Enhanced Photoassisted Li‐O2 Battery with Ce‐UiO‐66 Metal‐Organic Framework Based Photocathodes

Li‐O2 batteries have attracted extensive attention because of their theoretical specific energy equivalent to gasoline, but the poor electrical conductivity of Li2O2 leads to high overpotential, which limits their further development and application. Herein, this work introduces Ce‐UiO‐66, a metal‐organic framework material, as the photocatalyst to reduce the overpotential in the discharging and charging processes. With the simple in situ growth of Ce‐UiO‐66 on carbon cloth as an integrated photocathode, the photoassisted Li‐O2 batteries display decent discharge and charge voltages of 3.1 and 3.6 V, and a lifespan of 160 cycles. Moreover, theoretical calculations have been carried out to understand the band structure, spectroscopy, and photocatalytic property of Ce‐UiO‐66. The findings will encourage an enormous variety of novel MOFs‐based photocathodes for solar energy utilization systems.


Enhanced Photoassisted Li-O 2 Battery with Ce-UiO-66 Metal-Organic Framework Based Photocathodes
product Li 2 O 2 causes oxygen reduction reaction (ORR) and oxygen evolution reaction (OER) process slow, resulting in high overpotential and low round-trip efficiency. [8] Moreover, the high overpotential also leads to the side reaction between the electrode and electrolyte, leading to formation of several byproducts such as lithium carbonate. [9] In order to reduce the overpotential of Li-O 2 batteries, researchers have introduced various types of electrocatalysts, such as noble metals (Pt, [10] Pd [11] ), metal oxides (MnO 2 [12] ), metal sulfides (MoS 2 [13] ), and functional carbon materials. [14] However, the stability and selectivity of those catalysts are far from satisfactory, and the high cost of noble metals limits their application in nonaqueous  batteries. In addition, soluble catalysts also known as redox mediators (RMs), such as FePc, [15] LiI, [16] 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO), [17] and MoCl 5 , [18] have been introduced into the Li-O 2 batteries to reduce the overpotential. Nevertheless, the shuttle effect deteriorates the corrosion and passivation of the Li anode, leading to poor stability of the batteries. [19] Recently, photoelectric conversion technology has been centralized to metal air batteries. [20][21][22] Integrating photo electric conversion technology into Li-O 2 batteries can effectively improve the round-trip efficiency. [23,24] In the photoassisted Li-O 2 battery system, the electrons in valence band (VB) of the catalyst are excited to the conduction band (CB) upon illumination, leaving holes in the VB. During discharge, electrons on the CB reduce oxygen and produce lithium peroxide, while electrons from the anode reduce holes in the VB rather than oxygen. Therefore, the discharge voltage can be higher than the potential for Li 2 O 2 formation. During charging, the photoinduced holes oxidize lithium peroxide and produce oxygen, and simultaneously the electrons from the CB transfer to the anode through the external circuit to reduce lithium ions into lithium metal. [25] As a consequence, the photoassisted Li-O 2 battery displays a lower charge voltage, even below 2.96 V. Up to now, several inorganic semiconductors and polymeric semiconductors have been introduced to all types of Li-O 2 batteries for photoenergy utilization. [26,27] The representative photocatalysts TiO 2 and g-C 3 N 4 show excellent photocatalytic performance, but the problems of limited active sites and difficult modification deserve further improvement. [28][29][30] Metal-organic frameworks (MOFs) have been considered as promising platforms for electrocatalytic reactions due to their hierarchical pore structures, tunable compositions, and Li-O 2 batteries have attracted extensive attention because of their theoretical specific energy equivalent to gasoline, but the poor electrical conductivity of Li 2 O 2 leads to high overpotential, which limits their further development and application. Herein, this work introduces Ce-UiO-66, a metal-organic framework material, as the photocatalyst to reduce the overpotential in the discharging and charging processes. With the simple in situ growth of Ce-UiO-66 on carbon cloth as an integrated photocathode, the photoassisted Li-O 2 batteries display decent discharge and charge voltages of 3.1 and 3.6 V, and a lifespan of 160 cycles. Moreover, theoretical calculations have been carried out to understand the band structure, spectroscopy, and photocatalytic property of Ce-UiO-66. The findings will encourage an enormous variety of novel MOFs-based photocathodes for solar energy utilization systems.

Introduction
With the booming development of the electric vehicle industry, the energy density of lithium-ion batteries has reached the bottleneck and is difficult to meet people's needs. [1][2][3] Therefore, aprotic Li-O 2 batteries with the highest theoretical energy density have been attracting more and more attention. Li-O 2 batteries are mainly based on the chemical reaction of 2Li + O 2 ↔ Li 2 O 2 (E 0 = 2.96 V), with energy density of 11 700 Wh kg −1 (without considering the mass of oxygen). [4][5][6] The essence of Li-O 2 batteries is the formation and decomposition of lithium peroxide in the process of discharge and charge. [7] Although Li-O 2 batteries possess the highest energy density of all batteries, the poor conduction of discharge well-dispersed metal sites. [31][32][33] Various MOFs and their derivatives have been used to improve the kinetics of ORR and OER in the Li-O 2 battery. [34,35] Very recently, Yu et al. reported perovskite quantum dots encapsulated in MOFs as synergistic photocathode materials for the Li-O 2 battery and showed a high level of performance. [36] Theoretically, MOFs possess the potential for photocatalysis independently, with semiconductor characteristics and modifiable property endowed by organic-inorganic hybrid and porous characteristics. [37,38] UiO-66 MOFs caught our eyes owing to its large surface area, abundant active sites, and excellent chemical resistance toward different solvents and conditions, which has been widely applied to wastewater treatment, gas sorption, and photocatalysis. [39][40][41] Among UiO-66 MOFs, the light absorption range of Ce-UiO-66 is extended to the visible light region due to the low-lying empty 4f orbitals of Ce, making it optimum candidate for photoassisted Li-O 2 batteries. [42] Herein, the integrated photocathode was simply prepared by in situ growth of Ce-UiO-66 on carbon substrate. The ultraviolet-visible (UV-Vis) spectra and Mott-Schottky (M-S) plot showed that the band gap of Ce-UiO-66 is 2.71 eV, and the potentials of VB and CB are 4.06 and 1.35 V relative to Li / Li + respectively. Compared with the dropping method, in situ generated Ce-UiO-66 could be fully contact with the carbon cloth and directly served as an active component. Upon illumination, the discharge and charge voltages of Li-O 2 batteries with integrated photocathode were stable at 3.1 and 3.6 V, respectively, in contrast with 2.8 and 4.0 V for dropping photocathode. Furthermore, the batteries showed stable operation up to 160 cycles, which is three times longer than the batteries with dropping photocathode. In addition, the structure, spectroscopy, and catalytic property of Ce-UiO-66 were well investigated by the combination of theoretical calculations and experiment results. This work indicates that MOFs are promising candidates as photocatalysts for the development of photoassisted Li-O 2 batteries.

Results and Discussion
Ce-UiO-66 was synthesized through a simple solution reaction similar to the reported method. [43] The procedure for preparing the photocathode is illustrated in Scheme 1. The detailed synthesis process can be found in the Experimental Section. As can be seen in Figure 1a, the obtained material is identified as Ce-UiO-66 by X-ray diffraction (XRD), matching well with its theoretical patterns. [44] The corresponding peaks of integrated photocathode are identical to those of cathode prepared by conventional method, manifesting that Ce-UiO-66 has grown on the carbon substrate successfully. As illustrated in Figure 1b and Figure S1a (Supporting Information), the pristine Ce-UiO-66 is composed of several spheroids with irregular shape. Figure 1c and Figure S1b (Supporting Information) display the scanning electron microscopy (SEM) images of integrated photocathode. Compared with the dropping photocathode in Figure 1d and Figure S1c (Supporting Information), Ce-UiO-66 in integrated photocathode is evenly distributed and the agglomeration of particles is restrained. The in situ growth method enables Ce-UiO-66 to be tightly combined with the carbon cloth. On the contrary, Ce-UiO-66 only contacts the carbon cloth on the surface in the dropping photocathode. Additionally, the pores and active sites of Ce-UiO-66 could be effectively maintained without the introduction of Ketjen black (KB) and poly(1,1difluoroethylene) (PVDF). The well-dispersed distribution of the Ce-UiO-66 located in the photocathode would promote the generation and separation efficiency of photogenerated carriers, and thus enhance its photocatalytic performances.
It has been proven that the properties of MOFs vary with the valence state of metal active sites. [45] X-ray photoelectron spectroscopy (XPS) characterization was first carried out to analyze the chemical states of elements in the Ce-UiO-66. The high-resolution spectra of Ce (3d) and O (1s) are displayed in Figure 1e,f. The peaks at 900.7, 889.1, and 882.4 eV are attributed to 3d 5/2 level of Ce (IV), and the peaks at 918.4, 908.2, and Scheme 1. Photocathodes prepared by dropping method and in situ growth method. www.advmatinterfaces.de 903.3 eV are ascribed to 3d 3/2 level of Ce (IV). The observed two peaks at 905.4 and 884.9 eV can be assigned to Ce (III). [46] In the O 1s spectra (Figure 1f), the peak at 531.6 eV can be attributed to the oxygen species present in the crystal lattice (O Latt ), whereas the peak at 533.0 eV represents the adsorption oxygen species (O Ads ). [47] The extra existence of Ce 3+ can be attributed to the partial reduction of Ce 4+ ( , and oxygen vacancies are thus formed along with the valence state change of Ce. [48] Particularly, the presence of Ce 3+ and oxygen vacancies can promote the adsorption of O 2 and the carrier separation during the photocatalytic process. [46] In order to estimate the band structure and photocatalytic potential, the semiconductor characteristics of Ce-UiO-66 were measured in combination with theoretical calculations. The light absorbance of Ce-UiO-66 was first tested by ultravioletvisible (UV-Vis) spectroscopy. Ce-UiO-66 in Figure 2a exhibits an absorption edge at 458 nm, featuring intensive visible-light absorption. The absorption wavelength corresponds to the band gap (E g ) of 2.71 eV, approximate to the previous report. [46] In addition, as shown in Figure 2b, M-S measurements were further conducted to determine the flat-band (E fb ) potential. The E fb is determined to be 0.779 V versus saturated calomel electrode (SCE), which correlates to the electrochemical potential of the top of the VB. Thus, the CB is estimated at −1.931 V versus SCE. Based on the results of the M-S test, the VB and CB of Ce-UiO-66 are then calculated as 4.06 and 1.35 V versus Li/Li + , respectively. We also conducted the ultraviolet photoelectron spectroscopy (UPS) test to further confirm the VB of Ce-UiO-66. The ionization potential which is equivalent to the VB energy (E v ) of Ce-UiO-66 is calculated to be 5.56 eV by subtracting the width of the He I UPS spectra from the excitation energy (21.22 eV) ( Figure S2a, Supporting Information). [49] According to the reference standard for which 0 V versus Li/ Li + equals −1.46 eV versus vacuum level, the E v of Ce-UiO-66 is thus estimated at 4.1 V versus Li/Li + , close to the result of the M-S curve (4.06 V). The optimized structure of Ce-UiO-66 is shown in Figure S3 (Supporting Information). The calculated absorption spectrum is shown in Figure 2c, and the absorption edge is determined to be 417 nm, which is close to the experiment value. We also calculated the DOS of Ce-UiO-66 based on HSE06 hybrid functional as shown in Figure 2d, the magnetic moment is determined to be zero µB without considering spin polarization in the DOS calculation. The calculated band gap value is 2.62 eV, only 3% smaller than the experimental value, and the small difference could be due to the ignoring of the phonon scattering effect. [50] As a result, the energy levels of Ce-UiO-66 corresponding to the electrochemical potentials for VB and CB are deemed as 4.06 and 1.35 V versus Li/ Li + , respectively, which matches well with the thermodynamic requirements of photocatalytic ORR and OER in the Li-O 2 batteries.
The schematic diagram of photoassisted Li-O 2 battery system is displayed in Figure S4 (Supporting Information). Photoassisted Li-O 2 batteries were assembled with photocathodes as air cathodes in 2032-type coin cells with 10 mm holes in the cathode shells for illumination and O 2 diffusion, and a cold light from xenon lamp served as the light source to eliminate the temperature influence as much as possible. According to the band structure, schematic diagram of Ce-UiO-66 photocatalytic reduction of overpotential could be illustrated in Figure 3a.
Under illumination, the electrons in the VB of Ce-UiO-66 are excited to the CB, leaving holes in the VB. During discharge, oxygen is reduced by electrons on the CB and produce lithium  www.advmatinterfaces.de peroxide, while lithium is oxidized by the holes in the VB. Accordingly, the discharge voltage could be theoretically increased from 2.96 to 4.06 V. Conversely, the holes in the VB of Ce-UiO-66 oxidize Li 2 O 2 and generate O 2 under illumination, and the electrons in the CB will reduce Li + to Li at the anode with the aid of external voltage. Theoretically, the charge voltage could be reduced from 2.96 to 1.35 V. As a consequence, Ce-UiO-66 is a satisfactory semiconductor material for photoassisted Li-O 2 batteries. Though the results of the linear sweep voltammetry (LSV) in the Figure 3b and Figure S5a (Supporting Information), higher current densities were observed in batteries with both photocathodes under illumination, while batteries with integrated photocathode showed a more obvious increase than those with dropping photocathodes. Instead, the batteries with carbon photocathodes showed almost identical current densities in dark and light conditions ( Figure S5b, Supporting Information). To confirm the superiority of photoassisted Li-O 2 batteries with integrated photocathode, batteries with integrated photocathode, dropping photocathode, and carbon photocathode were assembled and tested in the same atmosphere and light source with a current density at 0.018 mA cm −2 . As expected, the performance of the integrated photocathode was better than that of the dropping photocathode in Figure 3e. During the discharge process, the discharge voltage of Li-O 2 batteries with integrated photocathode was stable at 3.1 V in contrast with 2.8 V for Li-O 2 batteries with dropping photocathode. Discharge voltages of both batteries based on Ce-UiO-66 were higher than that of the pure carbon electrode. Meanwhile, in the charge process, the effect of integrated photocathode on reducing the charge overpotential of Li-O 2 batteries was also the most significant, which was reduced from 4.5 to 3.6 V. In contrast, the charge voltage of Li-O 2 battery with dropping photocathode was reduced from 4.5 to 4.0 V. Obviously, Ce-UiO-66 has the ability to introduce solar power to the energy conversion of Li-O 2 batteries and its function could be further released by a more suitable electrode structure design. Overall, upon illumination, the discharge voltage of Li-O 2 batteries assembled with integrated photocathode was increased from 2.7 to 3.1 V, and the charge voltage was reduced from 4.5 to 3.6 V, with an increased roundtrip efficiency of 43.5%.
In order to exclude the possibility of direct redox reaction between lithium and Ce-UiO-66, the constant current roundtrip test of batteries in argon atmosphere was also carried out at the same condition. As shown in Figure S7 (Supporting Information), the discharge voltage of the photoassisted battery was even lower than 2 V, and the charge voltage attained almost close to 4.7 V, without obvious plateaus. Therefore, the key functional role of Ce-UiO-66 in photoassisted Li-O 2 batteries is the promotion of ORR and OER processes. The O 2 adsorption energy on the Ce site was calculated, the corresponding adsorption configuration and adsorption energy are shown in Figure S8 and Table S1 (Supporting Information). It can be found that the adsorption of O 2 on intact Ce-UiO-66 unite is inferior. From the Table S1 (Supporting Information), the O 2 adsorption energy of the configuration 4 is −0.34 eV, lowest in all five configurations. Therefore, the optimal adsorption site of O 2 is the bridge site between two Ce 3+ sites of the unit. In agreement with the XPS results, O 2 is more likely to adsorb on the defect sites of metal clusters in Ce-UiO-66. Based on the results of theoretical calculations and experimental measurements, the good photocatalytic properties of Ce-UiO-66 photocathode probably results from the good visible light-responsive core center, the strong O 2 activation of surface defects, and the boosted electron transfer. [51][52][53] The discharge product of the photoassisted Li-O 2 batteries was identified as Li 2 O 2 by XRD in Figure 4a, matching well with the standard powder diffraction file. After the charge process, the peaks of lithium peroxide disappeared and the XRD patterns were corresponded basically to those of pristine photocathode. In addition, the signal peak of Li 1s (55.4 eV) was found in the XPS patterns of the photocathode after discharge in Figure 4b. As expected, there was no signal peak of Li 1s after the charge process. The appearance and disappearance of peaks in Figure 4a,b also indicated that the discharge and charge of photoassisted Li-O 2 batteries are reversible processes. The photocathode after discharge was then tested by SEM to analyze the morphology of the discharge products. As can be seen in Figure 4c and Figure S9 (Supporting Information), large amounts of Li 2 O 2 particulates densely deposit on the surface of carbon fibers and the typical structure of Li 2 O 2 is toroidal. Not surprisingly, the SEM image of the photocathode after charge (Figure 4d) was similar to the original electrode (Figure 1c), meaning that Li 2 O 2 disappeared after the charge process.
The rate performance of photoassisted Li-O 2 batteries has also been conducted, and the corresponding results are shown in Figure S10 (Supporting Information). Obviously, the battery with integrated photocathode shows the best electrochemical performance in all currents. Compared with the Ce-UiO-66 photo cathodes, carbon photocathode cannot maintain a stable discharge platform at 0.08 mA, demonstrating the critical role of Ce-UiO-66 in photocatalytic processes. The cycling performance of Li-O 2 batteries with different photocathodes was firstly investigated by galvanostatic cycling at a current of 0.018 mA cm −2 . As displayed in Figure S11a,b (Supporting Information), the battery with dropping photocathode circulated only 19 cycles, accumulating 78 h. Particularly, the battery with integrated photo cathode lasted 80 discharge and charge cycles, with a total of 322 h. After 80 cycles, there was no obvious charge and discharge platform and the test was terminated. Nevertheless,

www.advmatinterfaces.de
Li-O 2 battery with carbon photocathode cannot circulate under the same conditions for the charge voltage was over 4.5 V even at the first cycle ( Figure S12, Supporting Information). As we know, Li-O 2 batteries belong to a semi-open and three-phase reaction system. In the photoassisted system, light could inevitably accelerate the volatilization of electrolyte, resulting in the instability of the Li-O 2 batteries. With the same capacity, higher current density can shorten the cycle time, which could reduce the influence of electrolyte volatilization on cycle stability. Li-O 2 batteries with different photocathodes were further tested at higher current densities. As shown in Figure 4e, the electrochemical performances of Li-O 2 battery assembled with integrated photocathode can maintain a decent level at five times current density (J = 0.09 mA cm −2 ). The photoassisted battery showed a discharge-charge voltage of 3.0/3.65 V at the first cycle, and lasted 160 cycles (12.8 mAh cm −2 ). In contrast, the battery with the dropping photocathode only circulated 40 cycles (3.2 mAh cm −2 ) at the same condition. Obviously, the well-dispersed distribution of the Ce-UiO-66 in the integrated photocathode can tightly contact with the carbon cloth and favor the charge transfer process, and thus promote photoelectrocatalysis and the utilization of solar energy in Li-O 2 batteries. Compared with the previous reported results of photoassisted Li-O 2 batteries in Figure 4f, [24,25,27,29,35,36,[54][55][56][57][58][59][60][61][62][63] the overall performance of the Ce-UiO-66 photocathode exceeds most results including discharge and charge voltages, discharge capacity, and cycle stability. It is believed that MOFs are promising candidates as photocatalysts for the development of photoassisted Li-O 2 batteries, and better performance can be obtained by the cooperation of structure designs, cocatalysts, heterojunctions, and guest molecules.

Conclusion
Ce-UiO-66 as a single-component photocatalyst has been successfully introduced into Li-O 2 batteries to accelerate the ORR and OER processes under illumination. The results of

Experimental Section
Materials Synthesis: An amount of 0.708 g p-phthalic acid was dissolved in 24 mL of N,N-dimethylformamide (DMF).An amount of 8 mL of 0.5333 mol L −1 cerium (IV) diammonium nitrate solution was also prepared. Then both kinds of solution were mixed and stirred in 95 °C water bath for 1 h. After cooling and filter, the filter residue was washed with DMF and absolute ethanol successively, and then dried in 70 °C vacuum oven to obtain Ce-UiO-66 product.
Photocathode Preparation: Ce-UiO-66, KB, and PVDF were mixed according to the mass ratio of 6:3:1. The mixture was dissolved in solvent N-methylpyrrolidone (NMP). After stirring 2 h, the solution was evenly dropped onto the carbon cloth. Then the carbon cloth was transferred to the vacuum oven and dried at 80 °C for 24 h to prepare the photocathode. This is the traditional dropping method to prepare photocathode. In contrast with traditional dropping method, the in situ growth method was basically the same as the method for preparing Ce-UiO-66, except that some carbon cloth was added to the heating solution.
Materials Characterization: XRD was performed on a Rigaku Ultima IV 2036E102 X-ray diffractometer with Cu Ka radiation (λ = 1.5406 Å) at 4° min −1 . SEM (JEOL JSM-7800F) was used to obtain the morphology and XPS measurements were performed on Thermo Scientific ESCALAB 250Xi. UV-Vis spectroscopy was performed on a SHIMADZU UV-2550 UV-Vis spectrophotometer in the range of 200-800 nm. UPS was performed on a Thermo ESCALAB XI+ with the bias voltage of −5 eV.
Cell Assembly and Electrochemical Measurements: All the chemicals were stored in an argon-filled glovebox with H 2 O and O 2 content <0.1 ppm and the batteries were stabilized for 4 h before tests. The electrolyte was a solution of 1 mol L −1 LiTFSI in TEGDME. The lightassisted Li-O 2 battery was assembled with a Li foil anode, a glass fiber filter (Whatman GF/A) encapsulated with the electrolyte, and Ce-UiO-66 photocathode as an oxygen electrode and photocathode in a 2032type coin cell with a 10 mm hole on the cathode shell as windows for illumination and O 2 diffusion. The batteries were tested in a homemade device filled with O 2 . All electrochemical measurements were carried out with a model CHI660E electrochemical analyzer (Chenhua Instrumental Co., Ltd, Shanghai, China) and LANDCT2001A at room temperature. A 300 W BBZM cold source Xe-lamp was used for illumination as solar source.
Determinations of Band Edges: The flat-band potential (E fb ) of a semiconductor can be determined by M-S equation from Equation (1). The M-S curves were measured in 0.5 mol L −1 Na 2 SO 4 solution at 1 KHz. A curve of 1/C 2 against E should yield a straight line from which E fb can be obtained from the E axis. The band gap can be determined by applying the formula from Equation (2). Tauc's curve showed a plot of αhν against E due to the results of UV absorption spectrum. When αhν is zero, the x-intercept equals the band gap (E g ) of the semiconductor material. And the E vb of the semiconductor can be calculated in Equation (3) since E fb is known. Consequently, the E cb can be determined after confirming the E g and E vb .
Computational Details: All the theoretical calculations were performed by the VASP (Vienna ab initio Simulation Package). [64] Perdew-Burke-Ernzerhof (PBE) functional [65] within generalized-gradient-approximation (GGA) was used to describe the exchange-correlation functional. The cutoff energy was set to 520 eV. The convergence criteria for selfconsistent field calculation and structure optimization were 10 −5 eV and 0.05 eV Å −1 , respectively. The Gaussian smearing with a width of 0.05 eV was used. DFT-D3 method of Grimme with zero-damping function was used to consider the vdW dispersion correction. [66] The primitive cell of Ce-UiO-66 which contains 114 atoms, including 28 H, 48 C, 32 O, and 6 Ce was used. 2 × 2 × 2 and 3 × 3 × 3 Gamma centered Monkhorst-Pack scheme were used for structure optimization and statical calculation to sample the Brillouin zone. [67] A single gamma point sampling in combination with HSE06 [68] hybrid functional was used for the density of state (DOS) calculation as in the previous report. [69] The absorption spectrum was calculated according to: where α is the absorption coefficient, ε 1 and ε 2 are the real and imaginary parts of the dielectric function, ω is the photon frequency, and c is the light speed. Postprocessing of the VASP computational data was carried out by VASPKIT code. [70] The adsorption of O 2 on the Ce site was calculated with one Ce-UiO-66 unit, which contains 90 atoms, including 28 H, 24 C, 32 O, and 6 Ce in a 30 × 30 × 30 (Å 3 ) cell. The adsorption of O 2 on the defect sites of Ce-UiO-66 was investigated by one unite metal clusters with two ligands removed. The revised Perdew-Burke-Ernzerhof (RPBE) functional [71] was applied to describe the exchange-correlation functional and the cutoff energy was set to 450 eV for calculating the adsorption energy. 1 × 1 × 1 Gamma centered Monkhorst-Pack scheme was used for structure optimization to sample the Brillouin zone. The O 2 adsorption energy was calculated based on: in which ΔE ads represents the adsorption energy of O 2 adsorbed on Ce site of the unit, E adsorption_configuration , E O2 , and E unit are the energy calculated by VASP of adsorption configuration, O 2 , and the unit, respectively.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.