Catalytic oxidation of CO by N₂O on neutral Y₂MO₅ (M = Y, Al) clusters: a density functional theory study

Abstract

Density functional theory (DFT) calculations are employed to investigate the full catalytic cycle of CO oxidation by N₂O on yttrium oxide clusters Y₂MO₅ (M = Y, Al) in the gas-phase. Extensive structural searches show that both the ground-state structures of Y₃O₅ and Y₂AlO₅ contain an oxygen radical (O_t˙) which plays an important role in CO oxidation. Energy profiles are calculated to determine the reaction mechanisms. Molecular electrostatic potential maps (MEPs) and natural bond orbital (NBO) analyses are employed to rationalize the reaction mechanisms. The results indicate that the whole catalytic cycle for the reaction CO + N₂O → CO₂ + N₂, conducted by yttrium oxide clusters Y₂MO₅ (M = Y, Al), is favored both thermodynamically and kinetically. Moreover, compared with the previous report on di-nuclear YAlO₃⁺˙ and Y₂O₃⁺˙, it's obvious we can conclude that tri-nuclear Y₃O₅ and Y₂AlO₅ exhibit greatly enhanced catalytic activity toward CO/N₂O couples.

---

1. Introduction

Catalytic oxidation of harmful gases, such as carbon monoxide (CO) into carbon dioxide (CO₂) and nitrous oxide (N₂O) into nitrogen (N₂), is of great importance both environmentally and economically.^1–10 Even though oxidation of CO by N₂O is exothermic, it does not occur directly at any measurable temperatures because of high energy barrier for the CO/N₂O couple.¹¹ It has been verified that some transition metal oxides are effective in catalytic oxidation of CO by N₂O.^12–18 Mechanistic investigation on the oxygen atom transfer process mediated by transition metal oxides which are involved in CO oxidation by N₂O is meaningful to design better catalysts with high activity and selectivity. Gas-phase cluster study has been emerged as an alternative approach to provide molecular-level insights into the complex chemical reaction processes occurring on the bulk transition metal oxide.^19–28

The catalytic oxidation of CO by N₂O at room temperature (RT) in the gas phase was demonstrated for the first time by Kappes and Staley with FeO⁺ as a catalyst.¹² In the ensuing decades, numerous relevant investigations were developed subsequently.^13–16 For example, Castleman and co-workers have reported that a series of zirconium oxide (ZrO₂)_x⁺ (x = 2–5) cations as well as (Zr_xO_2x+1)⁻ (x = 1–4) anions exhibit high activity in CO oxidation and the oxygen radical center with elongated zirconium–oxygen bond is the active site.^13–15 They also have shown that the initial zirconium oxide ions can be regenerated by treating oxygen-deficient zirconium-oxide clusters with N₂O. Besides, binary neutral metal oxide clusters ZrScO₄ and ZrNbO₅ were demonstrated to have similar reactivity with their isoelectronic ions Zr₂O₄⁺ and Zr₂O₅⁻ for CO oxidation and N₂O reduction.¹⁶ Some other metal-oxide clusters, such as Ce_mO_2m⁺ (m = 2–6), Ce_nO_2n+1⁻ (n = 4–21), (TiO₂)_nO⁻ (n = 3–25) and VO₃, were also exploited to bring out their oxidizing capacity towards CO.^29–32 These investigations have indicated the oxygen radical center plays a critical role in CO oxidation process.

Yttrium oxides are also promising for the catalytic application, yet the related studies on CO oxidation by yttrium oxides are relatively scarce.^11,33 Many early studies have focused their attentions on the electronic structure and bonding of yttrium oxide clusters for their unique properties.^34–38 The previous study has pointed out that the YAlO₃⁺˙ cluster contains oxygen radical center and shows better reactivity than homonuclear yttrium oxide cluster Y₂O₃⁺˙ in CO oxidation.¹¹ Recently, the terminal oxygen radical was also found in the neutral Y₃O₅ cluster.³⁸

In this work, we present our theoretical study on CO oxidations by neutral Y₂MO₅ (M = Y, Al) clusters employing density functional calculations. Extensive structural searches show that both Y₃O₅ and Y₂AlO₅ clusters are doublet state with a single unpaired electron located on the terminal oxygen atom. The energy profiles are calculated to get the detail mechanisms underlying the oxygen atom transfer processes involved in CO oxidation reaction. Qualitative explanation for the reactions of Y₂MO₅ (M = Y, Al) clusters with CO can be understood from the molecular electrostatic potential maps and natural bond orbital analyses. The regenerations of Y₂MO₅ (M = Y, Al) clusters via the reactions of oxygen-deficient clusters Y₂MO₄ (M = Y, Al) with N₂O are also calculated. A full catalytic cycle for the reaction of CO with N₂O on Y₂MO₅ (M = Y, Al) clusters is thus complete.

2. Computational methods

The theoretical calculations were performed at the DFT level using the BP86 functional.^39,40 The calculations were performed using analytical gradients with the Stuttgart relativistic small core basis set and efficient core potential^41,42 augmented with two f-type and one g-type polarization functions for yttrium [ζ(f) = 0.144, 0.546; ζ(g) = 0.249] as recommended by Martin and Sundermann⁴³ and the aug-cc-pvTZ basis set for oxygen, alumina, carbon and nitrogen.^44,45 Scalar relativistic effects, that is, the mass velocity and Darwin effects, were taken into account via the quasi-relativistic pseudopotentials. All the structures presented in the calculated energy profiles were fully optimized. The initial structures of the transition state were obtained by potential energy surface scans with appropriate coordinates. Vibrational frequency calculations were performed at the same level of theory to verify the nature of the stationary points as minima (zero imaginary frequency) or transition states (one imaginary frequency). The pathways of the reaction mechanisms were further confirmed by intrinsic reaction coordinate (IRC) calculations.^46,47

The relative energies of the low-lying Y₂MO_x (M = Y, Al; x = 4–5) structures were further evaluated via single-point calculations at the coupled cluster [CCSD(T)]^48–52 level with the Y/Stuttgart+2f1g/O/Al/aug-cc-pVTZ basis sets at the BP86 geometries. The results are summarized and available in ESI (Table S1†). In our current studies, the BP86 gave superior results in terms of energies when directly compared to the results of CCSD(T) calculations. The previous studies on yttrium oxide clusters^38,53 also have indicated that the BP86 calculations can give good consistency with the experimental results. As discussed below, we used the results with the BP86 functional for further discussion. All DFT calculations were performed with the Gaussian 09 software package⁵⁴ and the CCSD(T) calculations were done using MOLPRO 2010.1 package.⁵⁵ Natural bond orbital (NBO) analysis was performed by using the NBO 3.1 program.⁵⁶ Three-dimensional molecular structures were visualized using the GaussView 4.1.⁵⁷ Three-dimensional contours of the molecular orbitals were visualized using the VMD software.⁵⁸

3. Results and discussions

3.1 The structures and spin density analyses of Y₂MO_x (M = Y, Al; x = 4–5)

The optimized ground state structures and low-lying isomers within 0.40 eV for the neutral Y₃O_x and Y₂AlO_x (x = 4–5) clusters at the BP86 level are presented in Fig. 1 and 2. Alternative optimized geometries are shown in the ESI (Fig. S1 and 2†). In the following discussions, O_t, O_b and O_c stand for the terminal, bridging and capped oxygen atoms, respectively. The spin density analyses are performed to shed light on the nature of the radical character of these neutral ground states, as shown in Fig. 3.

Fig. 1 Optimized global minima and selected low-lying structures (within ∼0.40 eV) for Y₃O_x (x = 4–5) clusters at the BP86 level. The bond lengths are in angstroms.

Fig. 2 Optimized global minima and selected low-lying structures (within ∼0.40 eV) for Y₂AlO_x (x = 4–5) clusters at the BP86 level. The bond lengths are in angstroms.

Fig. 3 Numerical electron spin density (in |e|) for the ground states of Y₂MO_x (M = Y, Al; x = 4–5) clusters.

3.1.1 Y₃O₄ and Y₃O₅. In the previous work, we have reported the structures of Y₃O₄ and Y₃O₅ clusters.³⁸ The low-lying isomers within 0.40 eV for Y₃O₄ and Y₃O₅ are given in Fig. 1. The ground state of Y₃O₄ is a doublet state C_3v (²A₁) structure (Fig. 1a), which can be regarded as an umbellate structure with a capped oxygen atom and three bridging oxygen atoms. In our calculation, the bond lengths of Y–O_c and Y–O_b are 2.174 Å and 2.048 Å, respectively. Alternative optimized structures are located remarkably higher in energy (Fig. S1a–c†).

The ground state of Y₃O₅ has a propensity to adopt a doublet (²A′′) state with C_s symmetry (Fig. 1b) at the BP86 level. The Y–O_t bond length is 2.155 Å and can be considered as Y–O single bond. Two kinds of connection between caped oxygen and yttrium atoms are found in this structure, where the Y–O_c distances are presented to be 2.134 Å and 2.309 Å, respectively. Two (²A′) state structures showing in Fig. 1c and d are calculated to be 0.11 eV and 0.21 eV above the ground state, respectively. We also calculated the higher symmetry structure (D_3h) with two triply-bridging O atoms and three doubly-bridging O atoms for Y₃O₅, which converged to the C_s structure, as shown in Fig. 1d. The relative energies for these low-lying structures (within ∼0.20 eV) were further evaluated using single point CCSD(T) calculations. The results of CCSD(T) calculations are in line with those of DFT calculations.

3.1.2 Y₂AlO₄ and Y₂AlO₅. We carried out extensive structural searches for the ground-state structures on the potential energy surfaces of Y₂AlO₄ and Y₂AlO₅ clusters considering a variety of structural candidates including different spin states and geometries. The global minimum of Y₂AlO₄ is predicted to be a doublet state (²A′) with C_s symmetry (Fig. 2a), which shows resemblance with Y₃O₄ (C_3v, ²A₁) in geometry (Fig. 1a). The low-lying isomer C_2v (²A₁) (Fig. 2b) with four bridging O atoms is only 0.02 eV above the global minimum at BP86 level. Single-point CCSD(T) calculations were performed for these two isomers, revealing that the (C_s, ²A′) structure (Fig. 2a) is 0.12 eV more stable than the (C_2v, ²A₁) structure (Fig. 2b).

As for Y₂AlO₅, selected optimized structures with their relative energies are given in Fig. 2c–f. The ground state of Y₂AlO₅ is a doublet state C_s (²A′′) as shown in Fig. 2c. It can be viewed that the Al atom replaces the yttrium atom which is link with the terminal oxygen atom in Y₃O₅. The Al–O_t distance is 1.767 Å and can be labeled as Al–O single bond.⁵⁹ The (²A′) state with similar geometry is 0.12 eV above the ground state at the BP86 level. The C_2v (²B₂) (Fig. 2d) and C_2v (²B₁) (Fig. 2f), which own a terminal and four bridging oxygen atoms, are 0.08 eV and 0.31 eV higher in energy, respectively. These low-lying structures within 0.20 eV were further evaluated by CCSD(T) calculations and the isomer (C_s, ²A′′) (Fig. 2c) is still the lowest-energy structure.

3.1.3 Spin density analysis. All the ground states of Y₂MO_x (M = Y, Al; x = 4–5) clusters are doublet state. As mentioned above, the global minimum of Y₃O₄ is a C_3v structure with a capped and three bridging oxygen atoms. The spin is distributed equally on three yttrium atoms, as shown in Fig. 3a. After doping an aluminum atom, the unpaired electron is mainly located on aluminum atom in Y₂AlO₄ (Fig. 3b). Both Y₃O₅ and Y₂AlO₅ clusters contain an oxygen radical (O˙) as depicted in Fig. 3c and d, respectively. The singly occupied molecular orbitals (SOMOs) of Y₃O₅ and Y₂AlO₅ (Fig. 4) indicate that the unpaired electron is located on the 2p orbitals of terminal oxygen atom.

Fig. 4 The singly occupied molecular orbitals for the ground states of Y₂MO₅ (M = Y, Al) clusters.

To our knowledge, the oxygen-centre radical is highly reactive. It's able to oxidize a variety of stable molecules including CO and many small organic species at low temperature, such as CH₄, C₂H₄, C₂H₂ and so forth.^14,60–63 He and co-workers have studied the reactions of CO with Y₂O₃⁺˙ and YAlO₃⁺˙ by mass spectrometry and density functional theoretical.¹¹ According to their results, oxidation of the CO by Y₂O₃⁺˙ is prevented by an energy barrier, while oxidation of the CO with YAlO₃⁺˙ is favored. The differences are mainly produced by the spin distributions. In Y₂O₃⁺˙, the unpaired electron is delocalized over three bridging oxygen atoms. As for YAlO₃⁺˙, the spin is mainly located on the terminal oxygen atom of Al–O_t˙ unit. However, the single unpaired electrons of Y₃O₅ and Y₂AlO₅ clusters both located on the terminal oxygen atom of M–O_t˙ (M = Y, Al) subunit. In the next section, the mechanisms for CO oxidation by Y₃O₅ and Y₂AlO₅ are presented.

3.2 The reactions of Y₂MO₅ (M = Y, Al) with CO

The detail mechanisms for the reactions of Y₂MO₅ (M = Y, Al) with CO are investigated at the BP86 level. In our calculation, all the intermediates and transition structures in doublet state are much more stable than those in quartet state. So DFT calculations are performed for the molecule reactions of Y₂MO₅ (M = Y, Al) towards CO on the doublet ground state potential energy surfaces. The products are proposed according to the following equations:

Y₂MO₅ (M = Y, Al) + CO → Y₂MO₄ (M = Y, Al) + CO₂

We have considered the possible paths for the reactions and found the way in which CO reacts with the M–O_t˙ site is more preferred. Then we will focus on the mechanism for CO oxidation involving the oxygen radical center. Fig. 5 and 6 present the favored energy profiles calculated at the BP86 level in which IMn represent the intermediate structures and TSn correspond to the transition states, respectively. The detail of electron transfer procedures can be obtained by analyzing the spin density of the intermediates and transition states in the reactions as shown in Fig. S8 and 9.† The Cartesian coordinates for the optimized structures on the potential energy surfaces are given in Table S3.† Other possible reaction paths were also calculated and given in Fig. S3–S7.†

Fig. 5 DFT calculated energy profile for the reaction of Y₃O₅ cluster with CO₂. The reactants, intermediates, transition states and produces of the reaction are denoted as R, IMn, TSn and P, respectively. The energies are given in eV and bond lengths are in angstroms.

Fig. 6 DFT calculated energy profile for the reaction of Y₂AlO₅ cluster with CO. The reactants, intermediates, transition states and produces of the reaction are denoted as R, IMn, TSn and P, respectively. The energies are given in eV and bond lengths are in angstroms.

3.2.1 The reaction profiles of Y₂MO₅ (M = Y, Al) with CO.
3.2.1.1 Y₃O₅ with CO. The calculated energy profile for the reaction between Y₃O₅ and CO is presented in Fig. 5. The oxidation reaction occurs through the exothermic (0.29 eV) binding of the carbon atom of CO to the radical oxygen and the adjacent yttrium atom of Y₃O₅ forming an intermediate (IM1). In IM1, the distances of carbon atom to yttrium atom and the radical oxygen are 2.699 Å and 2.749 Å, respectively. Formation of the IM2 requires a transfer of spin from the terminal oxygen to the nonlinear CO₂ unit. This involves an easily surmountable barrier of 0.02 eV. In the process, the C–O_t distance decreases from 2.749 Å to 1.275 Å and Y–O_t length increases by 0.138 Å. From IM2, the nonlinear CO₂ subunit rotates around the Y–O_t bond involving a transition state (TS2) which is 0.10 eV higher in energy. The resulting complex IM3 is lower in energy by 2.64 eV than the reactants. There is another transition state (TS2′) from IM2 to IM3, which needs to overcome 0.54 eV of barrier, as shown in Fig. S3.† The unpaired electron transfers from the curving bent CO₂ motif back to the yttrium oxide cluster needs to overcome a barrier of 0.61 eV through the transition state (TS3). In this process, the distance of the terminal oxygen atom to the adjacent yttrium atom is elongated, that is, 2.325 Å for IM3, 2.510 Å for TS3 and 2.589 Å for IM4, respectively. The IM4 is 2.05 eV lower in energy than the reactants with a linear CO₂. Finally, dissociation of CO₂ from IM4 to form the products Y₃O₄ and CO₂ requires 0.25 eV of energy. The overall oxidation reaction is barrierless and calculated to be exothermic by 1.80 eV, indicating the oxidation reaction is favorable both kinetically and thermodynamically.
3.2.1.2 Y₂AlO₅ with CO. The calculated energy profile of Fig. 6 shows that the reaction of Y₂AlO₅ with CO proceeds according to a general mechanism involving the initial binding of the carbon atom of CO to the radical oxygen of Y₂AlO₅. The initial encounter complex IM1 is 2.11 eV more stable than the reactants and contains a bent CO₂ subunit that results from the transfer of single unpaired electron to CO. The bond lengths of C–O_t and Al–O_t are 1.314 Å and 1.800 Å, respectively. In the following steps, the CO₂ moiety is transferred from the Al site to the Y site of Y₂AlO₄ subunit, involving the transitions which are 0.54 eV and 0.53 eV higher in energy for TS1 and TS2, respectively. IM2 and IM3 are 2.34 eV and 2.20 eV lower in energy than the separated reactants, respectively. Spin density is appeared to be mainly located on the Y₂AlO₄ subunit for the IM2 and Al atom for IM3, respectively (Fig. S9†). With the increase of the distance from the carbon atom to the bridge oxygen atom and the decrease of the distance from the carbon atom to the capped oxygen atom, the CO₂ subunit is passed to the Y–O_c site forming IM4, requiring an energy barrier of 0.53 eV through the TS3. The CO₂-absorbing (CO₂ ready-desorbing) IM4 is 1.70 eV lower in energy than the reactants. Besides, the electron located on CO₂ unit of IM1 can directly transfer to Al atom forming the IM4 with a barrier of 1.28 eV, as depicted in Fig. S5.† Finally, loss of the CO₂ from IM4 requires 0.31 eV of energy, and the overall process is exothermic by 1.39 eV, resulting in Y₂AlO₄ and CO₂ products.

3.2.2 The molecular electrostatic potential and natural bond orbital analyses. An additional supporting finding for the reactions is acquired by considering the molecular electrostatic potential maps (MEPs) and natural bond orbital (NBO) analyses. As can be seen in Fig. S10† for Y₃O₅ cluster, the MEP exhibits positive values (the blue areas) which are preferred for CO-adsorption. In our calculations, the way that CO adsorbs on the yttrium atom featured by terminal oxygen atom is preferred both kinetically and thermodynamically. In contrast, CO binding to the yttrium atoms which connect with the capped oxygen and bridging oxygen atoms is impeded by an energy barrier, as shown in Fig. S4.† Just as many researchers have reported, metal atom adjacent to the oxygen-centered radical is more reactive toward CO.^16,29 The MEP of Y₂AlO₅ is positive over two yttrium atoms surface. However, the way that CO adsorbs on the yttrium atoms of the Y₂AlO₅ cluster is not favored as depicted in Fig. S7.† In Y₂AlO₅, the oxygen radical center area is relatively negative. Attacking at the oxygen radical center leads to the oxidation reaction, which is a preferable path without any energy barrier as shown previously.^14,16,17

The MEP of CO₂ molecular, shown in Fig. S10,† is positive over carbon atom and negative over oxygen atoms, respectively. In Y₃O₄, the yttrium atoms are the most positive (1.74|e|) sites and the capped oxygen is more negative (−1.34|e|) than bridging oxygen sites. Therefore, the linear CO₂ has a tendency to split from the Y–O_c bond. Analogously, the yttrium site (1.91|e|) is more positive than aluminum site (1.47|e|) in Y₂AlO₄, which may explain why the CO₂ molecular dissociates from the Y–O_c bond rather than Al–O_c bond.

3.3 The reaction of Y₂MO₄ (M = Y, Al) with N₂O

The Y₂MO₅ (M = Y, Al) clusters can be retrieved by treating Y₂MO₄ (M = Y, Al) clusters with N₂O. Then a full cycle is attainable in the reactions of CO with N₂O by neutral Y₂MO₅ (M = Y, Al) clusters (Fig. 7). DFT calculations at the BP86 level are performed for the reactions of Y₂MO₄ (M = Y, Al) with N₂O. The favorite paths for Y₃O₄ and Y₂AlO₄ towards N₂O are presented in Fig. 8 and 9, respectively. The others listed in the ESI (Fig. S11 and 12†).

Fig. 7 Proposed full catalytic cycle reaction involving CO and N₂O mediated by Y₂MO₅ (M = Y, Al) clusters.

Fig. 8 DFT calculated for the reaction of N₂O with Y₃O₄ cluster. The reactants, intermediates, transition states and produces of the reaction are denoted as R, IMn, TSn and P, respectively. The energies are given in eV and bond lengths are in angstroms.

Fig. 9 DFT calculated for the reaction of N₂O with Y₂AlO₄ cluster. The reactants, intermediates and produces of the reaction are denoted as R, IM1 and P, respectively. The energies are given in eV and bond lengths are in angstroms.

3.3.1 Y₃O₄ with N₂O. The calculated energy profile for the reaction between Y₃O₄ and N₂O is presented in Fig. 8. For Y₃O₄, the spin is delocalized equally over the three yttrium atoms. The adsorption of N₂O on the equivalent yttrium atom leads to the intermediate (IM1), which is 1.35 eV more stable than the reactants. With the increase of the distance between O atom and N atom (1.399 to 2.789 Å) and the decrease of the distance from O atom to Y atom (2.275 to 2.172 Å), the O atom is passed on smoothly to bind with Y atom forming IM2. The activation barrier of the oxygen transfer is calculated to be 0.05 eV, which corresponds primarily to the breaking of the bond of N–O. The IM2 is 1.68 eV lower in energy than the reactants. Dissociation of the N₂ unit generates the products Y₃O₅ and N₂, which requires 0.01 eV of energy. The total oxidation reaction is exothermic by 1.67 eV.

3.3.2 Y₂AlO₄ with N₂O. The possible paths for the Y₂AlO₄ cluster reducing N₂O are displayed in Fig. 9, S11 and 12.† The current study shows that the unpaired electron in Y₂AlO₄ cluster mainly located on the main-group metal aluminum atom. As the N₂O molecule attacking the aluminum atom, the unpaired electron will occupy the LUMO anti-bonding orbital of N₂O and the N–O distance will be lengthened to 2.824 Å. The interaction of the Al atom and N₂ unit is weak as their distance is 2.643 Å. The complex IM1 is 1.99 eV more stable than the separated reactants. Releasing the N₂ from IM1 is rather easy without any barrier. The overall process is barrier-less and calculated to be exothermic by 2.09 eV.

The different reactivity of Y₃O₄ versus Y₂AlO₄ in their reactions with N₂O is just another example of the role spin distributions often played in chemical reactions. In the doped Y₂AlO₄, the unpaired electron is mainly localized on the Al atom, while in the homonuclear Y₃O₄ cluster, the spin delocalized equally over the three bridging Y atoms. Consequently, the reaction patterns of Y₃O₄ and Y₂AlO₄ towards N₂O are different. Furthermore, the different spin distributions also induce a more negative change of Gibbs free energy for Y₂AlO₄ (−2.14 eV) in reacting with N₂O than that of Y₃O₄ (−1.79 eV). In other words, while doping a main group Al atom the reaction is favorable both thermodynamically and kinetically.

3.4 Comparison with di-nuclear yttrium oxide clusters

As mentioned above, the reactions of CO with Y₂MO₅ (M = Y, Al) are barrierless and exothermic, and the other half of the catalytic cycles for regenerations of Y₂MO₅ (M = Y, Al) by N₂O are favored both thermodynamically and kinetically. Thus the full catalytic cycle is attainable in this work. The catalytic ability of di-nuclear yttrium oxides clusters Y₂O₃⁺˙ and YAlO₃⁺˙ for CO/N₂O couple also have been studied by He and co-workers.¹¹ The oxidation of CO with Y₂O₃⁺˙ is impeded by an energy barrier. As one yttrium atom of Y₂O₃⁺˙ was replaced by one aluminum atom, the oxygen radical center was generated leading to the reaction barrier-less. Regeneration of YAlO₃⁺˙ from YAlO₂⁺˙ with N₂O dose not take place and the catalytic cycle cannot be closed. Combined our calculations with He et al. report, it's obvious to conclude that the tri-nuclear yttrium oxides clusters exhibit greatly enhanced catalytic activity in the reaction of CO with N₂O.

4. Conclusion

Extensive DFT calculations are employed to investigate the reaction mechanism of CO oxidation by Y₂MO₅ (M = Y, Al) clusters and the regeneration of Y₂MO₅ (M = Y, Al) clusters using N₂O. Theoretical investigations reveal that both the Y₃O₅ and Y₂AlO₅ clusters are doublet state with a single unpaired electron located on the terminal oxygen atom. Calculated energy profiles for the reactions of Y₃O₅ and Y₂AlO₅ with CO are energetically favorable and barrierless. The high reactivity for Y₂MO₅ clusters is ascribed to the oxygen radical center. The reaction mechanisms for Y₂MO₅ (M = Y, Al) clusters with CO may be understood qualitatively from the molecular electrostatic potential maps and natural bond orbital analyses. Regenerations of the initial species Y₂MO₅ (M = Y, Al) by N₂O constitute the other half of the catalytic cycle. Based on the previous study of di-nuclear YMO₂⁺˙ (M = Y, Al), we propose that by controlling the cluster size (or doping ratio), it's possible to generate neutral oxide cluster with localized radical oxygen center, which is highly active in redox reactions.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (21371034, 21373048 and 21301030), the Natural Science Foundation of Fujian Province for Distinguished Young Investigator Grant (2013J06004), and the Independent Research Project of State Key Laboratory of Photocatalysis on Energy and Environment (2014A02).

References

1. A. R. Ravishankara, J. S. Daniel and R. W. Portmann, Science, 2009, 326, 123.
2. B. K. Min and C. M. Friend, Chem. Rev., 2007, 107, 2709.
3. S. Royer and D. Duprez, ChemCatChem, 2011, 3, 24.
4. S. J. Lin, J. Cheng, C. F. Zhang, B. Wang, Y. F. Zhang and X. Huang, Phys. Chem. Chem. Phys., 2015, 17, 11499.
5. I. Lopes, A. Davidson and C. Thomas, Catal. Commun., 2007, 8, 2105.
6. G. W. Peng, L. R. Merte, J. Knudsen, R. T. Vang, E. Lagsgaard, F. Besenbacher and M. Marvrikakis, J. Phys. Chem. C, 2010, 114, 21579.
7. X. W. Xie, Y. Li, Z. Q. Liu, M. Haruta and W. J. Shen, Nature, 2009, 458, 746.
8. H. J. Freund, G. Meijer, M. Scheffler, R. Schlogl and M. Qolf, Angew. Chem., Int. Ed., 2011, 50, 10064.
9. W. E. Kaden, W. A. Kunkel, F. S. Roberts, M. Kane and S. L. Andersona, J. Chem. Phys., 2012, 136, 204705.
10. O. P. Balaj, I. Balteanu, T. T. J. Roβteuscher, M. K. Beyer and V. E. Bondybey, Angew. Chem., Int. Ed., 2004, 43, 6519.
11. J. B. Ma, Z. C. Wang, M. Schlangen, S. G. He and H. Schwarz, Angew. Chem., Int. Ed., 2013, 52, 1226.
12. M. M. Kappes and R. H. Staley, J. Am. Chem. Soc., 1981, 103, 1286.
13. G. E. Johnson, R. Mitrić, E. C. Tyo, V. Bonačić-Koutecký and A. W. Castleman Jr, J. Am. Chem. Soc., 2008, 130, 13912.
14. G. E. Johnson, R. Mitrić, M. Nössler, E. C. Tyo, V. Bonačić-Koutecký and A. W. Castleman Jr, J. Am. Chem. Soc., 2009, 131, 5460.
15. G. E. Johnson, R. Mitrić, V. Bonačić-Koutecký and A. W. Castleman Jr, Chem. Phys. Lett., 2009, 475, 1.
16. M. Nößler, R. Mitrić, V. Bonačić-Koutecký, G. E. Johnson, E. C. Tyo and A. W. Castleman Jr, Angew. Chem., Int. Ed., 2010, 49, 407.
17. Z. C. Wang, N. Dietl, R. Kretschmer, T. Weiske, M. Schlangen and H. Schwarz, Angew. Chem., Int. Ed., 2011, 50, 12351.
18. E. C. Tyo, M. Nößler, R. Mitrić, V. Bonačić-Koutecký and A. W. Castleman Jr, Phys. Chem. Chem. Phys., 2011, 13, 4243.
19. A. W. Castleman Jr, Catal. Lett., 2011, 141, 1243.
20. S. Yin and E. R. Bernstein, Int. J. Mass Spectrom., 2012, 321–322, 49.
21. S. M. Lang and T. M. Bernhardt, Phys. Chem. Chem. Phys., 2012, 14, 9255.
22. K. R. Asmis, Phys. Chem. Chem. Phys., 2012, 14, 9270.
23. H. Schwarz, Angew. Chem., Int. Ed., 2011, 50, 10096.
24. M. Schlangen and H. Schwarz, Catal. Lett., 2012, 142, 1265.
25. Z. C. Wang, N. Dietl, R. Kretschmer, J. B. Ma, T. Weiske, M. Schlangen and H. Schwarz, Angew. Chem., Int. Ed., 2012, 51, 3703.
26. Z. C. Wang, X. N. Wu, Y. X. Zhao, J. B. Ma, X. L. Ding and S. G. He, Chem.–Eur. J., 2011, 17, 3449.
27. Z. C. Wang, T. Weiske, R. Kretschmer, M. Schlangen, M. Kaupp and H. Schwarz, J. Am. Chem. Soc., 2011, 133, 16930.
28. Z. C. Wang, X. N. Wu, Y. X. Zhao, J. B. Ma, X. L. Ding and S. G. He, Chem. Phys. Lett., 2010, 489, 25.
29. X. N. Wu, Y. X. Zhao, W. Xue, Z. C. Wang, S. G. He and X. L. Ding, Phys. Chem. Chem. Phys., 2010, 12, 3984.
30. X. N. Wu, X. L. Ding, S. M. Bai, B. Xu, S. G. He and Q. Shi, J. Phys. Chem. C, 2011, 115, 13329.
31. J. B. Ma, B. Xu, J. H. Meng, X. N. Wu, X. L. Ding, X. N. Li and S. G. He, J. Am. Chem. Soc., 2013, 135, 2991.
32. Z. C. Wang, W. Xue, Y. P. Ma, X. L. Ding, S. G. He, F. Dong, S. Heinbuch, J. J. Rocca and E. R. Bernstein, J. Phys. Chem. A, 2008, 112, 5984.
33. M. R. Sievers and P. B. Armentrout, Inorg. Chem., 1999, 38, 397.
34. M. Knickelbein, Chem. Phys., 1995, 102, 1.
35. A. Pramann, Y. Nakamura and A. Nakajima, J. Phys. Chem. A, 2001, 105, 7534.
36. G. Gu, B. Dai, X. Ding and J. Yang, Eur. Phys. J. D, 2004, 29, 27.
37. B. Dai, K. Deng and J. Yang, Chem. Phys. Lett., 2002, 364, 188.
38. L. Xu, C. J. Xia, L. F. Wang, L. Xie, B. Wang, Y. F. Zhang and X. Huang, RSC Adv., 2014, 4, 60270.
39. A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098.
40. J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys., 1986, 33, 8822.
41. W. Küchle, M. Dolg, H. Stoll and H. Preuss, Pseudopotentials of the Stuttgart/Dresden group 1998, revision August 11, 1998, http://www.theochem.uni-stuttgart.de/pseudopotentiale.
42. D. Andrae, U. Häeußermann, M. Dolg, H. Stoll and H. Preuss, Theor. Chim. Acta, 1990, 77, 123.
43. J. M. L. Martin and A. Sundermann, J. Chem. Phys., 2001, 114, 3408.
44. R. A. Kendall, T. H. Dunning Jr and R. J. Harrison, J. Chem. Phys., 1992, 96, 6796.
45. T. H. Dunning Jr, J. Chem. Phys., 1989, 90, 1007.
46. C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1989, 90, 2154.
47. C. Gonzalez and H. B. Schlegel, J. Phys. Chem., 1990, 94, 5523.
48. G. D. Purvis III and R. J. Bartlett, J. Chem. Phys., 1982, 76, 1910.
49. G. E. Scuseria, C. L. Janssen and H. F. Schaefer III, J. Chem. Phys., 1988, 89, 7382.
50. K. Raghavachari, G. W. Trucks, J. A. Pople and M. Head-Gordon, Chem. Phys. Lett., 1989, 157, 479.
51. J. D. Watts, J. Gauss and R. J. Bartlett, J. Chem. Phys., 1993, 98, 8718.
52. R. J. Bartlett and M. Musial, Rev. Mod. Phys., 2007, 79, 291.
53. Y. Gong, C. F. Ding and M. F. Zhou, J. Phys. Chem. A, 2009, 113, 8569.
54. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylow, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian 09, Revision A.1, Gaussian, Inc., Wallingford CT, 2009.
55. H. J. Werner, P. J. Knowles, F. R. Manby, M. Schütz, P. Celani, G. Knizia, T. Korona, R. Lindh, A. Mitrushenkov, G. Rauhut, T. B. Adler, R. D. Amos, A. Bernhardsson, A. Berning, D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn, F. Eckert, E. Goll, C. HamPel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen, C. Köppl, Y. Liu, A. W. Lloyd, R. A. Mata, A. J. May, S. J. McNicholas, W. Meyer, M. E. Mura, A. Nicklass, D. P. O'Neill, P. Palmieri, K. Pflüger, R. Pitzer, M. Reiher, T. Shiozaki, H. Stoll, A. J. Stone, R. Tarroni, T. Thorsteinsson, M. Wang and A. Wolf, MOLPRO, version 2010.1, a package of ab initio programs, http://www.molpro.net.
56. E. D. Glendening, A. E. Reed, J. E. Carpenter and F. Weinhold, NBO Version 3.1.
57. R. Dennington II, T. Keith, J. Millam, K. Eppinnett, W. L. Hovell and R. Gilliland, GaussView: Version 4.1, Semichem Inc., Shawnee Mission, KS, 2007.
58. W. Humphrey, A. Dalke and K. Schulten, J. Mol. Graphics, 1996, 14, 33.
59. A. Martínez, F. J. TenorioJ and V. Ortiz, J. Phys. Chem. A, 2001, 105, 11291.
60. Y. X. Zhao, X. N. Wu, J. B. Ma, S. G. He and X. L. Ding, Phys. Chem. Chem. Phys., 2011, 13, 1925.
61. J. B. Ma, Z. C. Wang, M. Schlangen, S. G. He and H. Schwarz, Angew. Chem., Int. Ed., 2012, 51, 5991.
62. N. Dietl, M. Schlangen and H. Schwarz, Angew. Chem., Int. Ed., 2012, 51, 5544.
63. X. L. Ding, X. N. Ding, Y. X. Zhao and S. G. He, Acc. Chem. Res., 2012, 45, 382.

---

Footnote

† Electronic supplementary information (ESI) available: The others possible pathways for the redox reaction are presented in Fig. S3–S7, S11 and 12. Numerical electron spin density (in |e|) for the structures presented in the energy profiles of Fig. 5 and 6 are shown in Fig. S8 and 9. The Cartesian coordinates for the reactants, intermediates, transition states and the products shown in the energy profiles of Fig. 5, 6, 8 and 9 are listed in Table S3. See DOI: 10.1039/c5ra11695c

---