A Density Functional Study for the Reaction Mechanism of CO Oxidation on the Copper Cluster①

ZHOU Sheng-Hu YU Wei-Ling ZHANG Jing LI Yi ZHANG Yong-Fn CHEN Wen-Ki, b, c

?

A Density Functional Study for the Reaction Mechanism of CO Oxidation on the Copper Cluster①

ZHOU Sheng-HuaaYU Wei-LingaZHANG JingaLI YiaZHANG Yong-FanaCHEN Wen-Kaia, b, c②

a(350116)b(350116)c(()361005)

We have studied the reaction mechanism of CO oxidation on the Cu13cluster via density functional theory. There are two main reaction pathways to be considered: Eley-Rideal (ER) and Langmuir-Hinshelwood (LH) mechanisms, respectively. According to these two main reaction mechanisms, we have obtained five reaction pathways for the first CO oxidation (denoted as RER1,RER2,RLH1, RLH2and RLH3, respectively):RER1is CO(gas)+ O2 (ads)→O(ads)+ CO2(gas); RER2is CO(gas)+ O2(ads)→CO3(ads)→ O(ads)+ CO2(gas); RLH1refers to CO(ads)+ O2(ads)→ O(ads)+ CO2(gas); RLH2refers to CO(ads)+ O2(ads)→ OOCO(ads)→ O(ads)+ CO2(gas)and RLH3refers toO2 (ads)+ CO(ads)→ O(ads)+O(ads)+ CO(ads)→ O(ads)+ CO2(gas). These pathways have low energy barriers and are strongly exothermic, suggesting the Cu13cluster is very favorable catalyst for the first CO oxidation. However, there are higher energy barriers of 99. 8 and 45.4 kJ/mol in the process of producing and decomposing intermediates along the RLH2and RER2, indicating that RER1, RLH1and RLH3are superior pathways with lower energy barriers, especially the RER1channel. Thereafter, the second CO is more prone to react with the remaining oxygen atom on Cu13along the ER channel in comparison with the LH pathway, in which the moderate barrier is 70.0 kJ/mol and it is exothermic by 59.6 kJ/mol. Furthermore, the interaction between the absorbate and cluster is analyzed by electronic analysis to gain insights into high activity of the copper cluster.

reaction mechanism, CO oxidation, copper cluster, catalyst, electronic analysis;

1 INTRODUCTION

It has been a subject of extensive studies for CO oxidation, which has attracted particular attention in many applications, including the abatement of CO produced by the vehicle, gas purification for improvement efficiency of CO2lasers, full cell, developing CO gas sensors and removal for trace CO in a certain confined space[1-12]. And also the CO oxidation on the metal and their oxide surfaces have been fundamentally investigated and considered as an ideal reaction in heterogeneous catalysis[13-16]. It has been very effective for catalytic oxidation of CO on the precious metals like Pd, Pt, Au and so on with high activity and stability at low temperature for a long time[17-21]. Due to the limited availability, high cost of noble metals and sensitivity to sulfur poisoning, considerable attention has been paid to new catalysts and their oxides[22]. Among cheap transition metals, the copper is often viewed as a remarkable base metal catalyst for CO oxidation[23, 24]. Therefore, there are a few applications for catalytic oxidation of carbon monoxide over the copper catalysts, which have been viewed as a fundamental process in automotive exhaust controls, the operation of flue cells, the water-gas shift reaction, and reforming of alcohols[25-31]. In particular, the superior catalytic activities of copper and copper-based catalysts have obtained much attention for the CO oxidation in regular and hydrogen-rich (PROX) streams. And also there are changes thermodyna- mically for oxidation states of copper among three copper species, namely Cu, Cu2O and CuO[32]. It is found that there are inconsistent conclusions from studies on catalytic activities of copper with different oxidation states for CO oxidation[25, 33]. Jernigan and Somorjai obtained the kinetic result of CO oxidation on a thin film of every oxidation state of copper, which increases in the order of CuOis oxygen defect phase and metastable state, which can influence the study for catalytic activity of CuObecause of apparent structural sensitivity, and illustrated the rate of CO catalytic oxidation is strong on the basis of initial oxidation state under mild conditions, but the initial phase differences have a few effects on the rate at high temperature for a long time.

Moreover, it is known that copper metal bulk materials exhibit chemical inertness to be used as currency, but corresponding metal clusters present chemical active. The clusters have many coordinated surface sites and a large surface-to-volume ratio in comparison with the corresponding bulk materials. Interaction among geometric, elastic and electronic parameters contributes to unique properties of clusters[37, 38]. There are intense investigations for heterogeneous catalysis on the clusters surface currently[39, 40]. The number of atoms in the clusters significantly affects the metal clusters properties. Adding or removing a simple atom into or from the clusters may dramatically change the clusters properties. Meanwhile, there exist many isomers of a metal cluster. And the properties of metal clusters may be also strongly sensitive to their structures. Recently, the considerable information of transition metal clusters on structure spectroscopy and reac- tivity is available with the development of theoretical and experimental techniques. The photoelectron spectroscopy (PES) is often applied to the investiga- tions of copper anion clusters[41-44]. Except the PES studies, the theoretical methods including all kinds of density functional[45-49]are employed in the investi- gations for energy, binding energy, electron affinity, ground state structure and so on. The structures for all kinds of transition metal clusters containing 13 atoms have been studied by many theories and experiments to know they might be completely icosahedral and very stable[50-52].

However, it is still unclear for the reaction mechanism of CO oxidation on copper cluster. Kim and co-workers followed LH mechanism to present the high catalytic activity of Ag13cluster for CO oxidation[53]. Meanwhile, silver is copper-group metal so that its properties are similar to the pro- perties of copper. However, it is found that copper is much more suitable for the oxygen adsorption than silver, which can obviously influence the reaction pathway of CO oxidation on the Cu13cluster. Knowing the mechanism of CO oxidation on copper clusters is meaningful in designing suitable cheap metal catalysts. We have obtained two main reaction pathways of CO oxidation on the copper cluster: the ER and LH pathways, respectively. However, there are two possible processes to be denoted as RER1andRER2for the first CO oxidation as a function of the ER pathway and also three possible processes to be indicated as RLH1, RLH2, and RLH3for the LH mechanism.

Herein, we present a systematic research on the reaction mechanism of CO oxidation on the copper cluster, which is the aim of work. All calculated results involved in reaction process are obtained and compared with each other, which can provide a better understanding on CO oxidation on copper cluster and present whether the Cu13cluster can be used as better catalysts for CO oxidation. Furthermore, we have illustrated how the intermediates are formed during the reaction.

2 COMPUTATIONAL METHOD

To investigate the reaction mechanism of CO oxidation on Cu13cluster, the program package of Materials Studio of Accelrys Inc has been performed to optimize the geometries and search the transition states[54, 55]. Generalized gradient approximation (GGA) with exchange-correlation function proposed by Perdew and Wang scheme[56]was performed. In calculation, the effective core potential was employ- ed for inner electrons of copper atom, but the C and O atoms were treated with all-electron basis set. Meanwhile, the double-numerical basis with polari- zation functions (DNP) has been also employed in the computational calculation. The smearing method for a window size with 0.005 Hartree was employed. The energy, maximum force and maximum displace- ment for convergence tolerance were 10-5Hartree, 0.002 Hartree/? and 0.005 ?, respectively. The transition states in the processes of CO oxidation on clusters were determined by the complete LST/QST method, which means linear synchronous transition and quadratic synchronous transition, respectively. Every atom in the cluster was relaxed so that the cluster and absorbates were fully optimized geome- trically[57]. Meanwhile, the adsorption energy is defined as follows:

ads=system– (adsorbate+cluster) (1)

where thesystemis the energy of the adsorption system,absorbatemeans the energy of the absorbate alone, andclusterstands for the ground state energy of the bare cluster.

3 RESULTS AND DISCUSSION

3.1 Investigation for absorbates on the surface of cluster

The configurations for CO and O2adsorption on the cluster are shown in Fig. 1, respectively. It is found that the adsorption of O2parallel on the cluster is the most stable. And also O2can obtain much more electrons from the cluster than CO on the cluster, especially the O2parallel on the copper cluster. Meanwhile, it is noted that the bond distances of free CO and O2molecules are 1.141 and 1.224 ?, respectively. The O–O bonds for O2parallel on the cluster are much more obviously elongated, as illustrated in Fig. 1. The fact implies that CO mole- cule is more likely to attack O2parallel on the cluster. However, there are two adsorption configurations for O2parallel on the cluster, denoted as b1 and b2 in Fig. 1. The O2adsorption in b1 is a bit more stable than that in b2 as a function of the comparison between their adsorption energies. To obtain an optimized O2adsorption system, the density of states (DOS) on O2and Cu before and after adsorption have been calculated and given in Fig. 2. The 3orbital of Cu plays a major role for properties of Cu13cluster so that we can obtain the density of states of Cu 3orbital. Both density of states for the configurations in Fig. 1(b) are similar to each other, which is in accordance with their stabilities. However, it is found that the bond distance of O2on the cluster is more obviously elongated for b2 adsorption system. Meanwhile, the Mulliken charge demonstrated in Table 1 can clearly suggest that two oxygen atoms of the O2molecule in Fig. 1 (b2) obtain much more electric charges from the cluster to activate O2. Therefore, the adsorption system in Fig. 1(b2) is the most optimized. Note that the adsorption energies of CO and O2on the Ag13cluster are –84.91 and –62.72 kJ/mol via DFT-based calculation[53], which are far weaker than that on the THE Cu13cluster. And also the coadsorption energy of CO and O2on the Ag13cluster is –147.62 kJ/mol, which is very close to the value for an elementary step for CO oxidation on the Ag13along the LH mechanism. However, it is much weaker than the coadsorption of CO and O2on Cu13cluster with the adsorption energy of –353.99 kJ/mol, which is in good agreement with the results of –304.79 kJ/mol[58]. That influences the catalytic activity of cluster. And it is known that gold clusters possess high activities for CO oxidation along the LH pathway[59]. Meanwhile, the coadsorption of CO and O2on the gold clusters is an elementary step. Instead, it is very weak for O2adsorption on the gold clusters, which can be enhanced via CO preadsorp- tion on the gold clusters[59]. And also we have obtained the adsorption energy for CO on the Cu13cluster to be –165.87 kJ/mol, which is more stable in comparison with CO on the Au16cluster with the adsorption energy of –104.20 kJ/mol[60]. Moreover, the coadsorption for CO and O2on the Cu13cluster is also much more stable than on many gold clusters.

Fig. 1. Optimized geometries of CO and O2adsorption on the surface of Cu13cluster. The bond lengths are given in angstrom (?)

Fig. 2. Density of states (DOS) for O2molecule andorbital of Cu before and after adsorption, in which the adsorption systems are in accordance with Fig. 1(b). The dashed line represents the Fermi level

Table 1. Adsorption Energy and Mulliken Charge of CO (O2) Molecule on the Cluster as a Function of Adsorption Systems on Fig. 1, respectively

To understand the mechanism of CO oxidation on the cluster, the intermediates formed by CO and O2are also studied. We have obtained two stable configurations of intermediates: carbonate-like (CO3) and peroxo-type (OOCO), as shown in Fig. 3. It is found that the O–O bond is increased from 1.224 to 1.498 ? or even broken for intermediates, which can promote the forming of new C–O bond to obtain the carbonate-like and peroxo-type, respectively. And also the adsorption energies of intermediates are –621.93 and –323.72 kJ/mol, indicating the stabilities of the carbonate-like and peroxo-type on the copper cluster. Furthermore, there exist electrostatic interaction between carbon with positive electric charges and oxygen atom with negative electric charges for these two inter- mediates, which can enhance their stabilities as a function of recent publication[60-64]. Therefore, we have obtained that the carbonate-like (CO3) on the cluster may be more stable.

Fig. 3. Optimized stable intermediates: (a) carbonated-like (CO3), (b) peroxo-type (OOCO). The bond lengths in angstrom are indicated on the adsorption systems shown in the first row, and also Mulliken charges of two intermediates are shown in the second row

Fig. 4. Density of states for the Cu, O2and CO before and after adsorption, in which the adsorption systems are as a function of the corresponding configurations in Fig. 3. The dashed line represents the Fermi level

To make clear the interactions between carbon monoxide and oxygen molecules in carbonate-like (CO3) and peroxo-type (OOCO) on the Cu13cluster, the density of states (DOS) on absorbates and copper are obtained, as indicated in Fig. 4. For the system of CO3on the copper cluster, it is found that the energies for 5, 1, 2and 6orbitals of O2molecule are lower after adsorption, and also the 3orbital of copper is also delocalization toward lower energy region and overlaps with O22πorbital which is also more obvious delocalization, as demonstrated in Fig. 4 (a3). That indicates O2molecule obtains electrons form 3orbital of copper to make the O–O bond rupture and lower the system energy. Meanwhile, the 2πorbitals of CO and O2are overlapped with each other to form the new C–O bond between CO and O2. For system of OOCO on the copper cluster, 2πorbitals of the CO and O2can obviously interact with Cu 3orbital, and significantly overlap with each other, as indicated in Fig. 4 (b3) and 4 (b4), which suggest the new C–O bond between CO and O2is obtained. Furthermore, the delocalization range for 2πorbital of O2and CO after adsorption shown in Fig. 4 (a3) is wider toward the lower energy region than in Fig. 4 (b3) and (b4), indicating the system of carbonate-like (CO3) on the cluster is more stable.

3.2 Reaction profiles of carbon monoxide catalytic oxidation on the Cu13 cluster

The reaction mechanisms of carbon monoxide oxidation on the Cu13cluster have been studied via DFT-based calculation. There are five possible reac- tion pathways for the first CO oxidation on the copper cluster, shown in Figs. 5 to 9.

The first pathway denoted as RER1, as shown in Fig. 5, indicates the CO directly attacks the O2absorption on the Cu13cluster to obtain CO2leaving from the cluster. Note that the reaction along this pathway is highly exothermic, and its energy barrier is only 19.91 kJ/mol, which suggests this reaction pathway is prone to proceed. Meanwhile, the O–O bond is breaking and also the O–C bond is forming along the reaction channel, which promotes CO oxidation.

Fig. 5. Energy profile for the first CO oxidation on the Cu13cluster as a function of the RER1pathway, in which bond lengths are given in angstrom

Fig. 6 indicates the second pathway, denoted as RER2. The O2absorption on the surface is firstly attacked by the gaseous CO to obtain an intermediate shown in Fig. 3 (a). This intermediate decomposes to obtain CO2molecule leaving from the surface and oxygen atom on the cluster. In this process, the gaseous CO breaks the O–O bond via transition state 1 with an energy barrier of 19.23 kJ/mol, and forms chemical bonds with two oxygen atoms to obtain a very stable intermediate with three-membered ring, which is exothermic by the 398.51 kJ/mol. However, the decomposition of the intermediate on the surface needs to surmount a higher energy barrier of 45.40 kJ/mol, in which there exist O–Cu and O–C bonds scission and it is endothermic by 24.74 kJ/mol. Therefore the formation of CO3is much easier than dissociation, which may make the second CO oxi- dation be less possible along this pathway in comparison with RER1channel. Instead, the potential energy of transition state 2 is much smaller than the energy of the reactant demonstrated in Fig. 6, indi- cating this pathway is also suitable.

Fig. 6. Energy profile for the CO oxidation on the Cu13cluster as a function of the RER2channel, in which bond lengths are given in angstrom

Fig. 7. Energy profile for the CO oxidation on the Cu13 cluster as a function of the path RLH1, in which bond lengths are given in angstrom

Fig. 7 illustrates the third possible pathway, denoted as RLH1, which presents that both CO and O2absorb on the surface to generate gaseous CO2and oxygen atom on the cluster. Note that the bond distance of O–O is elongated and the Cu–O bond length is shorter on the reactant, which suggest there is stronger interaction between the oxygen and copper atoms. And also the process is exothermic by 237.72 kJ/mol along CO(ads)+ O2 (ads)→O(ads)+ CO2(gas)

with a low energy barrier of 22.69 kJ/mol. Therefore, it is found this third reaction pathway is more prone to proceed than the second one, but it may be a bit less suitable than the first channel because of the weaker exothermicity. All in all, this pathway is also favorable.

Fig. 8. Energy profile for the CO oxidation on the Cu13cluster as a function of the RLH2, in which the bond lengths are given in angstrom

The forth pathway begins with the coadsorption mode of CO and O2molecules on copper cluster along the RLH2channel, as shown in Fig. 8. They surmount the relative large energy barrier to obtain the intermediate b demonstrated in Fig. 3(b) via transition state 1, in which there is a high energy barrier of 99.78 kJ/mol and it is endothermic by 30.29 kJ/mol. Furthermore the rupture of O–O bond in OOCO generates the CO2molecule leaving from the cluster with an energy barrier of 16.82 kJ/mol, indicating intermediate b is easier to decompose. Therefore, the reaction along RLH2is feasible but less possible than other pathways mentioned above, especially the RER1and RLH1.

Fig. 9. Energy profile for the CO oxidation on the Cu13cluster as a function of RLH3, in which the bond lengths are given in angstrom

There is also a reaction pathway to be denoted as RLH3. The absorbed O2molecule is firstly dissociated, which surmounts the energy barrier of 22.26 kJ/mol and is exothermic by 172.68 kJ/mol. And then gaseous CO reacts with one of the oxygen atoms on the cluster to obtain CO2leaving form the surface, in which there is an energy barrier of 9.27 kJ/mol and it is exothermic by 74.86 kJ/mol. The RLH3pathway has low energy barrier, which is similar to the RER1and RLH1pathways. However, it is weakly exother- mic and has a bit more complex reaction process in comparison with the RER1and RLH1channels, which may influence activity of the cluster. As such, this pathway is very prone to proceed.

There are also two reaction pathways for the second CO molecule oxidation to be considered, denoted as ER and LH mechanisms. The gaseous CO reacts with the remaining oxygen atom on the cluster to generate the CO2leaving from the surface, and both of the CO molecule and oxygen atom on the cluster react with each other to obtain the gaseous CO2, respectively. Therefore, we have firstly ob- tained Fig. 10. It is found that there is energy barrier of 70.00 kJ/mol and it is exothermic by 59.57 kJ/mol, indicating the second CO oxidation is a rate-limiting step in a complete catalytic cycle. However, the reaction along LH, as illustrated in Fig. 11, needs to surmount the energy barrier of 87.09 kJ/mol and is endothermic by 63.35 kJ/mol. Thereby, it is found that the second CO oxidation is more possible to be along the ER channel via comparison. Although there are higher energy barrier and lower exothermicity in the process of the second CO oxidation, which is ratelimiting step in this reaction, the high exo- thermicity and low energy barrier in the process of the first CO oxidation can promote the second CO oxidation.

Fig. 10. Energy profile for the oxidation of the second CO on the cluster along the ER channel, in which bond lengths are given in angstrom

Fig. 11. Energy profile for the oxidation of the second CO on the cluster along the LH channel, in which bond lengths are given in angstrom

Table 2. Energy Barrier and the Reaction Energy for CO Oxidation on the Cu13 Cluster along Different Mechanisms, in Which the Eb1 and Eb2 Denote the Energy Barrier of the Formation and the Decomposition for Intermediate in the Corresponding Processes, and also theΔEr1 and ΔEr2 Indicate the Reaction Energy of Formation and the Decomposition of Intermediate

4 CONCLUSION

In summary, we have obtained that five reaction pathways can be suitable for the first CO oxidation on the Cu13cluster. However, it is found that the carbonate-like (CO3) is very stable on the Cu13cluster surface, which can hinder the CO3decompo- sition to make RER2channel less possible. Meanwhile, there is a relative high energy barrier for transition state 1 in the RLH2channel, which weakens the activity of cluster. Therefore, RER1, RLH1and RLH3are more possible channels in the reaction. In particular, the RER1has a simpler pathway, lower energy barrier and higher exothermicity in comparison with other pathways for the first CO oxidation. And also ER pathway is more prone to proceed for the second CO oxidation than LH. Our calculation indicates that Cu13cluster exhibits a large catalytic activity for CO oxidization by the O2molecule, which can be concluded by its specific geometry and extraor- dinarily electronic properties between the cluster and absorbates. Furthermore, it is found that the structure of the Cu13cluster has relaxation in the reaction, which may influence the catalytic activity of substrate for CO oxidation. Thus, the core atom in the Cu13cluster is replaced by the other atom to obtain bimetallic cluster,which can lower the repulsive forces in the cluster to make structure more stable and protect the configuration from the relaxation in the reaction as a function of the previous study. We will search a better heterogeneous core atom or also consider the supported metal cluster to improve the catalytic activity of cluster for CO oxidation, enhance the stability of the configura- tion and also expect that the present results can provide information for future research.

(1) Haruta, M.Size-and support-dependency in the catalysis of gold1997,36,153-166.

(2) Okumura, M.;Tsubota, S.;Haruta, M.Vital role of moisture in the catalytic activity of supported gold nanoparticles2004,43,2129-2132.

(3) Bond, G.C.;Thompson, D.T.Catalysis by gold1999,41,319-388.

(4) Sanchez, A.;Abbet, S.;Heiz, U.;Schneider, W.D.;H?kkinen, H.;Barnett, R.;Landman, U.When gold is not noble: nanoscale gold catalysts1999,103,9573-9578.

(5) Kimble, M.L.;Castleman, A.W.;Mitric, R.;Bürgel, C.;Bonac?ic-Koutecky, V.Reactivity of atomic gold anions toward oxygen and the oxidation of CO: Experiment and theory2004,126,2526-2535.

(6) Kim, H.Y.;Lee, H.M.;Pala, R.G.S.;Shapovalov, V.;Metiu, H.CO oxidation by rutile TiO2(110) doped with V, W, Cr, Mo, and Mn2008,112,12398-12408.

(7) Wang, H. F.;Gong, X. Q.;Guo, Y. L.;Guo, Y.;Lu, G.;Hu, P.Structure and catalytic activity of gold in low-temperature CO oxidation2009,113,6124-6131.

(8) Wang, F.;Zhang, D.;Xu, X.;Ding, Y.Theoretical study of the CO oxidation mediated by Au3+, Au3, and Au3?: Mechanism and charge state effect of gold on its catalytic activity2009,113, 18032-18039.

(9) Wan, H.;Li, D.;Dai, Y.;Hu, Y.;Zhang, Y.;Liu, L.;Zhao, B.;Liu, B.;Sun, K.;Dong, L.Effect of CO pretreatment on the performance of CuO/CeO2/-Al2O3catalysts in CO+O2reactions2009,360,26-32.

(10) Rozovskii, A.Y.;Kipnis, M.;Volnina, E.;Samokhin, P.;Lin, G.;Kukina, M.Laws of selective CO oxidation over a Ru/Al2O3catalyst in the surface ignition regime: ii. Transition states2009,50,691-704.

(11) Pala, R.G.S.;Tang, W.;Sushchikh, M.M.;Park, J. N.;Forman, A.J.;Wu, G.;Kleiman-Shwarsctein, A.;Zhang, J.;McFarland, E.W.;Metiu, H.Co oxidation by Ti-and Al-doped ZnO: Oxygen activation by adsorption on the dopant2009,266,50-58.

(12) Farkas, A.;Mellau, G.C.;Over, H.Novel insight in the CO oxidation on RuO2(110) by in situ reflection? absorption infrared spectroscopy2009,113,14341-14355.

(13) McClure, S.M.;Goodman, D.W.New insights into catalytic CO oxidation on pt-group metals at elevated pressures2009,469,1-13.

(14) Xu, J.;White, T.;Li, P.;He, C.;Yu, J.;Yuan, W.;Han, Y. F.Biphasic Pd?Au alloy catalyst for low-temperature CO oxidation2010,132,10398-10406.

(15) Suchorski, Y.;Wrobel, R.;Becker, S.;Weiss, H.CO oxidation on a CeO/Pt(111) inverse model catalyst surface: catalytic promotion and tuning of kinetic phase diagrams2008,112,20012-20017.

(16) Eck, S.;Castellarin-Cudia, C.;Surnev, S.;Prince, K.;Ramsey, M.;Netzer, F.Adsorption and reaction of CO on a ceria–Rh(111)“inverse model catalyst” surface2003,536,166-176.

(17) Zhu, H.;Qin, Z.;Shan, W.;Shen, W.;Wang, J.Pd/CeO2–TiO2catalyst for CO oxidation at low temperature: a tpr study with H2and CO as reducing agents2004,225,267-277.

(18) Budroni, G.;Corma, A.Gold-organic-inorganic highsurfacearea materials as precursors of highly active catalysts2006,45,3328-3331.

(19) Mari?o, F.;Descorme, C.;Duprez, D.Noble metal catalysts for the preferential oxidation of carbon monoxide in the presence of hydrogen (prox)2004,54,59-66.

(20) Fu, Q.;Weber, A.;Flytzani-Stephanopoulos, M.Nanostructured Au–CeO2catalysts for low-temperature water-gas shift2001,77,87-95.

(21) Zeng, Q.S.;Sun, B.Z.;Zhao, W.N.;Lin, H.X.;Li, Y.;Chen, W.K.Adsorption of Co2B2and Ni2B2clusters on the TiO2(110) surface: a density functional study2013,32,341-348.

(22) Luo, M. F.;Song, Y. P.;Lu, J. Q.;Wang, X. Y.;Pu, Z. Y.Identification of CuO species in high surface area CuO-CeO2catalysts and their catalytic activities for co oxidation2007,111,12686-12692.

(23) Wang, J.B.;Tsai, D. H.;Huang, T. J.Synergistic catalysis of carbon monoxide oxidation over copper oxide supported on samaria-doped ceria2002,208,370-380.

(24) Martínez-Arias, A.;Hungría, A.;Fernández-García, M.;Conesa, J.;Munuera, G.Interfacial redox processes under CO/O2in a nanoceria-supported copper oxide catalyst2004,108,17983-17991.

(25) Jernigan, G.;Somorjai, G.Carbon monoxide oxidation over three different oxidation states of copper: metallic copper, copper(I) oxide, and copper(II) oxide-a surface science and kinetic study1994,147,567-577.

(26) Gamarra, D.;Belver, C.;Fernández-García, M.;Martínez-Arias, A.Selective CO oxidation in excess H2over copper-ceria catalysts: identification of active entities/species2007,129,12064-12065.

(27) Szanyi, J.;Goodman, D.W.CO oxidation on a Cu(100) catalyst1993,21,165-174.

(28) Domagala, M.E.;Campbell, C.T.The mechanism of CO oxidation over Cu(110): effect of CO gas energy1991,9,65-70.

(29) Newsome, D.S.The water-gas shift reaction1980,21,275-318.

(30) Rao, K.N.;Bharali, P.;Thrimurthulu, G.;Reddy, B.M.Supported copper-ceria catalysts for low temperature CO oxidation2010,11,863-866.

(31) Li, Y.;Xing, Y. Y.;Liu, R. R.;Hu, J. M.;Zhang, Y. F.;Chen, W. K.Density functional theory study of the adsorption of C2H2on the Cu/Pt(111) bimetallic surfaces2013,32,1211-1221.

(32) O'Keeffe, M.;Moore, W.J.Thermodynamics of the formation and migration of defects in cuprous oxide2004,36,3009-3013.

(33) Nagase, K.;Zheng, Y.;Kodama, Y.;Kakuta, J.Dynamic study of the oxidation state of copper in the course of carbon monoxide oxidation over powdered CuO and Cu2O1999,187,123-130.

(34) Huang, T. J.;Tsai, D. H.CO oxidation behavior of copper and copper oxides2003,87,173-178.

(35) Sadykov, V.;Tikhov, S.Comment on “carbon monoxide oxidation over three different oxidation states of copper: metallic copper, copper(I) oxide, and copper(II) oxide—a surface science and kinetic study” by gg jernigan and ga somorjai1997,165,279-283.

(36) Kryukova, G.;Zaikovskii, V.;Sadykov, V.;Tikhov, S.;Popovskii, V.;Bulgakov, N.Study of the nature of extended defects of copper oxide1988,74,191-199.

(37) White, B.;Yin, M.;Hall, A.;Le, D.;Stolbov, S.;Rahman, T.;Turro, N.;O'Brien, S.Complete Co oxidation over Cu2O nanoparticles supported on silica gel2006,6,2095-2098.

(38) Jia, T. T.;Sun, B. Z.;Lin, H. X.;Li, Y.;Chen, W. K.Bonding of hydroxyl and epoxy groups on graphene: Insights from density functional calculations. 2013,32,1475-1484.

(39) Haruta, M.Catalysis: gold rush2005,437,1098-1099.

(40) Hutchings, G.J.;Haruta, M.A golden age of catalysis: a perspective2005,291,2-5.

(41) Cha, C.Y.;Gantef?r, G.;Eberhardt, W.Photoelectron spectroscopy of Cunclusters: comparison with jellium model predictions1993,99,6308-6312.

(42) Taylor, K.;Pettiette-Hall, C.;Cheshnovsky, O.;Smalley, R.Ultraviolet photoelectron spectra of coinage metal clusters1992,96,3319-3329.

(43) Ho, J.;Ervin, K.M.;Lineberger, W.Photoelectron spectroscopy of metal cluster anions: Cun, Agn, and Aun1990,93,6987-7002.

(44) Cheshnovsky, O.;Taylor, K.;Conceicao, J.;Smalley, R.Ultraviolet photoelectron spectra of mass-selected copper clusters: evolution of the 3band1990,64,1785-1788.

(45) Fernández, E.M.;Soler, J.M.;Garzón, I.L.;Balbás, L.C.Trends in the structure and bonding of noble metal clusters2004,70,165403-7.

(46) Jaque, P.;Toro-Labbé, A.Characterization of copper clusters through the use of density functional theory reactivity descriptors2002,117,3208-3218.

(47) Jug, K.;Zimmermann, B.;Calaminici, P.;K?ster, A.M.Structure and stability of small copper clusters2002,116,4497-4507.

(48) Calaminici, E.P.;K?ster, A.;Russo, N.;Salahub, D.A density functional study of small copper clusters: Cun(n < 5)1996,105,9546-9556.

(49) Jackson, K.A.First-principles study of the structural and electronic properties of Cu clusters1993,47,9715-9722.

(50) Okamoto, Y.Density-functional calculations of icosahedral M13(M = Pt and Au) clusters on graphene sheets and flakes2006,420,382-386.

(51) Wang, L. L.;Johnson, D.Density functional study of structural trends for late-transition-metal 13-atom clusters2007,75,235405-5.

(52) Zhang, M.;Fournier, R.Density-functional-theory study of 13-atom metal clusters M13, M =Ta–Pt2009,79,043203-6.

(53) Kim, H.Y.;Kim, D.H.;Ryu, J.H.;Lee, H.M.Design of robust and reactive nanoparticles with atomic precision: 13Ag-Ih and 12AgX (X = Pd, Pt, Au, Ni, or Cu) core-shell nanoparticles2009,113,15559-15564.

(54) Delley, B.From molecules to solids with the dmol3 approach2000,113,7756-7764.

(55) Delley, B.An all-electron numerical method for solving the local density functional for polyatomic molecules1990,92,508-517.

(56) Perdew, J.P.;Wang, Y.Accurate and simple analytic representation of the electron-gas correlation energy1992,45,13244-13249.

(57) Liu, X.;Yang, Z.;Zhang, R.;Li, Q.;Li, Y.Density functional theory study of mechanism of N2O decomposition over Cu-ZSM-5 zeolites2012,116,20262-20268.

(58) Hirabayashi, S.;Ichihashi, M.;Kawazoe, Y.;Kondow, T.Comparison of adsorption probabilities of O2and CO on copper cluster cations and anions2012,116,8799-8806.

(59) Gao, Y.;Shao, N.;Pei, Y.;Chen, Z.;Zeng, X.C.Catalytic activities of subnanometer gold clusters (Au16–Au18, Au20, and Au27–Au35) for CO oxidation2011,5,7818-7829.

(60) Chen, H.T.;Chang, J.G.;Ju, S.P.;Chen, H.L.First-principle calculations on CO oxidation catalyzed by a gold nanoparticle2010,31,258-265.

(61) Haruta, M.;Tsubota, S.;Kobayashi, T.;Kageyama, H.;Genet, M.J.;Delmon, B.Low-temperature oxidation of CO over gold supported on TiO2,-Fe2O3, and Co3O41993,144,175-192.

(62) Su, H. Y.;Yang, M. M.;Bao, X. H.;Li, W. X.The effect of water on the CO oxidation on Ag(111) and Au(111) surfaces: a first-principle study2008,112,17303-17310.

(63) Prestianni, A.;Martorana, A.;Ciofini, I.;Labat, F.;Adamo, C.CO oxidation on cationic gold clusters: a theoretical study2008,112,18061-18066.

(64) Faji?n, J.L.;Cordeiro, M.N.D.;Gomes, J.R.DFT study of the CO oxidation on the Au(321) surface2008,112,17291-17302.

19 December 2017;

12 April 2018

① This project was supported by the National Natural Science Foundation of China (Nos. 51574090 and 21773030) and Natural Science Foundation of Fujian Province (2017J01409)

. Professor, majoring in computational chemistry. E-mail: wkchen@fzu.edu.cn

10.14102/j.cnki.0254-5861.2011-1927