Cu32H18L12 cluster, which could readily proceed via two sequential protonꢀ
reduction steps to regenerate the two reacted lattice hydrides. We found that the first
hydrogen goes to the µ3ꢀH1 vacancy with ꢁG=0.05 eV and the second goes to the µ4ꢀH1
vacancy with ꢁG=ꢀ0.06 eV (Figure 2 right). One can see that in the more favorable
latticeꢀhydride channel, the potentialꢀlimiting step for the HCOOH formation is the first
8
ACS Paragon Plus Environment

Page 9 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
hydride transfer to form the HCOO* intermediate. We also considered alternative latticeꢀ
hydride mechanisms (six pathways in total; see Figure 3): for example, the first hydride
addition to CO2 is followed by a protonꢀreduction step instead (pathway II, III and IV in
Figure 3), or the first protonꢀreduction step is followed by a latticeꢀhydride reduction step
(pathway VI in Figure 3). But we found that they are less favorable than the doubleꢀ
hydride reductionꢀregeneration path (pathway I in Figure 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
Figure 2. Reaction scheme for CO2 electroreduction on Cu32H20L12 to form HCOOH via the protonꢀ
reduction channel (to left) and the latticeꢀhydride channel (to right). L12 is omitted when labelling the
intermediates.
Figure 3. Reaction scheme for CO2 electroreduction on Cu32H20L12 to form HCOOH via six different
pathways. The calculated free energy difference (ꢁG) for each elementary step is shown under each
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 23
1
2
3
4
5
6
intermediate state. The most favorable reaction pathway is highlighted in blue, which is the same as
the one in Figure 2 right. (L=S2PH2, not shown when labeling the cluster state in the scheme).
3.3. CO2 reduction to CO on the Cu32H20L12 nanocluster. CO2 reduction to CO is
7
another 2eꢀreduction process that competes with HCOOH formation and is initiated by 1st
H addition to O (Table 2). We also explored the binding of COOH* and CO*
intermediates at different hydrides and Cu sites (Tables S3ꢀS4 and Figures S4ꢀS6). Again,
we found that µ3ꢀH1 and its associated Cu site are the most active for COOH* and CO*
binding, while the energy of COOH* varies up to 0.70 eV for the hydride channel and
0.49 eV for the protonꢀreduction channel and the energy of CO* on the Cu site varies up
to 0.30 eV. So we focus on the µ3ꢀH1 and its associated Cu site for mechanistic insight
into CO formation. Figure 4 shows the reaction pathways for CO2 reduction to CO on the
Cu32H20L12 cluster via the latticeꢀhydride (right) and protonꢀreduction (left) channels.
Formation of Cu32H19L12ꢀCOOH via the latticeꢀhydride mechanism is much more
preferred (ꢁG=0.81 eV) than formation of Cu32H20L12ꢀCOOH via the protonꢀreduction
channel (ꢁG=1.53 eV). In the optimized structure of Cu32H19L12ꢀCOOH (Figure S5a), the
C atom of COOH* is bonded to two Cu atoms on the Cu3 facet of µ3ꢀH1 forming
tetrahedral coordination. The LDOS analysis showed strong hybridization between the
2pꢀorbital of the C atom and the 3dꢀorbitals of the bonded Cu atoms (Figure S7b).
Cu32H19L12ꢀCOOH is the precursor for COꢀOH bond cleavage, leading to CO and OH*
(ꢁG= ꢀ0.43 eV). This is followed by formation of H2O from the OH* intermediate via the
latticeꢀhydride mechanism (ꢁG=ꢀ0.16 eV). The two reacted hydrides can then be
regenerated by electrochemical protonꢀreduction processes. Here we note that the two
transferring hydrides for the CO formation are the same as those for the HCOOH
formation: first, the µ3ꢀH1 adds to O of CO2 to form COOH*; then, the OH* intermediate
from COꢀOH cleavage reacts with µ4ꢀH1 to form H2O. Again, we also checked other
possible reaction pathways (all the possible six pathways are shown in Figure 5), and the
lattice hydride channel (pathway I in Figure 5) is the most favorable pathway for CO
formation.
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
10
ACS Paragon Plus Environment

Page 11 of 23
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 4. Reaction scheme for CO2 electroreduction on Cu32H20L12 to form CO via the protonꢀ
reduction channel (to left) and the latticeꢀhydride channel (to right). L12 is omitted when labelling the
intermediates.
Figure 5. Reaction scheme for CO2 electroreduction on Cu32H20L12 to form CO via six different
pathways. The calculated free energy difference (ꢁG) for each elementary step is shown under each
intermediate state. The most favorable reaction pathway (I) is highlighted in blue, which is the same
as the one in Figure 4 right. (L=S2PH2, not shown when labeling the cluster state in the scheme).
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 23
1
2
3
4
5
6
7
8
9
Previous studies showed that the pꢀblock dopants such as P and S atoms embedded on
metal surfaces play a significant role in modulating the COOH* and CO* binding free
energies because these intermediates can directly bind on the pꢀblock dopant.65,66 We
found that the binding free energies of COOH* and CO* on the S and P sites of the ꢀ
S2PH2 ligands are higher than on Cu (Table S4), indicating that the S and P atoms from
the ligands do not play a direct role in binding the key intermediates from CO2 reduction
on the Cu32H20L12 cluster
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
3.4. Comparison of HCOOH vs CO formation from the lattice hydride mechanism
on the Cu32H20L12 nanocluster. From Figures 2 and 4 one can see that both HCOOH
and CO formations from CO2 prefer the latticeꢀhydride mechanism on Cu32H20L12. From
the freeꢀenergy diagrams, the potentialꢀdetermining step for both the HCOOH and CO
formation derives from the first hydride transfer, with the onset potential of ꢀ0.32 V and ꢀ
0.81 V vs. SHE, respectively; in other words, the overpotential is 0.32 V for HCOOH
formation and 0.81 V for CO formation. Previous theoretical studies of CO2 reduction on Cu
(211) showed that the overpotential for HCOOH and CO formation is about 0.41 eV,26
while CO formation on Cu(100) and Cu(111) needs higher overpotential (0.6 ~ 0.9 V).28
Recent experimental reports showed that nanostructured copper electrodes can reduce the
overpotential to below 0.4 V for both HCOOH and CO, compared to the polycrystalline
copper electrode.67,68 So our predicted overpotential for HCOOH is lower while that for
CO is higher on Cu32H20L12 than the reported ones on Cu surfaces or nanostructures.
To further shed light on the selectivity of HCOOH over CO on Cu32H20L12 via the
latticeꢀhydride mechanism, we located the transition states for the key steps leading to
HCOOH and CO formations (Figure 6). One can see that the HCOOH pathway is more
favorable than the CO pathway on the Cu32H20L12 cluster. The key reason is that CO
formation is energetically limited by formation of the COOH* intermediate (activation
energy: 2.12 eV) which is much less favored than the HCOO* intermediate (activation
energy: 0.89 eV). The preference of the HCOO* intermediate is due to the negatively
charged lattice hydrides favoring addition to the positively charged C in CO2, while O
atoms on HCOO* are stabilized by the positively charged surface Cu atoms. This
indicates that HCOOH should be the main product of CO2 reduction on the Cu32H20L12
cluster.
12
ACS Paragon Plus Environment

Page 13 of 23
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 6. Freeꢀenergy diagrams for HCOOH and CO formations on the Cu32H20L12 cluster via the
latticeꢀhydride mechanism. TS = transition state. L12 is omitted when labelling the intermediates. The
energy barriers are shown in parentheses.
HCOOH formation and the subsequent hydride regeneration on the Cu32H20L12
cluster are summarized in Figure 7. For comparison, CO formation cycle is provided in
Figure S8. One can clearly see that, starting from Cu32H20L12 and CO2, the µ3ꢀH1 hydride
will first add to C to yield Cu32H19L12ꢀHCOO (Figure 7a). Notably, the two O atoms of
the HCOO* intermediate are strongly bonded to the surface Cu forming a fiveꢀmembered
ring (Figure 7 top). The HCOO* subsequently reacts with µ4ꢀH1 to release the HCOOH
product (Figure 7b). The resulting Cu32H18L12 cluster with two hydride vacancies will
then proceed through the protonꢀreduction processes to refill the two latticeꢀhydride sites
(Figure 7c,d). We found that during HCOOH formation, the local structure of the Cu
atoms around the two reacting hydrides did not change much, even after the loss of the
interstitial µ4ꢀH1. This indicates that the structural stability of the Cu cluster framework
can accommodate the loss of the interstitial hydrides, thereby facilitating the latticeꢀ
hydride mechanism to produce HCOOH. For comparison, CO formation is provided in
Figure S8, which follows a similar cycle of the latticeꢀhydride mechanism.
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 23
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 7. The overall mechanism of HCOOH formation from CO2 reduction on Cu32H20L12 via the
latticeꢀhydride channel; the structures of the key intermediates and transition states are shown. Color
code: orange, Cu; green, hydride; red, oxygen; grey, carbon; L=S2PH2, not shown.
3.5. Hydrogen evolution reaction (HER) on the Cu32H20L12 nanocluster. HER is the
major competing reaction for CO2 electroreduction at low overpotentials, so it is
necessary to compare HER vs CO2 reduction on the Cu32H20L12 nanocluster to fully
evaluate its viability as an electrocatalyst for CO2 reduction. The many hydride atoms in
the Cu32H20L12 nanocluster favour the Heyrovsky mechanism whereby a proton from the
aqueous solution reacts with a hydride atom to generate H2 (Cu32H20 + H+ + eꢀ → Cu32H19
+ H2). The lost hydride will be replenished by adsorbing another proton through the
Volmer reaction (Cu32H19 + H+ + eꢀ → Cu32H20). Figure 8a shows the calculated free
energy for hydrogen evolution through the HeyrovskyꢀVolmer mechanism at different
hydride sites. We can see that at lower overpotential (< 0.3 eV), the µ4ꢀH1, µ4ꢀH2 and µ5ꢀ
H will desorb from the interstitial binding sites to generate H2, while the removal of the
surface capping hydrides (µ3ꢀH1, µ3ꢀH2, µ3ꢀH3 and µ3ꢀH4) to form H2 will require a
14
ACS Paragon Plus Environment

Page 15 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
higher overpotential (0.6 ~ 1 eV). Figure 8b shows the kinetics for the Heyrovsky
reaction at the interstitial µ4ꢀH1, µ4ꢀH2 and µ5ꢀH sites, and the activation barrier is 2.06
eV, 2.39 eV, and 1.58 eV, respectively. These barriers are too high to be competitive with
the HCOOH formation (0.89 eV). In other words, at the low overpotential, HCOOH
formation is kinetically preferred than HER. However, at high overpotentials, all these
hydride channels will be open for HER.
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 8. (a) Free energy diagram for hydrogen evolution via the HeyrovskyꢀVolmer mechanism
(Cu32H20 + H+ + eꢀ → Cu32H19 + H2; Cu32H19 + H+ + eꢀ → Cu32H20) at different hydrides on Cu32H20.
(b) Minimum freeꢀenergy path for the Heyrovsky reaction at the µ4ꢀH1, µ4ꢀH2 and µ5ꢀH sites. TS =
transition state; free energy barrier heights are shown in parentheses.
3.6. Electrocatalysis experiments of CO2 reduction on the Cu32H20L12 nanocluster.
To verify our theoretical prediction that CO2 is selectively reduced to HCOOH at low
overpotentials instead of to CO or losing to HER, we synthesized the Cu32H20L12
nanocluster and performed controlled potential electrolysis (CPE) in 0.1 M KHCO3 and
0.4 M KCl (pH 6.8) to examine its electrocatalytic activity. The asꢀsynthesized clusters
were characterized and verified to be Cu32H20L12 by various techniques, including
elemental analysis, ESIꢀMS, and ambient temperature 1H and 31P{1H} NMR
spectroscopy (Figure S9). For the electrochemical test, the Cu32H20L12 clusters were first
immobilized on carbon black (C) particles with Nafion that serves as a proton conductor
as well as a binder.69 The resulting particles were then dropcast on a GDL to form a
composite Cu32/C/GDL electrode.
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 23
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 9. (a) Average current densities (black circles) and cumulative faradaic efficiencies for H2,
HCOOH, and CO (stacked bars) obtained at different overpotentials. (b) Product selectivity for H2,
HCOOH, and CO produced at different overpotentials. Controlled potential electrolysis (CPE) was
conducted for 90 min in 0.1 M KHCO3 and 0.4 M KCl (pH 6.8) on a Cu32/C/GDL electrode (1 cm2).
Figure 9 shows the results of CO2 electroreduction on the Cu32H20L12 nanocluster.
As can be seen from Figure 9a, the average current density during 90 min CPE becomes
significant at the overpotential of 0.3 V and increases with overpotential. Figure 9a also
shows the cumulative Faradaic efficiency of the product formation after 90 min CPE at
different overpotentials: H2, HCOOH, and CO were detected as the main products
accounting for the cumulative Faradaic efficiency of >90%. The selectivity of these three
products is compared as a function of overpotential in Figure 9b. One can see that the
Cu32H20L12 nanocluster predominantly produces HCOOH (as HCOOꢀ at pH 6.8) at low
overpotentials (89% at 0.3 V and 83% at 0.4 V), with minor amounts of CO and H2. By
contrast, the product selectivity dramatically changes when the overpotential is higher
than 0.5 V where H2 is predominantly produced (85% at 0.5 V and 94% at 0.6 V). These
experimental results are in excellent agreement with our DFT prediction that HCOOH
formation is favored at low overpotential, while HER dominates at higher overpotential.
The Cu32H20L12 nanocluster is quite stable during electrocatalysis at pH 6.8: the Faradaic
efficiency of HCOOH (and H2 and CO) formation monitored at different CPE time (at the
overpotential of 0.3 V) was found to be unchanged during the first 90 min and slightly
16
ACS Paragon Plus Environment

Page 17 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
decreased at 180 min (Figure S10). We further analyzed the turnover number of the
Cu32H20L12 cluster for HCOOH production and found it to be 1740 moles of HCOOH per
mole of Cu32H20L12 after 90 min CPE, supporting the replenishment of the lattice
hydrides during the CO2 reduction. For comparison, Cu(0) nanoparticles produced mainly
CO at low overpotentials,67 while Cu foil produced mainly H2 at low overpotentials.8
3.7. Implications of the lattice hydride mechanism. So far we have shown that the key
steps and product selectivity of electroreduction of CO2 on a typical atomically precise
copperꢀhydride cluster. The vast majority of previous work of electroreduction of CO2 on
Cu has focused on flat surfaces and large Cu nanoparticles. It has been found that on the
pure Cu surface, HCOOH and CO formations are competitive in the initial
electrochemical step of CO2 reduction but CO is the major product.13 Very recent
theoretical studies of CO2 reduction on Cu(100)70 found that HCOOꢀ formation needs to
overcome a higher free energy barrier (0.80 eV) than CO formation (0.43 eV). By
contrast, on the Cu32H20L12 cluster examined here, the presence of the negatively charged
hydride facilitates the formation of HCOO* intermediate, favoring formation of HCOOH
instead of CO. So the structurally precise Cuꢀhydride clusters offer distinctly different
pathways and product selectivity. Moreover, to the best of our knowledge, our work is
the first time that the latticeꢀhydride mechanism has been proposed for electrochemical
CO2 reduction on a heterogeneous or nanocluster catalyst.
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
We have now demonstrated electroreduction of CO2 in aqueous 0.1 M KHCO3 on
the copperꢀhydride clusters. This electrolyte has also been used in electroreduction of
CO2 to CO on the Au25 nanocluster previously.71 In the future, we think that it will be
also worthwhile to use nonꢀaqueous electrolytes such as organic solvents or ionic liquids
for CO2 reduction on the copperꢀhydride clusters. As demonstrated experimentally, these
nonꢀaqueous electrolytes can supply protons in two ways: (1) by mixing a low
concentration (~100 ppm) of water in the organic or ionic liquid electrolyte;72 (2) by
using a protic solvent such as methanol.73 In addition, the latticeꢀhydride mechanism of
CO2 reduction can be a general pathway beyond the Cu clusters, given the recent success
in synthesizing other hydrideꢀcontaining transitionꢀmetal clusters.74,75
17
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 23
1
2
3
4
4. CONCLUSIONS
5
6
7
8
In summary, we have demonstrated that structurally precise copperꢀhydride
nanoclusters offer unique product selectivity for electroreduction of CO2 at low
overpotentials. Based on a prototypical such cluster, Cu32H20L12, we predicted from
density functional theory that both the hydride ligands and the surface Cu play a critical
role in promoting the reduction process, leading to HCOOH instead of CO at low
overpotentials; more important, CO2 reduction was also predicted to be kinetically
preferred than hydrogen evolution. We found that HCOOH formation proceeds through
two steps of latticeꢀhydride reduction: CO2 reacts directly with the capping hydride to
form HCOO* which then reacts with another interstitial hydride to form HCOOH. The
reacted hydrides can be regenerated facilely by proton reduction. The predictions were
confirmed by our electrochemical testing of CO2 reduction on the synthesized Cu32H20L12
cluster: HCOOꢀ was found to be the main product (> 80% selectivity), with minor
amounts of CO and H2, at low overpotentials with a Faradaic efficiency over 90% with a
turnover number of 1740 for the first 90 min. Our work hence paves the way for the use
of the structurally precise copperꢀhydride nanoclusters to explore the electrocatalytic CO2
reduction. The latticeꢀhydride mechanism can be a general path on hydrideꢀcontaining
transitionꢀmetal nanoclusters and hence may offer unique product selectivity than
traditional electrocatalysts for CO2 reduction.
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
ACKNOWLEDGEMENTS
This work was supported by the University of California, Riverside. This research
used resources of the National Energy Research Scientific Computing Center, a DOE
Office of Science User Facility supported by the Office of Science of the U.S.
Department of Energy under Contract DEꢀAC02ꢀ05CH11231. D.L. acknowledges
support from the Korea CCS R&D Center (KCRC) grant (NRFꢀ2014M1A8A1074219)
and the NRF grants (NRFꢀ2014R1A2A1A11051032 and 2009ꢀ0093823). C. W. Liu
acknowledges support from the Ministry of Science and Technology in Taiwan (MOST
103ꢀ2113ꢀMꢀ259ꢀ003)
.
18
ACS Paragon Plus Environment

Page 19 of 23
Journal of the American Chemical Society
1
2
3
4
Supporting Information Available:
5
6
7
8
• Details of corrections in terms of freeꢀenergy profiles.
• Adsorption sites, electronic structure, geometry, and energetics of key
intermediates.
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 of CO production cycle.
• Synthesis details and characterization data.
• Faradaic efficiency formation monitored at different times at a fixed overpotential.
REFERENCES
(1) Davis, S. J.; Caldeira, K.; Matthews, H. D. Science 2010, 329, 1330ꢀ1333.
(2) Parrenin, F.; MassonꢀDelmotte, V.; Kohler, P.; Raynaud, D.; Paillard, D.; Schwander, J.; Barbante,
C.; Landais, A.; Wegner, A.; Jouzel, J. Science 2013, 339, 1060ꢀ1063.
(3) Olah, G. A.; Prakash, G. K. S.; Goeppert, A. J. Am. Chem. Soc. 2011, 133, 12881ꢀ12898.
(4) Song, C. S. Catal. Today 2006, 115, 2ꢀ32.
(5) Whipple, D. T.; Kenis, P. J. A. J. Phys. Chem. Lett. 2010, 1, 3451ꢀ3458.
(6) Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Chem. Soc. Rev. 2009, 38, 89ꢀ99.
(7) Hori, Y. In Modern Aspects of Electrochemistry, No 42; Springer, New York: 2008, p 89ꢀ189.
(8) Hori, Y.; Murata, A.; Takahashi, R. J. Chem. Soc., Faraday Trans. I 1989, 85, 2309ꢀ2326.
(9) Hori, Y.; Kikuchi, K.; Suzuki, S. Chem. Lett. 1985, 1695ꢀ1698.
(10) Gattrell, M.; Gupta, N.; Co, A. J. Electroanal. Chem. 2006, 594, 1ꢀ19.
(11) Hori, Y.; Takahashi, R.; Yoshinami, Y.; Murata, A. J. Phys. Chem. B 1997, 101, 7075ꢀ7081.
(12) Hori, Y.; Kikuchi, K.; Murata, A.; Suzuki, S. Chem. Lett. 1986, 897ꢀ898.
(13) Hori, Y.; Murata, A.; Takahashi, R.; Suzuki, S. J. Am. Chem. Soc. 1987, 109, 5022ꢀ5023.
(14) Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo, T. F. Energy Environ. Sci. 2012, 5, 7050ꢀ7059.
(15) Kuhl, K. P.; Hatsukade, T.; Cave, E. R.; Abram, D. N.; Kibsgaard, J.; Jaramillo, T. F. J. Am.
Chem. Soc. 2014, 136, 14107ꢀ14113.
(16) Hori, Y.; Takahashi, I.; Koga, O.; Hoshi, N. J. Phys. Chem. B 2002, 106, 15ꢀ17.
(17) Schouten, K. J. P.; Kwon, Y.; van der Ham, C. J. M.; Qin, Z.; Koper, M. T. M. Chem. Sci. 2011,
2, 1902ꢀ1909.
(18) Schouten, K. J. P.; Qin, Z. S.; Gallent, E. P.; Koper, M. T. M. J. Am. Chem. Soc. 2012, 134,
9864ꢀ9867.
(19) Cook, R. L.; Macduff, R. C.; Sammells, A. F. J. Electrochem. Soc. 1989, 136, 1982ꢀ1984.
(20) Raciti, D.; Livi, K. J.; Wang, C. Nano Lett. 2015, 15, 6829ꢀ6835.
19
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 23
1
2
3
4
5
(21) Loiudice, A.; Lobaccaro, P.; Kamali, E. A.; Thao, T.; Huang, B. H.; Ager, J. W.; Buonsanti, R.
Angew. Chem. Int. Ed. 2016, 55, 5789ꢀ5792.
6
7
8
(22) Yang, K. D.; Ko, W. R.; Lee, J. H.; Kim, S. J.; Lee, H.; Lee, M. H.; Nam, K. T. Angew. Chem. Int.
Ed. 2017, 56, 796ꢀ800.
9
(23) Huang, Y.; Handoko, A. D.; Hirunsit, P.; Yeo, B. S. ACS Catal. 2017, 7, 1749ꢀ1756.
(24) Ma, S.; Sadakiyo, M.; Heima, M.; Luo, R.; Haasch, R. T.; Gold, J. I.; Yamauchi, M.; Kenis, P. J.
A. J. Am. Chem. Soc. 2017, 139, 47ꢀ50.
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
(25) Huan, T. N.; Simon, P.; Rousse, G.; Genois, I.; Artero, V.; Fontecave, M. Chem. Sci. 2017, 8,
742ꢀ747.
(26) Peterson, A. A.; AbildꢀPedersen, F.; Studt, F.; Rossmeisl, J.; Norskov, J. K. Energy Environ. Sci.
2010, 3, 1311ꢀ1315.
(27) Nie, X. W.; Esopi, M. R.; Janik, M. J.; Asthagiri, A. Angew. Chem. Int. Ed. 2013, 52, 2459ꢀ2462.
(28) Durand, W. J.; Peterson, A. A.; Studt, F.; AbildꢀPedersen, F.; Norskov, J. K. Surf. Sci. 2011, 605,
1354ꢀ1359.
(29) Peterson, A. A.; Norskov, J. K. J. Phys. Chem. Lett. 2012, 3, 251ꢀ258.
(30) CalleꢀVallejo, F.; Koper, M. T. M. Angew. Chem. Int. Ed. 2013, 52, 7282ꢀ7285.
(31) Xiao, H.; Cheng, T.; Goddard, W. A.; Sundararaman, R. J. Am. Chem. Soc. 2016, 138, 483ꢀ486.
(32) Montoya, J. H.; Shi, C.; Chan, K.; Norskov, J. K. J. Phys. Chem. Lett. 2015, 6, 2032ꢀ2037.
(33) Manthiram, K.; Beberwyck, B. J.; Aivisatos, A. P. J. Am. Chem. Soc. 2014, 136, 13319ꢀ13325.
(34) Reske, R.; Mistry, H.; Behafarid, F.; Cuenya, B. R.; Strasser, P. J. Am. Chem. Soc. 2014, 136,
6978ꢀ6986.
(35) Baturina, O. A.; Lu, Q.; Padilla, M. A.; Xin, L.; Li, W. Z.; Serov, A.; Artyushkova, K.; Atanassov,
P.; Xu, F.; Epshteyn, A.; Brintlinger, T.; Schuette, M.; Collins, G. E. ACS Catal. 2014, 4, 3682ꢀ3695.
(36) Jin, R. C. Nanoscale 2015, 7, 1549ꢀ1565.
(37) Jin, R.; Zeng, C.; Zhou, M.; Chen, Y. Chem. Rev. 2016, 116, 10346ꢀ10413.
(38) Joshi, C. P.; Bootharaju, M. S.; Bakr, O. M. J. Phys. Chem. Lett. 2015, 6, 3023ꢀ3035.
(39) Desireddy, A.; Conn, B. E.; Guo, J. S.; Yoon, B.; Barnett, R. N.; Monahan, B. M.; Kirschbaum,
K.; Griffith, W. P.; Whetten, R. L.; Landman, U.; Bigioni, T. P. Nature 2013, 501, 399ꢀ402.
(40) Dhayal, R. S.; Liao, J.ꢀH.; Liu, Y.ꢀC.; Chiang, M.ꢀH.; Kahlal, S.; Saillard, J.ꢀY.; Liu, C. W.
Angew. Chem. Int. Ed. 2015, 54, 3702ꢀ3706.
(41) Dhayal, R. S.; Lin, Y.ꢀR.; Liao, J.ꢀH.; Chen, Y.ꢀJ.; Liu, Y.ꢀC.; Chiang, M.ꢀH.; Kahlal, S.; Saillard,
J.ꢀY.; Liu, C. W. Chem. Eur.J 2016, 22, 9943ꢀ9947.
(42) Yang, H.; Wang, Y.; Huang, H.; Gell, L.; Lehtovaara, L.; Malola, S.; Häkkinen, H.; Zheng, N.
Nat. Commun. 2013, 4, 2422.
(43) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Science 2007, 318,
430ꢀ433.
20
ACS Paragon Plus Environment

Page 21 of 23
Journal of the American Chemical Society
1
2
3
4
5
(44) Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R. W. J. Am. Chem. Soc. 2008, 130,
3754ꢀ3755.
6
7
8
9
(45) McKenzie, L. C.; Zaikova, T. O.; Hutchison, J. E. J. Am. Chem. Soc. 2014, 136, 13426ꢀ13435.
(46) Maity, P.; Tsunoyama, H.; Yamauchi, M.; Xie, S. H.; Tsukuda, T. J. Am. Chem. Soc. 2011, 133,
20123ꢀ20125.
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
(47) Wan, X.ꢀK.; Tang, Q.; Yuan, S.ꢀF.; Jiang, D.ꢀE.; Wang, Q.ꢀM. J. Am. Chem. Soc. 2015, 137, 652ꢀ
655.
(48) Wan, X.ꢀK.; Yuan, S.ꢀF.; Tang, Q.; Jiang, D.ꢀE.; Wang, Q.ꢀM. Angew. Chem. Int. Ed. 2015, 54,
5977ꢀ5980.
(49) Wan, X.ꢀK.; Xu, W. W.; Yuan, S.ꢀF.; Gao, Y.; Zeng, X.ꢀC.; Wang, Q.ꢀM. Angew. Chem., Int. Ed..
2015, 127, 9683ꢀ9686.
(50) Dhayal, R. S.; van Zyl, W. E.; Liu, C. W. Acc. Chem. Res. 2016, 49, 86ꢀ95.
(51) Nguyen, T.ꢀA. D.; Goldsmith, B. R.; Zaman, H. T.; Wu, G.; Peters, B.; Hayton, T. W. Chem. Eur.
J. 2015, 21, 5341ꢀ5344.
(52) Huertos, M. A.; Cano, I.; Bandeira, N. A. G.; BenetꢀBuchholz, J.; Bo, C.; van Leeuwen, P. Chem.
Eur. J. 2014, 20, 16121ꢀ16127.
(53) Dhayal, R. S.; Liao, J. H.; Wang, X. P.; Liu, Y. C.; Chiang, M. H.; Kahlal, S.; Saillard, J. Y.; Liu,
C. W. Angew. Chem. Int. Ed. 2015, 54, 13604ꢀ13608.
(54) Dhayal, R. S.; Liao, J.ꢀH.; Lin, Y.ꢀR.; Liao, P.ꢀK.; Kahlal, S.; Saillard, J.ꢀY.; Liu, C. W. J. Am.
Chem. Soc. 2013, 135, 4704ꢀ4707.
(55) Nguyen, T.ꢀA. D.; Jones, Z. R.; Goldsmith, B. R.; Buratto, W. R.; Wu, G.; Scott, S. L.; Hayton, T.
W. J. Am. Chem. Soc. 2015, 137, 13319ꢀ13324.
(56) Edwards, A. J.; Dhayal, R. S.; Liao, P. K.; Liao, J. H.; Chiang, M. H.; Piltz, R. O.; Kahlal, S.;
Saillard, J. Y.; Liu, C. W. Angew. Chem. Int. Ed. 2014, 53, 7214ꢀ7218.
(57) Dhayal, R. S.; Liao, J.ꢀH.; Kahlal, S.; Wang, X.; Liu, Y.ꢀC.; Chiang, M.ꢀH.; van Zyl, W. E.;
Saillard, J.ꢀY.; Liu, C. W. Chem. Eur. J. 2015, 21, 8369ꢀ8374.
(58) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169ꢀ11186.
(59) Blochl, P. E. Phys. Rev. B 1994, 50, 17953ꢀ17979.
(60) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865ꢀ3868.
(61) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010, 132, 154104.
(62) Henkelman, G.; Uberuaga, B. P.; Jonsson, H. J. Chem. Phys. 2000, 113, 9901ꢀ9904.
(63) Nørskov, J. K.; Rossmeisl, J.; Logadottir, A.; Lindqvist, L.; Kitchin, J. R.; Bligaard, T.; Jónsson,
H. J. Phys. Chem. B 2004, 108, 17886ꢀ17892.
(64) Kortlever, R.; Shen, J.; Schouten, K. J. P.; CalleꢀVallejo, F.; Koper, M. T. M. J. Phys. Chem. Lett.
2015, 6, 4073ꢀ4082.
21
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 23
1
2
3
4
5
(65) Lim, H.ꢀK.; Shin, H.; Goddard, W. A.; Hwang, Y. J.; Min, B. K.; Kim, H. J. Am. Chem. Soc.
2014, 136, 11355ꢀ11361.
6
7
8
9
(66) Shin, H.; Ha, Y.; Kim, H. J. Phys. Chem. Lett. 2016, 7, 4124ꢀ4129.
(67) Li, C. W.; Kanan, M. W. J. Am. Chem. Soc. 2012, 134, 7231ꢀ7234.
(68) Qiao, J. L.; Jiang, P.; Liu, J. S.; Zhang, J. J. Electrochem. Commun. 2014, 38, 8ꢀ11.
(69) Kwak, K.; Choi, W.; Tang, Q.; Kim, M.; Lee, Y.; Jiang, D. E.; Lee, D. Nat. Commun. 2017, 8,
14723.
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
(70) Cheng, T.; Xiao, H.; Goddard, W. A. J. Am. Chem. Soc. 2016, 138, 13802ꢀ13805.
(71) Kauffman, D. R.; Alfonso, D.; Matranga, C.; Qian, H.; Jin, R. J. Am. Chem. Soc. 2012, 134,
10237ꢀ10243.
(72) Asadi, M.; Kim, K.; Liu, C.; Addepalli, A. V.; Abbasi, P.; Yasaei, P.; Phillips, P.; Behranginia,
A.; Cerrato, J. M.; Haasch, R.; Zapol, P.; Kumar, B.; Klie, R. F.; Abiade, J.; Curtiss, L. A.; Salehiꢀ
Khojin, A. Science 2016, 353, 467ꢀ470.
(73) Murugananthan, M.; Kaneco, S.; Katsumata, H.; Suzuki, T.; Kumaravel, M. In Green Carbon
Dioxide; John Wiley & Sons, Inc.: 2014, p 191ꢀ214.
(74) Bootharaju, M. S.; Dey, R.; Gevers, L. E.; Hedhili, M. N.; Basset, J. M.; Bakr, O. M. J. Am.
Chem. Soc. 2016, 138, 13770ꢀ13773.
(75) Ohki, Y.; Shimizu, Y.; Araake, R.; Tada, M.; Sameera, W. M. C.; Ito, J.ꢀI.; Nishiyama, H.
Angew.Chem. Int.Ed. 2016, 55, 15821ꢀ15825.
22
ACS Paragon Plus Environment

Page 23 of 23
Journal of the American Chemical Society
1
2
3
4
TOC GRAPHIC
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
23
ACS Paragon Plus Environment