Catalytic CO Reduction To Valuable Chemicals Using NiFe Based Nanoclusters A First Principles Theoretical Evaluation

View Article Online
View Journal
PCCP
Physical Chemistry Chemical Physics
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: L. Gong, J. Chen
and Y. Mu, Phys. Chem. Chem. Phys., 2017, DOI: 10.1039/C7CP06155B.
Volume 18 Number 1 7 January 2016 Pages 1–636 This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
PCCP
Physical Chemistry Chemical Physics Accepted Manuscripts are published online shortly after
www.rsc.org/pccp
acceptance, before technical editing, formatting and proof reading.

Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the

author guidelines.
Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s
ISSN 1463-9076 standard Terms & Conditions and the ethical guidelines, outlined
PERSPECTIVE
in our author and reviewer resource centre, still apply. In no
Darya Radziuk and Helmuth Möhwald
Ultrasonically treated liquid interfaces for progress in cleaning and
separation processes
event shall the Royal Society of Chemistry be held responsible
for any errors or omissions in this Accepted Manuscript or any
consequences arising from the use of any information it contains.
rsc.li/pccp
Page 1 of 36 Physical Chemistry Chemical Physics
View Article Online
DOI: 10.1039/C7CP06155B
1 Catalytic CO2 reduction to valuable chemicals using NiFe-based nanoclusters: A
2 first-principle theoretical evaluation
Physical Chemistry Chemical Physics Accepted Manuscript

3
4 Li Gonga, Jie-Jie Chena, Yang Mua,b*

Published on 25 September 2017. Downloaded by LA TROBE UNIVERSITY on 01/10/2017 20:16:01.
a
6 CAS Key Laboratory of Urban Pollutant Conversion, Collaborative Innovation
7 Centre of Suzhou Nano Science and Technology, Department of Chemistry,
8 University of Science and Technology of China, Hefei, China

b
9 Anhui Province Key Laboratory of Polar Environment and Global Change,
10 University of Science and Technology of China, Hefei, China
11
12
13
14
15
16 *Corresponding author:
17 Prof. Yang Mu, Fax: +86 551 63607907, E-mail: yangmu@ustc.edu.cn
18
1
Physical Chemistry Chemical Physics Page 2 of 36
View Article Online
DOI: 10.1039/C7CP06155B
19 Abstract
20 Converting CO2 into valuable chemicals and fuels is one of the most practical routes

21 for reducing CO2 emissions while fossil fuels continue to dominate the energy sector.
22 Nobel-metal-free NiFe bimetal nanoparticles have shown the good catalytic activity
23 in CO2 conversion. Herein we theoretically evaluated the catalytic performance and
24 possible mechanisms of NiFe-based nanoclusters for hydrogenating CO2 to form
25 formic acid and CO through bicarbonate by using a periodic and self-consistent
26 density functional theory (DFT) simulation. The theoretical results illustrated that the
27 NiFe nanoclusters could have a good catalytic activity and selectivity for HCO3-
28 reduction to formic acid and the possible pathway is that HCO3- preferred to react
29 with adsorbed H atoms of H2 on NiFe alloy nanoclusters through the carbon atom
30 site. Moreover, the NiFe alloy nanoclusters with Fe atom exposed on the surface of
31 Ni cluster showed the better performance with a lower energy barrier compared to
32 that with Fe doped in the corner of Ni cluster. However, the generation of CO from
33 HCO3- reduction was shown to be neither thermodynamically nor kinetically
34 favorable on NiFe alloy nanoclusters. Additionally, the simulation results also
35 suggested that it was thermodynamically unfavorable for a further hydrogenated
36 reduction of formic acid to formaldehyde on NiFe alloy nanoclusters themselves as
37 well as supported on graphene. In summary, a molecular-level insight of CO2
38 reduction to valuable products on NiFe nanoclusters was offered in this study, which
39 may provide some useful information for guiding the design of NiFe-based catalytic
40 materials for efficient CO2 conversion to useful fuels.
41
2
View Article Online
DOI: 10.1039/C7CP06155B
42 Introduction
43

44 Worldwide energy crisis and environmental problems induced by increasing CO2
45 emissions are serious issues faced by modern society. CO2 reduction into useful fuels
46 could not only help to reduce CO2 emission, but also provide valuable fuels for
47 energy and thus decrease the energy shortage.1, 2

Over the past three decades,
48 researchers have evaluated many methods for CO2 reduction into valuable products,
49 including photocatalytic, electrocatalytic and photo-electrochemical combined
50 ones.3-5 Among these approaches, catalytic reduction of CO2 by direct hydrogenation
51 with generated or in-site hydrogen is currently considered to be the most promising
52 feasible synthetic route in terms of economy and ecology.
53 Tremendous research efforts have been devoted to developing non-precious
54 metal catalysts for CO2 reduction, in which bicarbonate was usually adopted as CO2
55 source in water.6-9 In particular, earth-abundant 3d transition metals like Fe, Co and
56 Ni have been considered as promising catalytic materials for replacement of precious
57 catalysts due to their high theoretical activity and low cost. Nickel (Ni) is an
58 earth-abundant first row transitional metal with corrosion-resistance and ductility,
59 and it has been found that Ni nanoparticles could be a potential heterogeneous
60 catalyst for selective CO2 reduction to formic acid through bicarbonate.10,11
61 Various studies have been performed to improve the catalytic performance of
62 Ni-based catalysts, such as through decorating Ni with CrOx,12 supporting Ni on
63 carbon materials,13 and alloying Ni with foreign metals.14,15 Bimetallic alloy
64 catalysts are reported to generally exhibit enhanced catalytic performances over
65 traditional monometallic ones, due to the intimate interaction and special structure
66 configuration of two metals, which results in the alterations of the electronic and
3
View Article Online
DOI: 10.1039/C7CP06155B
67 chemical properties of the meatals.16,17 Nanosized bimetallic catalysts, especially
68 those based on nonprecious, low-cost, and abundant metals, with regards to their

69 superior performance and atomic economy, are highly demanded for various
70 interdisciplinary chemical processes such as hydrogenation/dehydrogenation,

71 fuel-cell electro catalysis ammonia decomposition, and so on.18-21 Recently
72 noble-metal-free NiFe alloy has got more and more attentions among different
73 bimetallic catalysts. Xu et al. reported that NiFe alloy nanoparticles with distinct
74 surface properties outperformed its parent metals for the complete decomposition of
75 hydrous hydrazine.22 The study from Cui et al. indicated the excellent catalytic
76 activity of NiFe alloy in the oxygen evolution reaction from water oxidation.23
77 Additionally, NiFe bimetal catalyst also showed the high catalytic activity of
78 4-chlorophenol dechlorination in water.24 Since NiFe bimetal alloy has been
79 demonstrated to have the excellent catalytic activity, it would expect that this kind of
80 materials might be a potential catalyst for CO2 reduction. However, to the best of our
81 knowledge, so far there are no efforts to explore CO2 reduction to valuable products
82 on the noble-metal-free NiFe alloy nanoparticles.
83 In this study we theoretically evaluated the catalytic performance of the
84 NiFe-based nanoclusters for CO2 reduction to valuable products through bicarbonate
85 using density functional theory (DFT). The individual adsorption of H2 and HCO3- as
86 well as their co-adsorption on different structures of the NiFe bimetal catalyst were
87 firstly investigated. Following, both thermodynamic and kinetic properties as well as
88 possible reaction mechanisms of CO2 reduction to formic acid and CO through
89 bicarbonate on different NiFe alloy nanoclusters were further explored. Finally, the
90 feasibility of further reducing formic acid on various NiFe nanoclusters as well as
91 supported on graphene was studied by using DFT simulations. In this way, a
4
View Article Online
DOI: 10.1039/C7CP06155B
92 molecular-level insight into catalytic conversion of CO2 to valuable products through
93 bicarbonate on NiFe-based nanoclusters would be offered, which may provide useful

94 information for guiding the design of new CO2-conversion materials.
95
96
97 Computational methods and models
98
99 Computational methods
100 First-principles DFT calculations were performed in the CASTEP module of
101 Materials Studio software package (Accelrys software Inc., San Diego, CA).25 The
102 intermediate structures were optimized using plane-wave basis sets at the generalized
103 gradient approximation (GGA) Perdew-Burke-Ernzerhof (PBE) level.26 The ultrasoft
104 pseudopotential was constructed using the pseudopotential generator implemented in
105 the CASTEP.27 The ultrasoft pseudopotentials to improve transferability and reduce
106 the number of plane waves required in the expansion of the Kohn-Sham orbitals. The
107 energy cut-off of 400 eV for the plane wave basis was used throughout the study,
108 which has been proven to be enough for the study of Ni13-based nanoclusters in
109 15×15×15 Å3supercell.28 The tolerances of the energy was set out to be 1×10-4 eV/cell.
110 The Brillouin zone integration was obtained with variable numbers of k-points,
111 generated by the Monkhorst-pack algorithm.29 Spin-polarized calculations were used
112 throughout the whole calculations. Only the gamma point was adopted during
113 geometry optimization, while 3 × 3 × 3 k-points were used for the determination of
114 phonon density of states. Phonon frequencies calculated by Finite displacement
115 method can be obtained with no limitations on the settings or the physical properties
116 of the system. Phonon dispersion and density of states were computed using the
5
View Article Online
DOI: 10.1039/C7CP06155B
117 method of Finite displacement. All the atoms were fully relaxed and optimized
118 without symmetry constraints. The linear synchronous transit (LST)/quadratic

119 synchronous transit (QST) methods were employed to search for the transition state
120 structure and calculate the energy barrier (Ea).30,31

121 The adsorption energy (∆Eads), a key quantity in predicting adhesive property of
122 adsorption system, was calculated to estimate the strength of a molecule-surface
123 interaction and to seize the most energetically stable adsorption model. The ∆Eads can
124 be expressed as:
125 ∆Eads = Esystem – (Esurf + Ereact) (1)
126 where the total energy values Esystem, Esurf, Ereact are relative to supercells simulating a
127 molecule interacting with nanoclusters, the nanoclusters, and an isolated molecule in
128 the vacuum, respectively. In general, a negative ∆Eads indicates that the molecule
129 adsorption is exothermic and thus the adsorption system is energetic stable.
130 The Gibb’s free energy difference (∆G) of each elementary reaction is given by
131 the following equation:
132 ∆G = ∆E + ∆ZPVE – T∆S (2)
133 where ∆E is the total energy change, directly obtained from DFT calculations, ∆ZPVE
134 is the change in zero-point energies, T is temperature (298.15 K), and ∆S is the change
135 in entropy.
136
137 Computational models
138 According to “magic numbers” n of Ni nanocluster that the icosahedral (Ih) Ni13
139 exhibits a high binding energy as well as structural stability,32 Ni13 cluster was chosen
140 for Ni nanoparticles while both Ni@Ni11Fe and Fe@Ni12 clusters for NiFe ones. Such
141 clusters are able to provide a higher geometric or electronic stability than others and
6
View Article Online
DOI: 10.1039/C7CP06155B
142 thus have been commonly adopted for metal nanoparticles during simulation.33,34 Ni13
143 nanocluster with Ih symmetry had one atom at the center and other 12 identical atoms

144 on the spherical shell surface. While for NiFe alloys, two structures were constructed
145 to explore its catalytic activities. The central or a shell Ni atom of Ni13 cluster was
146 replaced with one Fe atom to obtain the Ni@Ni11Fe and Fe@Ni12 core-shell structures,
147 respectively. The optimized structures of Ni13, Ni@Ni11Fe and Fe@Ni12 clusters are
148 shown in Fig. 1. Evidently, both Ni13 and Ni@Ni11Fe clusters had an Ih symmetric
149 structure, which belongs to a regular icosahedral structure and has all the (111) facets
150 on the surface. Ni@Ni11Fe and Fe@Ni12 were placed in a 15×15×15 Å3 cubic
151 supercell during the calculation of CO2 reduction process. Moreover, a (5×5) supercell
152 of graphene with the periodic boundary conditions on the x-y plane was employed to
153 investigate catalytic activities of graphene-supported NiFe alloys for CO2 reduction
154 process
155
156
157 Results and discussion
158 Several previous studies have adopted bicarbonate as the primary CO2 source for the
159 reduction on Ni nanocluster, Au electrode as well as Au-based catalysts.35-37 CO was
160 the dominant product in CO2 reduction through bicarbonate on Au-based catalysts,
161 while formic acid was selectively formed on the Ni nanocluster. In addition, it has
162 been found that CO was the most important intermediate in CO2 reduction to methane,
163 methanol or long chain hydrocarbons.38,39 Therefore, the formation of both formic
164 acid and CO from CO2 reduction on NiFe bimetals was taken into account in order to
165 understand the HCO3- reduction mechanisms on NiFe bimetal surface as follows.
166
7
View Article Online
DOI: 10.1039/C7CP06155B
167 H2 and HCO3- adsorption on NiFe nanoclusters
168 Reactants adsorption and activation on catalysts are important issues in CO2 reduction

169 process, which can help us to better understand the activity of catalysts.40 Thus,
170 understanding of the competitive adsorption between CO2 and HCO3- is critical for
171 exploring the possible mechanisms of CO2 reduction through HCO3- on Ni-based
172 catalysts. The adsorption of HCO3- and H2 on the Ni@Ni11Fe and Fe@Ni12
173 nanoclusters were initially explored by optimizing the structures in molecularly or
174 dissociatively adsorbed state. The location and orientation of the adsorbed H2
175 molecule as well as the adsorption sites of Ni@Ni11Fe and Fe@Ni12 nanoclusters had
176 significant effects on the adsorbed structure and the relative stability of each adsorbed
177 structure. A number of stable configurations with H-H of H2 elongated were found for
178 the adsorption of H2 on the Ni@Ni11Fe and Fe@Ni12 nanoclusters, as shown in Fig.
179 S1 of the supporting information (SI). The estimated H2 adsorption energies for
180 different configurations on the Ni@Ni11Fe and Fe@Ni12 nanoclusters are given in
181 Table 1. The preferred H2 adsorption structure on the Ni@Ni11Fe nanocluster released
182 an Eads of 0.81 eV. The two dissociative H atoms of H2 bridged on one Ni site of
183 Ni@Ni11Fe into adsorbed surface hydrogen (Ni-Hads), as illustrated in the
184 H2-Ni@Ni11Fe-t1 structure of Fig. S1. The parallel coordination with two H atoms
185 adsorbed on the two Ni sites of Fe@Ni12 was the most preferred H2 adsorption
186 configuration for Fe@Ni12 nanocluster, releasing an Eads of 1.58 eV, as shown in the
187 H2-Fe@Ni12-b2 structure of Fig. S1, which is almost 2-fold higher than that of H2
188 adsorption structure on Ni@Ni11Fe. This finding indicated that surface hydrogen
189 (Ni-Hads) was more likely to be formed on the Fe@Ni12 nanocluster than on the
190 Ni@Ni11Fe nanocluster.
191 As shown in Fig. S2, six HCO3- adsorption configurations on Ni@Ni11Fe and
8
View Article Online
DOI: 10.1039/C7CP06155B
192 Fe@Ni12 nanoclusters were explored, with respects to HCO3- adsorption sites and the
193 orientation of HCO3-. The adsorption energies of the six HCO3- adsorption

194 configurations with measurements of the distances of formed O-Ni and O-Fe bonds
195 (dO-Ni, dO-Fe) for each arrangement are listed in Table 2. HCO3- adsorption energies of
196 the six configurations ranged from -1.28 to -3.62 eV, higher than those reported for
197 CO2 adsorption on the active (100) surface of Ni and Fe,7 implying that HCO3- could
198 competitively adsorbed on the NiFe alloy surface with the existence of CO2. Two
199 adsorption structures were obtained with adsorption energies of only -2.12 and -1.28
200 eV for the Ni@Ni11Fe nanocluster, respectively. In the relatively stable HCO3-
201 adsorption geometry on Ni@Ni11Fe (Fig. S2a), O1 and O2 atoms bridged two surface
202 Ni atoms with shorter O-Ni bonds of 1.97 and 2.02 Å, respectively, while the O1 and
203 O2 atoms were bond through one Ni atom on Ni@Ni11Fe nanocluster in the less stable
204 configuration (Fig. S2b). The strongest nanocluster-HCO3- interaction was found for
205 the configuration of Fig. S2e on Fe@Ni12 nanocluster in which O1 and O2 atoms of
206 HCO3- were bonded through a Ni atom and Fe atom of Fe@Ni12 nanocluster. This
207 configuration had the lowest adsorption energy of -3.62 eV. Moreover, the formed
208 O-Fe and O-Ni bonds of 1.98 Å and 1.96 Å were relatively shorter than those in other
209 adsorption configurations indicating strong interaction of HCO3- molecule with
210 Fe@Ni12 nanocluster. Marginally less favorable adsorption configuration was the
211 situation where HCO3- adsorbed on Fe@Ni12 nanocluster through two O-Ni bonds
212 with the bonds distance of 1.97 Å and 1.92 Å, respectively (Fig. S2f). In summary,
213 HCO3- could be stably adsorbed on the two NiFe nanoclusters but more energy
214 favorable for Fe@Ni12 one.
215 Co-adsorption of reactants are not only highly important such as in the case of
216 hydrogenation reactions on Pd-based catalysts41, but also meaningful to map out the
9
View Article Online
DOI: 10.1039/C7CP06155B
217 HCO3- reduction mechanisms on the Ni-based catalysts. Since both H2 and HCO3-
218 molecule could be stably adsorbed on Ni@Ni11Fe and Fe@Ni12 nanoclusters,

219 geometry optimizations of H2 and HCO3- co-adsorption on Ni@Ni11Fe and Fe@Ni12
220 nanoclusters were conducted to determine the optimized configurations for HCO3-
221 reduction process on two nanoclusters, respectively. Based on the above results for
222 individual H2 and HCO3- adsorption on nanoclusters, the optimized geometries for the
223 co-adsorption of H2 and HCO3- on Ni@Ni11Fe and Fe@Ni12 nanoclusters are shown
224 in Fig. 2, and consequently the estimated adsorption energy (Eco-ads) are presented in
225 Table S1. Only one favorable geometry was considered for the co-adsorption of H2
226 and HCO3- on Ni@Ni11Fe nanocluster with an Eco-ads of -2.61 eV (Fig. 2a). Among
227 three optimized geometries for the co-adsorption of H2 and HCO3- on Fe@Ni12
228 nanocluster, the most energy-favored structure is shown in Fig. 2d with a lowest
229 binding energy of -4.60 eV, in which H2 molecule bridged two Ni atoms while HCO3-
230 was adsorbed through binding O1 and O2 atoms with one Ni atom and one Fe atom
231 from nanocluster. Moreover, it was found that the presence of surface H atom at the
232 NiFe alloy nanoclusters could leave the HCO3- adsorption geometry almost
233 unchanged. Similar result was also obtained for CO2 reduction on Ni (111) surface
234 reported by Mavrikakis et al.42 The co-adsorption configurations presented in Figs. 2a
235 and 2d were chosen for the following simulations.
236
237 HCO3- reduction to formic acid and CO on Ni@Ni11Fe and Fe@Ni12 nanoclusters
238 Based on the studies of CO2 reduction on Ni catalysts as well as through bicarbonate,
10, 36, 43
239 paths for HCO3- reduction to formic acid and CO with dissociative H2
240 adsorbed on the NiFe nanoclusters are depicted in Scheme S1 of the SI. For HCO3-
241 reduction to formic acid (Paths 1 and 2 in Scheme S1), H2 and HCO3- co-adsorbed
10
View Article Online
DOI: 10.1039/C7CP06155B
242 initially on the NiFe nanoclusters, and then HCO3- reacted with H atom at either C or
243 O1 site. After that, the formed intermediate reacted with another H atom of H2.

244 Eventually, the adsorbed HCOO- was formed with a H2O taken off from the
245 nanocluster. The CO formation through bicarbonate on NiFe nanoclusters is depicted

246 in Path 3 of Scheme S1 with the O1 atom dissociated from HCO3- initially. Path 1
247 proceeded through states 1, 2, and 3 to final state 4, while Path 2 through states 1, 2, 5
248 and 6 to final state 4 with CO formation through states 1, 7 and 8. CO2-, a key reaction
249 intermediate in gas CO2 reduction,31 could be generated in both Paths 1 and 2 of
250 Scheme S1. Path 1 is similar to the one proposed in a previous study.44 In Path 2,
251 HCO3- reduction process involves the step mentioned in the research HCO3-
252 hydrogenation on the rhodium complex: OH elimination from the HCO3-, resulting in
253 a CO2 molecule bound to the catalysts (step 5 → 6 ).45 Calculated free energies (G) of
254 the geometry optimized reaction intermediates in three paths are given in Table S2,
255 and consequently free energy changes (∆G) of each step in three paths could be
256 estimated (Table S3). As a result, the free energy profiles for all the steps of HCO3-
257 reduction to HCOO- with intermediate structures through the Paths 1 and 2 as well as
258 CO formation on Ni@Ni11Fe and Fe@Ni12 nanoclusters are compared in Figs. 3, 4
259 and 5, respectively.
260 HCO3- reduction to formic acid on Ni@Ni11Fe and Fe@Ni12 nanoclusters. As
261 illustrated in Fig. 3, the overall process for HCO3- reduction to formic acid on
262 Ni@Ni11Fe nanocluster was thermodynamically feasible with the total ∆G of -2.46 eV
263 from state 1 to 4. The co-adsorption of H2 and HCO3- on Ni@Ni11Fe nanocluster was
264 exothermic with ∆G of -2.82 eV, which implies that the process was
265 thermodynamically favorable. The process from state 2 to 3, i.e. an active H atom of
266 H2 reacting with HCO3- at the C site, was the rate-limiting step in Path 1 with an
11
View Article Online
DOI: 10.1039/C7CP06155B
267 energy barrier (Ea) of 1.61 eV and a free energy change (∆G) of 1.22 eV. While in the
268 pathway 2, HCO3- reacting with an active H atom of H2 at O3 site of HCO3- (i.e. Step

269 2 → 5) was the rate-controlling step, with higher Ea of 2.70 eV and ∆G of 1.71 eV
270 compared to the pathway 1. Eventually HCOO- was formed spontaneously in both
271 pathways on Ni@Ni11Fe nanocluster. In conclusion, HCO3- reduction to formic acid
272 process over Ni@Ni11Fe nanocluster in water is more likely to proceed through Path 1
273 dominantly. This finding is as same as our previous study for HCO3- reduction on Ni
274 nanocluster, but Ni@Ni11Fe nanocluster showed better catalytic activity for HCO3-
275 reduction to formic acid compared to Ni one as it afford a lower energy barrier (1.62
276 vs 3.24 eV).43
277 As depicted in Fig. 4, HCO3- reduction to formic acid process could also
278 spontaneously occur on Fe@Ni12 nanocluster with ∆G of -2.67 eV. Firstly, HCO3- and
279 H2 moved close to the surface and settled in a stable state of co-adsorption on
280 Fe@Ni12 nanocluster (i.e. step 1 → 2). The following step in both paths (i.e. step 2 →
281 3 in Path 1, and step 2 → 5 in Path 2) was the rate-controlling one for HCO3-
282 reduction to formic acid on the Fe@Ni12 nanocluster. Moreover, adsorbed HCO3-
283 reacted with an active H atom of H2 at the C site (i.e. step 2 → 3 in Path 1) with an
284 energy barrier of 0.91 eV, which was lower than that in Path 2 (i.e. step 2 → 5). Such
285 a low energy barrier for the rate-controlling step indicates that Path 1 might be the
286 dominant one for HCO3- reduction to formic acid on the Fe@Ni12 nanocluster. Finally,
287 formic acid was formed on Fe@Ni12 nanocluster spontaneously by reacting with the
288 other H atom of H2. Compared with Ni@Ni11Fe and Ni13 nanoclusters, Fe@Ni12 one
289 afforded the lowest energy barriers for HCO3- reduction to formic acid.
290 In summary, the above results demonstrated that hydrogenated reduction of
291 HCO3- to formic acid could take place spontaneously on NiFe nanoclusters through
12
View Article Online
DOI: 10.1039/C7CP06155B
292 Path 1 at 298.15 K and 1 atm. This is consistent with previous experimental results
293 that HCO3- could reduce to formic acid on Ni nanocluster acid at ambient

294 conditions.37 The theoretical analysis illustrated that the NiFe alloyed nanoclusters
295 (Ni@Ni11Fe and Fe@Ni12) showed a better catalytic activity for HCO3- reduction to
296 formic acid with lower energy barriers compared to Ni13 one.43 This is likely due to
297 that HCO3- and H2 could stably co-adsorbed on NiFe alloy nanoclusters and moreover
298 their interactions with NiFe alloy nanoclusters were stronger compared with Ni13 one,
299 resulting in an easy activation of H2 and HCO3- during HCO3- reduction. On the other
300 hand, above results also illustrated that catalytic ability of Ni13 nanocluster for HCO3-
301 reduction could be significantly enhanced by introducing the foreign metal Fe. This
302 finding is consistent with several previous studies, in which the catalytic performance
303 of such as Pd, Au, and Cu in CO2 reduction was remarkably improved when
304 combined with foreign metals.46, 47 In addition, Ni-Fe cubanes as analogs to the active
305 sites of carbon monoxide dehydrogenase (CODH) enzymes have also been proven to
306 be able to lower the energy barrier for CO2 reduction compared to other cubanes.48
307 Moreover, the theoretical simulations suggested that Fe@Ni12 nanocluster with Fe
308 doped on Ni surface had a superior catalytic activity for the reduction of HCO3- to
309 formic acid because of a lower energy barrier for the rate-limiting step. This might be
310 due to that H2 and HCO3- molecules could be more efficiently activated on the surface
311 of Fe@Ni12 nanocluster resulting from the co-function of both Ni and Fe sites on the
312 surface of the Fe@Ni12 nanocluster. These results would provide a useful guideline
313 for the design of NiFe-based catalysts.
314 CO formation through bicarbonate reduction on Ni@Ni11Fe and Fe@Ni12
315 nanoclusters. COOH, a key intermediate for the low-temperature water gas shift
316 reaction on Cu (111),49 could be produced from HCO3- reduction with simultaneous
13
View Article Online
DOI: 10.1039/C7CP06155B
317 NiO formation and then hydrogenated to formic acid on NiFe alloy surface, as
318 depicted in Path 3 of Scheme S1.

319 On Ni@Ni11Fe nanocluster, O1 atom dissociated from HCO3- molecule and
320 bridged on Ni@Ni11Fe surface with the formation of COOH, which bridged on
321 Ni@Ni11Fe surface with C and O2 atoms, as illustrated in Figure 5. The calculated
322 reaction energy and energy barrier for COOH formation were 0.01 and 2.50 eV,
323 respectively. The energy barrier for COOH formation in Path 3 was about 0.89 eV
324 higher than that of the rate-limiting step for formic acid generation in Path 1. This
325 implies that the formation of formic acid on Ni@Ni11Fe surface is more likely
326 through Path 1 rather than Path 3 via COOH intermediate, which is similar with CO2
327 reduction on Ni (111) surface reported by Mavrikakis et al.42 OH group in COOH
328 reacted with H atom resulting in CO generation on Ni@Ni11Fe surface (Steps 7 → 8
329 in Scheme S1) with a reaction energy of 1.2 eV, implying that CO formation from
330 HCO3- reduction is thermodynamically unfavorable on Ni@Ni11Fe surface. As
331 illustrated in Figure 5, CO molecule was vertical to Ni@Ni11Fe surface with C atom
332 interacting with three Ni atoms, which is consistent with experimental results that
333 CO could be adsorbed on Ni surface through hollow site.50 It is shown that the
334 energy barrier of CO formation (Ea = 1.68 eV) in Path 3 was lower than that of
335 COOH formation, but still higher than that of the rate-limiting step for formic acid
336 formation in Path 1. Therefore, CO is difficult to be formed in Path 3 neither
337 thermodynamically or kinetically on Ni@Ni11Fe surface.
338 On Fe@Ni12 surface, O1 atom also dissociated from HCO3- molecule and
339 bridged on two Ni atoms on Ni@Ni11Fe surface with COOH formation initially, as
340 illustrated in Figure 5. COOH was bridged on Fe@Ni12 surface through Ni and Fe
341 atoms with a calculated formation energy of 0.44 eV. The energy barrier of COOH
14
View Article Online
DOI: 10.1039/C7CP06155B
342 formation on Fe@Ni12 surface was estimated at 2.72 eV, about two times higher than
343 that of the rate-limiting step for formic acid formation in Path 1, indicating that it is

344 hard to generate COOH on Fe@Ni12 surface. Following, CO was formed
345 accompanying with the departure of a H2O molecule from Fe@Ni12 surface, which
346 resulted in a reaction energy of 3.31 eV and an energy barrier of only 0.77 eV,
347 respectively. This implies that the formation of CO on Fe@Ni12 surface is difficult
348 thermodynamically but favorable in terms of kinetics. Obviously, HCO3- conversion
349 to CO is difficult to proceed on Fe@Ni12 surface and the rate-limiting step of Path 3
350 still lies on the step of COOH formation. Lower energy barrier of CO formation
351 from COOH on Fe@Ni12 surface compared to Ni@Ni11Fe surface (Ea = 0.77 eV vs
352 1.68eV) demonstrates the superior catalytic activity of Fe@Ni12 nanocluster.
353 However, the energy barrier of the rate-limiting step in Path 3 on Fe@Ni12 surface,
354 i.e., COOH formation, is much higher that of formic acid formation in Path 1.
355 In summary, HCO3- conversion to CO in Path 3 is thermodynamically
356 unfavorable on the surface of NiFe alloy at room temperature. Previous experimental
357 results showed that CO formation form HCO3- reduction could happen at high
358 temperature or with an external potential.7, 8 For instance, it was reported that CO
359 formation from CO2 on Ni catalysts surface required temperature higher than 160 oC.7
360 Besides, Xu et al. found that CO was mainly from the conversion of HCO3- rather
361 than CO2 molecule on Au surface but with external potential addition.8 Different from
362 CO formation with CO2 reduction on Ni surfaces,3 COOH is a key intermediate in CO
363 formation from HCO3- reduction. The energy barriers of both COOH formation step
364 and rate-limiting step of HCO3- conversion to CO on the surface of NiFe alloy are
365 lower than previously reported on Ni13 nanocluster surface, indicating improvement in
366 catalytic activity of NiFe alloyed nanoclusters compared to Ni13 one. Besides, energy
15
View Article Online
DOI: 10.1039/C7CP06155B
367 barrier of CO formation from COOH is the lowest among all the steps involving in
368 HCO3- conversion on NiFe alloyed nanoclusters, implying the great potential of NiFe

369 alloys in gas phase CO2 conversion to methane, methanol as well as long chain
370 hydrocarbons based on previous studies.9, 10 However, HCO3- conversion to CO on

371 NiFe alloys surface is kinetically unfavorable, which could result in selective
372 reduction of HCO3- to formic acid. Therefore, lower the energy barrier of COOH
373 formation on NiFe alloys surface would be the key issue in improving HCO3-
374 conversion to CO, leading to the further formation of methane, methanol as well as
375 long chain hydrocarbons.
376
377 Possibility of HCOOH reduction to HCHO on NiFe-based catalysts.
378 HCHO is of great industrial significance as it is a crucial building block for daily life
379 commodities and a major reactive C1 source.51-53 Herein we present the possibility of
380 formic acid reduction to formaldehyde on the NiFe-based catalysts.
381 A series of DFT calculations were conducted to explore the possibility of
382 HCOOH further reduction to HCHO, shown in Scheme S2, on Ni@Ni11Fe and
383 Fe@Ni12 nanoclusters (HCOO-* → CHO-*, * denotes adsorbed state). The adsorption
384 of HCOO- on Ni@Ni11Fe and Fe@Ni12 nanoclusters were theoretically investigated.
385 As shown in Table S4 and Fig. S3, HCOO- could be stably adsorbed on Ni@Ni11Fe
386 and Fe@Ni12 nanoclusters, where HCOO- tended to bridge on two Ni atoms through
387 O1 and O2 atoms on Ni@Ni11Fe nanoclusters, while two Ni atoms or one Ni and one
388 Fe atom for Fe@Ni12 nanoclusters. Calculated thermodynamic properties (∆G) and
389 energy barrier (Ea) for HCOO-* reduction to CHO-* on Ni@Ni11Fe, Fe@Ni12 at 298.15
390 K and 1 atm are listed in Table S5 and the optimized reaction intermediates are given
391 in Fig. 6. The ∆G of CHO- formation from HCOO-* reduction on Ni@Ni11Fe and
16
View Article Online
DOI: 10.1039/C7CP06155B
392 Fe@Ni12 nanoclusters were 1.49 and 3.23 eV, respectively. This indicated that
393 HCOOH reduction to HCHO on these three catalysts at 298.15 K and 1 atm could

394 only happen with extra energy of 2.54, 1.49, and 3.23 eV, respectively. The first step,
395 i.e. HCOO-* reacting with an active H of H2 to form HCOOH, was the rate-limiting
396 step on Ni@Ni11Fe and Fe@Ni12 nanoclusters with Ea of 2.62 and 2.67 eV,
397 respectively. Similar with the boron-doped diamond electrode, these results indicated
398 that NiFe nanoclusters were also inefficient for HCO3- reduction to formaldehyde
399 through formic acid.54
400 Graphene has been proved to be an efficient carrier to support metals and metal
401 oxides to enhance their catalytic activities.55-57 Therefore, we explored
402 graphene-supported Ni@Ni11Fe and Fe@Ni12 nanoclusters for formic acid (HCOO-*)
403 reduction to formaldehyde (CHO-*) with the hope to find a more efficient catalyst.
404 The interactions of Ni@Ni11Fe and Fe@Ni12 nanoclusters with graphene were
405 explored, with the adsorption configurations shown in Fig. 7 and the adsorption
406 energy given in Table S6. Accordingly, Ni@Ni11Fe nanocluster could be stably
407 adsorbed on graphene either with a Ni atom of Ni@Ni11Fe interaction, i.e.,
408 Ni@Ni11Fe/G-1, or with the plane formed by three Ni atoms, i.e. Ni@Ni11Fe/G-2 in
409 Fig. 7. For Fe@Ni12 nanocluster supported on graphene, four configurations were
410 taken into consideration with regard to the adsorption orientation and the interacting
411 sites. The interaction of graphene with Fe@Ni12 nanocluster through the plan formed
412 by three Ni atoms (Fe@Ni12-2 in Fig. 7) or one Fe and two Ni (Fe@Ni12-4 in Fig. 7)
413 were much more stronger than the cases, where Fe@Ni12 nanocluster settled on
414 graphene with one Ni atom or one Fe binding with the C atoms of graphene
415 (Fe@Ni12-1 and Fe@Ni12-3 in Fig. 7). As a result, the configurations of
416 Ni@Ni11Fe/G-1, Fe@Ni12/G-2 and Fe@Ni12/G-4 were considered for the adsorption
17
View Article Online
DOI: 10.1039/C7CP06155B
417 and activation of HCOO- and H2 in the following simulations.
418 The estimated adsorption energies of H2 and HCOO- on graphene-supported

419 Ni@Ni11Fe and Fe@Ni12 nanoclusters are given in Tables S7 and S8, respectively,
420 and the HCOO- adsorption configurations are presented in Fig. S4. The results
421 showed that H2 could be stably adsorbed and activated on graphene-supported
422 Ni@Ni11Fe and Fe@Ni12 nanoclusters. However, the positive adsorption energies of
423 HCOO- adsorbed on graphene supported Ni@Ni11Fe and Fe@Ni12 nanoclusters
424 indicates that HCOO- was unable to be stably adsorbed and activated on graphene
425 supported Ni@Ni11Fe and Fe@Ni12 nanoclusters, as shown in Table S8.
426 The calculated thermodynamic properties (∆G) and energy barrier (Ea) in formic
427 acid reduction to formaldehyde process on graphene supported Ni@Ni11Fe and
428 Fe@Ni12 nanoclusters at 298.15 K and 1 atm are given in Table S9, with the geometry
429 optimized reaction intermediates illustrated in Fig. 8. The results indicated that formic
430 acid reduction to formaldehyde process on graphene-supported NiFe alloy
431 nanoclusters was still thermodynamically unfavorable. In terms of kinetic properties
432 of the formic acid reduction to formaldehyde process on graphene-supported
433 Ni@Ni11Fe and Fe@Ni12 nanoclusters, the rate-limiting step of the process lay in the
434 second step, i.e. HCOOH+ H+ + e-→ CHO- + H2O, with an Ea of 4.34 and 4.20 eV,
435 respectively.
436 In summary, further hydrogenated reduction of formic acid to formaldehyde was
437 thermodynamically unfeasible on either Ni@Ni11Fe and Fe@Ni12 nanoclusters or
438 graphene-supported ones at 298.15 K and 1 atm. This might be able to further explain
439 the selectivity of formic acid formation on Ni nanoclusters (Fig. 5 and Table S5).43,44
440 Among different catalysts, a lowest extra energy of 1.17 eV would be required to
441 make this reaction occurring on the graphene supported Ni@Ni11Fe. In terms of
18
View Article Online
DOI: 10.1039/C7CP06155B
442 kinetic, the energy barriers of the rate-limiting step in formic acid reduction to
443 formaldehyde process on Ni@Ni11Fe and Fe@Ni12 nanoclusters were relatively lower

444 than that of Ni13 one, suggesting the improved kinetics of formic acid reduction. On
445 the other hand, thanks to the adsorption and activation of H2 on graphene supported
446 Ni@Ni11Fe and Fe@Ni12 nanoclusters, the Ea values of the first step for reduction
447 formic acid to formaldehyde on graphene-supported Ni@Ni11Fe and Fe@Ni12
448 nanoclusters were lower than those on Ni@Ni11Fe, Fe@Ni12 and Ni13 ones (TableS5).
449 However, the Ea values of the rate-limiting step for this reaction on
450 graphene-supported Ni@Ni11Fe and Fe@Ni12 nanoclusters were higher compared to
451 others, which might be due to the fact that HCOO- cannot be stably adsorbed and
452 activated on graphene supported Ni@Ni11Fe and Fe@Ni12 nanoclusters. Consequently,
453 although graphene supported Ni@Ni11Fe and Fe@Ni12 nanoclusters could not be used
454 for the reduction of HCOO-, they might be the good catalysts for activation of H2 in
455 hydrogenation reactions for such as chemical industry and water containment
456 treatment.48,49
457
458
459 Conclusions
460 The catalytic performance and possible mechanisms for CO2 reduction through HCO3-
461 on NiFe alloy nanoclusters have been explored theoretically in this study. The results
462 illustrated that Ni@Ni11Fe and Fe@Ni12 nanoclusters had good catalytic activity and
463 selectivity for CO2 reduction through HCO3- to formic acid. Compared with
464 Ni@Ni11Fe nanocluster, Fe@Ni12 one showed superior catalytic activity for CO2
465 reduction. Additionally, the results indicated that the formation of CO was difficult
466 through HCO3- on NiFe nanoclusters either thermodynamically or kinetically,
19
View Article Online
DOI: 10.1039/C7CP06155B
467 resulting in selective conversion of HCO3- to formic acid. Moreover, the simulation
468 results also suggested that it was thermodynamically unfavorable for a further

469 hydrogenated reduction of formic acid to formaldehyde on NiFe alloy nanocluster
470 themselves as well as supported on graphene.

471
472
473 Acknowledgements
474 The authors wish to thank the Natural Science Foundation of China (51538012 and
475 51478446), the Recruitment Program of Global Experts, and the Fundamental
476 Research Funds for the Central Universities for financially supporting this study.
477 Supercomputing center of USTC is acknowledged for computational support.
478
479
480 Notes and references
481 1. C. W. Li, J. Ciston and M. W. Kanan, Nature, 2014, 508, 504-507.
482 2. R. X. Liu, J. H. Guo, G. Ma, P. Jiang, D. H. Zhang, D. X. Li, L. Chen, Y. T. Guo
483 and G. L. Ge, ACS Appl Mater Inter, 2016, 8, 16833-16844.
484 3. Y. Tamaki, T. Morimoto, K. Koike and O. Ishitani, P Natl Acad Sci USA, 2012,
485 109, 15673-15678.
486 4. S. C. Roy, O. K. Varghese, M. Paulose and C. A. Grimes, ACS Nano, 2010, 4,
487 1259-1278.
488 5. X. X. Chang, T. Wang and J. L. Gong, Energ Environ Sci, 2016, 9, 2177-2196.
489 6. M. D. Porosoff, X. F. Yang, J. A. Boscoboinik and J. G. G. Chen, Angew Chem
490 Int Edit, 2014, 53, 6705-6709.
491 7. C. Liu, T. R. Cundari and A. K. Wilson, J Phys Chem C, 2012, 116, 5681-5688.
20
View Article Online
DOI: 10.1039/C7CP06155B
492 8. A. Dhakshinamoorthy, S. Navalon, A. Corma and H. Garcia, Energ Environ Sci,
493 2012, 5, 9217-9233.

494 9. K. P. Kuhl, E. R. Cave, D. N. Abram and T. F. Jaramillo, Energ Environ Sci,
495 2012, 5, 7050-7059.

496 10. G. A. Bhaduri and L. Siller, Catal Sci Technol, 2013, 3, 1234-1239.
497 11. G. A. Bhaduri, M. A. H. Alamiry and L. Siller, J Nanomater, 2015, 581785.
498 12. S. F. Lu, J. Pan, A. B. Huang, L. Zhuang and J. T. Lu, P Natl Acad Sci USA,
499 2008, 105, 20611-20614.
500 13. Z. B. Zhuang, S. A. Giles, J. Zheng, G. R. Jenness, S. Caratzoulas, D. G.
501 Vlachos and Y. S. Yan, Nat Commun, 2016, 7.
502 14. J. W. Wang, J. W. Liu, N. T. Yang, S. S. Huang, Y. H. Sun and Y. Zhu, Nanoscale,
503 2016, 8, 3949-3953.
504 15. S. De, J. G. Zhang, R. Luque and N. Yan, Energ Environ Sci, 2016, 9,
505 3314-3347.
506 16. S. Guo, W. Tian and L. Shi, Transit Metal Chem, 2012, 37, 757-763.
507 17. D. A. Hansgen, D. G. Vlachos and J. G. G. Chen, Nat Chem, 2010, 2, 484-489.
508 18. F. Studt, I. Sharafutdinov, F. Abild-Pedersen, C. F. Elkjaer, J. S. Hummelshoj, S.
509 Dahl, I. Chorkendorff and J. K. Norskov, Nat Chem, 2014, 6, 320-324.
510 19. W. J. Liu, T. T. Qian and H. Jiang, Chem Eng J, 2014, 236, 448-463.
511 20. G. Krishnan, R. F. Negrea, C. Ghica, G. H. ten Brink, B. J. Kooi and G.
512 Palasantzas, Nanoscale, 2014, 6, 11963-11970.
513 21. A. C. Alba-Rubio, J. L. G. Fierro, L. León-Reina, R. Mariscal, J. A. Dumesic
514 and M. López Granados, Appl Catal B: Environ, 2017, 202, 269-280.
515 22. S. K. Singh, A. K. Singh, K. Aranishi and Q. Xu, J Am Chem Soc, 2011, 133,
516 19638-19641.
21
View Article Online
DOI: 10.1039/C7CP06155B
517 23. X. J. Cui, P. J. Ren, D. H. Deng, J. Deng and X. H. Bao, Energ Environ Sci,
518 2016, 9, 123-129.

519 24. F. Y. Xu, S. B. Deng, J. Xu, W. Zhang, M. Wu, B. Wang, J. Huang and G. Yu,
520 Environ Sci Technol, 2012, 46, 4576-4582.

521 25. M. D. Segall, P. J. D. Lindan, M. J. Probert, C. J. Pickard, P. J. Hasnip, S. J.
522 Clark and M. C. Payne, J Phys-Condens Mat, 2002, 14, 2717-2744.
523 26. J. P. Perdew, K. Burke and M. Ernzerhof, Phys Rev Lett, 1996, 77, 3865-3868.
524 27. D. Vanderbilt, Phys Rev B, 1990, 41, 7892-7895.
525 28. P. Cheng, S. L. Zhang, P. Wang, S. P. Huang and H. P. Tian, Comput Theor
526 Chem, 2013, 1020, 136-142.
527 29. H. J. Monkhorst and J. D. Pack, Phys Rev B, 1976, 13, 5188-5192.
528 30. T. A. Halgren and W. N. Lipscomb, Chem Phys Lett, 1977, 49, 225-232.
529 31. N. Govind, M. Petersen, G. Fitzgerald, D. King-Smith and J. Andzelm, Comp
530 Mater Sci, 2003, 28, 250-258.
531 32. L. Lian, C. X. Su and P. B. Armentrout, J Chem Phys, 1992, 96, 7542-7554.
532 33. M. Sakurai, K. Watanabe, K. Sumiyama and K. Suzuki, J Chem Phys, 1999, 111,
533 235-238.
534 34. X. B. Han, Y. G. Li, Z. M. Zhang, H. Q. Tan, Y. Lu and E. B. Wang, J Am Chem
535 Soc, 2015, 137, 5486-5493.
536 35. M. Dunwell, Q. Lu, J. M. Heyes, J. Rosen, J. G. G. Chen, Y. S. Yan, F. Jiao and
537 B. J. Xu, J Am Chem Soc, 2017, 139, 3774-3783.
538 36. N. Sreekanth and K. L. Phani, Chem Commun, 2014, 50, 11143-11146.
539 37. G. D. Yao, F. Y. Chen, H. Zhang, R. T. He and F. M. Jin, ACS Sym Ser, 2015,
540 1194, 251-264.
541 38. A. A. Peterson and J. K. Norskov, J Phys Chem Lett, 2012, 3, 251-258.
22
View Article Online
DOI: 10.1039/C7CP06155B
542 39. Z. W. Ulissi, A. J. Medford, T. Bligaard and J. K. Norskov, Nat Commun, 2017,
543 8.

544 40. J. Y. Ye, C. J. Liu and Q. Ge, J Phys Chem C, 2012, 116, 7817-7825.
545 41. G. Canduela-Rodriguez, M. K. Sabbe, M. F. Reyniers, J. F. Joly and G. B. Marin,

546 Phys Chem Chem Phys, 2014, 16, 23754-23768.
547 42. G. W. Peng, S. J. Sibener, G. C. Schatz, S. T. Ceyer and M. Mavrikakis, J Phys
548 Chem C, 2012, 116, 3001-3006.
549 43. C.-S. He, L. Gong, J. Zhang, P.-P. He and Y. Mu, J CO2 Util, 2017, 19, 157-164.
550 44. F. Y. Chen, G. D. Yao, Z. B. Huo and F. M. Jin, RSC Adv, 2015, 5, 11257-11260.
551 45. G. Kovacs, G. Schubert, F. Joo and I. Papai, Catal Today, 2006, 115, 53-60.
552 46. S. Sarfraz, A. T. Garcia-Esparza, A. Jedidi, L. Cavallo and K. Takanabe, ACS
553 Catal, 2016, 6, 2842-2851.
554 47. H. A. Hansen, C. Shi, A. C. Lausche, A. A. Peterson and J. K. Norskov, Phys
555 Chem Chem Phys, 2016, 18, 9194-9201.
556 48. J. B. Varley, H. A. Hansen, N. L. Ammitzboll, L. C. Grabow, A. A. Peterson, J.
557 Rossmeisl and J. K. Norskov, ACS Catal, 2013, 3, 2640-2643.
558 49. A. A. Gokhale, J. A. Dumesic and M. Mavrikakis, J Am Chem Soc, 2008, 130,
559 1402-1414.
560 50. C. Heine, B. A. J. Lechner, H. Bluhm and M. Salmeron, J Am Chem Soc, 2016,
561 138, 13246-13252.
562 51. S. Bontemps, L. Vendier and S. Sabo-Etienne, J Am Chem Soc, 2014, 136,
563 4419-4425.
564 52. J. Shen, R. Kortlever, R. Kas, Y. Y. Birdja, O. Diaz-Morales, Y. Kwon, I.
565 Ledezma-Yanez, K. J. P. Schouten, G. Mul and M. T. M. Koper, Nat Commun,
566 2015, 6.
23
View Article Online
DOI: 10.1039/C7CP06155B
567 53. L. E. Heim, H. Konnerth and M. H. G. Prechtl, Green Chem, 2017, 19,
568 2347-2355.

569 54. K. Nakata, T. Ozaki, C. Terashima, A. Fujishima and Y. Einaga, Angew Chem
570 Int Edit, 2014, 53, 871-874.

571 55. S. Shahgaldi and J. Hamelin, Carbon, 2015, 94, 705-728.
572 56. Y. W. Cheng, H. B. Zhang, C. V. Varanasi and J. Liu, Energ Environ Sci, 2013, 6,
573 3314-3321.
574 57. J. Wang, X. B. Zhang, Z. L. Wang, L. M. Wang and Y. Zhang, Energ Environ Sci,
575 2012, 5, 6885-6888.
576
24
View Article Online
DOI: 10.1039/C7CP06155B
577 Table 1 Binding energy (Eads) of H2 dissociative adsorption on Ni@Ni11Fe and
578 Fe@Ni12 nanoclusters with different configurations in Fig. S1

Configuration Eads (eV)
H2-Ni@Ni11Fe-t1 -0.81
H2-Ni@Ni11Fe-t2 -0.30
H2-Ni@Ni11Fe-b -0.25
H2-Fe@Ni12-t1 -0.12
H2-Fe@Ni12-t2 -0.70
H2-Fe@Ni12-t3 -0.06
H2-Fe@Ni12-t4 -0.71
H2-Fe@Ni12-b1 -0.17
H2-Fe@Ni12-b2 -1.58
579
580
25
View Article Online
DOI: 10.1039/C7CP06155B
581 Table 2 Binding energy (Eads), and the distance of HCO3- adsorption on Ni@Ni11Fe
582 (dO1-Ni(Fe)) and Fe@Ni12 (dO2-Ni(Fe)) nanoclusters in Fig. S2

System Eads (eV) dO1-Ni(Fe) (Å) dO2-Ni(Fe) (Å)
a -2.12 2.01 1.96

b -1.28 2.22 2.07
c -2.96 2.15 2.07
d -1.55 2.02 2.19
e -3.62 1.96 1.98
f -3.22 1.97 1.92
583
584
26
View Article Online
DOI: 10.1039/C7CP06155B
585 Figure Legends
586

587
588 Fig. 1 Geometry optimized structures of Ni@Ni12Fe and Fe@Ni12 clusters (Purple:
589 Ni atom, and gold: Fe atom).
590
591 Fig. 2 Geometry optimized structures of H2 and HCO3- co-adsorption on Ni@Ni11Fe
592 and Fe@Ni12 nanoclusters (a for Ni@Ni11Fe nanocluster; and b, c, d for
593 Fe@Ni12 nanocluster; Purple: Ni atom, gold: Fe atoms, white: H atom, grey:
594 C atom, and red O atoms).
595
596 Fig. 3 Energy profile of HCO3- reduction to formic acid (HCOO-*) on Ni@Ni11Fe
597 nanocluster (The geometry optimized structures of reactants, intermediates,
598 and products are schematically illustrated in the respective insets, and TS
599 represents transition states).
600
601 Fig. 4 Energy profile of HCO3- reduction to formic acid (HCOO-*) on Fe@Ni12
602 nanocluster (The geometry optimized structures of reactants, intermediates,
603 and products are schematically illustrated in the respective insets, and *
604 denotes an adsorbed state).
605
606 Fig. 5 Molecular structures as well as calculated reaction energies and energy barriers
607 of HCO3- conversion to CO on Ni@Ni11Fe and Fe@Ni12 surface (units: eV).
608
27
View Article Online
DOI: 10.1039/C7CP06155B
609 Fig. 6 Optimized structures of reaction intermediates in HCOOH-* reduction to CHO-*
610 process on Ni13, Ni@Ni11Fe and Fe@Ni12 nanoclusters (a for adsorbed HCOO-,

611 b for adsorbed HCOOH-, and c for adsorbed CHO-; Purple: Ni atom, gold: Fe
612 atoms, white: H atom, grey: C atom, and red: O atom).

613
614 Fig. 7 Geometry optimized of graphene supported Ni@Ni11Fe and Fe@Ni12
615 nanoclusters (G for graphene; Purple: Ni atom, gold: Fe atoms, and grey: C
616 atom).
617
618 Fig. 8 Optimized structures of reaction intermediates in HCOOH- reduction to
619 CHO-process on graphene supported Ni@Ni11Fe and Fe@Ni12 nanoclusters (a
620 for HCOO-, b for HCOOH-, and c for CHO-; G for graphene; Purple: Ni atom,
621 gold: Fe atoms, white: H atom, grey: C atom, and red: O atom).
622
28
Page 29 of 36
626
625
624
623
29
Fig. 1
DOI: 10.1039/C7CP06155B
View Article Online

629
628
627
30
Fig. 2
DOI: 10.1039/C7CP06155B
View Article Online

Page 30 of 36
Published on 25 September 2017. Downloaded by LA TROBE UNIVERSITY on 01/10/2017
634
633
632
631
630
Page 31 of 36
31
Fig. 3

Published on 25 September 2017. Downloaded by LA TROBE UNIVERSITY on 01/10/2017
637
636
635
32
Fig. 4
Page 32 of 36

Page 33 of 36
641
640
639
638
33
Fig. 5
DOI: 10.1039/C7CP06155B
View Article Online

645
644
643
642
34
Fig. 6
DOI: 10.1039/C7CP06155B
View Article Online

Page 34 of 36
Page 35 of 36
649
648
647
646
35
Fig. 7
DOI: 10.1039/C7CP06155B
View Article Online

653
652
651
650
36
Fig. 8
DOI: 10.1039/C7CP06155B
View Article Online

Page 36 of 36

Catalytic CO Reduction To Valuable Chemicals Using NiFe Based Nanoclusters A First Principles Theoretical Evaluation

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Catalytic CO Reduction To Valuable Chemicals Using NiFe Based Nanoclusters A First Principles Theoretical Evaluation

Uploaded by

Copyright:

Available Formats

View Article Online

acceptance, before technical editing, formatting and proof reading.

You can find more information about Accepted Manuscripts in the

Please note that technical editing may introduce minor changes

1 Catalytic CO2 reduction to valuable chemicals using NiFe-based nanoclusters: A

2 first-principle theoretical evaluation

Physical Chemistry Chemical Physics Accepted Manuscript

4 Li Gonga, Jie-Jie Chena, Yang Mua,b*

7 Centre of Suzhou Nano Science and Technology, Department of Chemistry,

8 University of Science and Technology of China, Hefei, China

10 University of Science and Technology of China, Hefei, China

17 Prof. Yang Mu, Fax: +86 551 63607907, E-mail: yangmu@ustc.edu.cn

Physical Chemistry Chemical Physics Accepted Manuscript

23 in CO2 conversion. Herein we theoretically evaluated the catalytic performance and

24 possible mechanisms of NiFe-based nanoclusters for hydrogenating CO2 to form

25 formic acid and CO through bicarbonate by using a periodic and self-consistent

33 HCO3- reduction was shown to be neither thermodynamically nor kinetically

34 favorable on NiFe alloy nanoclusters. Additionally, the simulation results also

35 suggested that it was thermodynamically unfavorable for a further hydrogenated

36 reduction of formic acid to formaldehyde on NiFe alloy nanoclusters themselves as

37 well as supported on graphene. In summary, a molecular-level insight of CO2

40 materials for efficient CO2 conversion to useful fuels.

Physical Chemistry Chemical Physics Accepted Manuscript

47 energy and thus decrease the energy shortage.1, 2

49 including photocatalytic, electrocatalytic and photo-electrochemical combined

50 ones.3-5 Among these approaches, catalytic reduction of CO2 by direct hydrogenation

51 with generated or in-site hydrogen is currently considered to be the most promising

52 feasible synthetic route in terms of economy and ecology.

53 Tremendous research efforts have been devoted to developing non-precious

55 source in water.6-9 In particular, earth-abundant 3d transition metals like Fe, Co and

56 Ni have been considered as promising catalytic materials for replacement of precious

58 earth-abundant first row transitional metal with corrosion-resistance and ductility,

59 and it has been found that Ni nanoparticles could be a potential heterogeneous

60 catalyst for selective CO2 reduction to formic acid through bicarbonate.10,11

61 Various studies have been performed to improve the catalytic performance of

62 Ni-based catalysts, such as through decorating Ni with CrOx,12 supporting Ni on

63 carbon materials,13 and alloying Ni with foreign metals.14,15 Bimetallic alloy

64 catalysts are reported to generally exhibit enhanced catalytic performances over

67 chemical properties of the meatals.16,17 Nanosized bimetallic catalysts, especially

Physical Chemistry Chemical Physics Accepted Manuscript

70 interdisciplinary chemical processes such as hydrogenation/dehydrogenation,

71 fuel-cell electro catalysis ammonia decomposition, and so on.18-21 Recently

78 4-chlorophenol dechlorination in water.24 Since NiFe bimetal alloy has been

82 on the noble-metal-free NiFe alloy nanoparticles.

83 In this study we theoretically evaluated the catalytic performance of the

84 NiFe-based nanoclusters for CO2 reduction to valuable products through bicarbonate

87 firstly investigated. Following, both thermodynamic and kinetic properties as well as

88 possible reaction mechanisms of CO2 reduction to formic acid and CO through

90 feasibility of further reducing formic acid on various NiFe nanoclusters as well as

91 supported on graphene was studied by using DFT simulations. In this way, a

92 molecular-level insight into catalytic conversion of CO2 to valuable products through

93 bicarbonate on NiFe-based nanoclusters would be offered, which may provide useful

Physical Chemistry Chemical Physics Accepted Manuscript

97 Computational methods and models

100 First-principles DFT calculations were performed in the CASTEP module of

103 gradient approximation (GGA) Perdew-Burke-Ernzerhof (PBE) level.26 The ultrasoft

104 pseudopotential was constructed using the pseudopotential generator implemented in

111 generated by the Monkhorst-pack algorithm.29 Spin-polarized calculations were used

114 phonon density of states. Phonon frequencies calculated by Finite displacement

118 without symmetry constraints. The linear synchronous transit (LST)/quadratic

Physical Chemistry Chemical Physics Accepted Manuscript

120 structure and calculate the energy barrier (Ea).30,31

122 adsorption system, was calculated to estimate the strength of a molecule-surface