Synthesis and Oxygen Storage Capacities of Yttrium-Doped CeO2 with a Cubic Fluorite Structure

Doping CeO2 with Y cations was achieved in this study using three strategies: doping only during the hydrothermal process (H-Y-doped CeO2), doping only during the impregnation process (I-Y-doped CeO2), and doping during both the hydrothermal and impregnation processes (H/I-Y-doped CeO2). During the three synthesis strategies of Y-doped CeO2, these Y ions could be incorporated into the CeO2 lattice in the +3 state while holding the cubic fluorite structure, and no impurity phases were detected. Pure CeO2 crystal itself contained a certain number of intrinsic VO defects, and Y-doping was beneficial for the creation of extrinsic VO defects. The relative concentrations of VO defects were quantified by the values of A592/A464 obtained from Raman spectra, which were 1.47, 0.93, and 1.16 for the H-Y-, I-Y-, and H/I-Y-doped CeO2, respectively, and were higher than that of the undoped one (0.67). Moreover, the OSCs of the three Y-doped CeO2 were enhanced, and the sequence of OSCs was: H-Y-doped CeO2 (0.372 mmol/g) > H/I-Y-doped CeO2 (0.353 mmol/g) > I-Y-doped CeO2 (0.248 mmol/g) > Undoped CeO2 (0.153 mmol/g); this result was in good agreement with the Raman spectroscopy results.


Introduction
With the continuous development of science and technology, the development and utilization of energy and resources have become a hot topic [1][2][3]. Rare-earth elements, known as "industrial monosodium glutamate", "industrial vitamin", and "mother of new materials", are precious strategic metal resources. Cerium (Ce) is the most abundant rare-earth element in the Earth's crust. Its common valence states are +3 and +4, and the corresponding oxides are cerium sesquioxide (Ce 2 O 3 ) and cerium dioxide (CeO 2 ) [4][5][6]. Ce 2 O 3 is unstable in air and easily oxidizes to CeO 2 . Interestingly, there are not only Ce 4+ ions in CeO 2 crystals but also trace amounts of Ce 3+ ions. The oxidation/reduction cycle composed of Ce 3+ and Ce 4+ states (Ce 3+ ↔Ce 4+ ) enables CeO 2 to store and release oxygen, referred to as oxygen storage capability (OSC). In other words, CeO 2 can release oxygen under reducing conditions, forming nonstoichiometric oxides CeO 2−x , and the CeO 2−x can store oxygen by filling oxygen vacancies under oxidizing conditions [7,8]. In the atmosphere or an oxygen-rich environment, CeO 2 can store some oxygen in its own lattice, and these stored oxygen atoms can be released quickly when the partial pressure of ambient oxygen decreases. Precisely because of this ability, CeO 2 is considered an excellent catalyst in CO 2 methanation [9,10], hydrodeoxygenation of xylitol and fatty acids [11,12], NO x conversion [13,14], and so on.
A flow chart of the synthesis procedures employed for the undoped and Y-doped CeO 2 samples is shown in Figure 1. The reference CeO 2 without Pr-doping was prepared by the hydrothermal process using Route 1 (R1), as shown in Figure 1, denoted as Undoped CeO 2 . The CeO 2 samples doped with Pr cations during only the impregnation process or the hydrothermal process were synthesized as shown in Route 2 (R2) and Route 3 (R3) of Figure 1, denoted as I-Y-doped CeO 2 and H-Y-doped CeO 2 , respectively, while the one doped during both the impregnation and hydrothermal processes was synthesized as shown in Route 4 (R4) of Figure 1, denoted as H/I-Y-doped CeO 2 .
Undoped CeO 2 was synthesized using R1: 4.0 mmol Ce(NO 3 ) 3 ·6H 2 O, 25 mL EG, and 5 mL distilled water were added into a 50 mL Teflon-lined stainless-steel autoclave and sealed at 200 • C for 24 h. Afterward, the Ce precursor was collected, washed, and dried in turn. Finally, Undoped CeO 2 was obtained by subsequent calcination in air at 500 • C for 2 h.
I-Y-doped CeO 2 was synthesized using R2: the Ce precursor synthesized in R1 was impregnated into a saturated solution of Y 3+ ions at room temperature for 24 h. After filtration and drying, the I-Y-doped CeO 2 was obtained by subsequent calcination in air at 500 • C for 2 h.
H-Y-doped CeO 2 was synthesized using R3: 3.84 mmol Ce(NO 3 ) 3 ·6H 2 O, 0.16 mmol Y(NO 3 ) 3 ·6H 2 O, 25 mL EG, and 5 mL distilled water were added into a 50 mL Teflon-lined stainless-steel autoclave and sealed at 200 • C for 24 h. Afterward, the Ce/Y precursor was collected, washed, and dried in turn. Finally, H-Y CeO 2 was obtained by subsequent calcination in air at 500 • C for 2 h.
H/I-Y-doped CeO 2 was synthesized using R4: the Ce/Y precursor synthesized in R3 was impregnated into a saturated solution of Y 3+ ions at room temperature for 24 h. After filtration and drying, the H/I-Y-doped CeO 2 was obtained by subsequent calcination in air at 500 • C for 2 h. turn. Finally, Undoped CeO2 was obtained by subsequent calcination in air at 500 °C for 2 h. I-Y-doped CeO2 was synthesized using R2: the Ce precursor synthesized in R1 was impregnated into a saturated solution of Y 3+ ions at room temperature for 24 h. After filtration and drying, the I-Y-doped CeO2 was obtained by subsequent calcination in air at 500 °C for 2 h.
H-Y-doped CeO2 was synthesized using R3: 3.84 mmol Ce(NO3)3•6H2O, 0.16 mmol Y(NO3)3•6H2O, 25 mL EG, and 5 mL distilled water were added into a 50 mL Teflon-lined stainless-steel autoclave and sealed at 200 °C for 24 h. Afterward, the Ce/Y precursor was collected, washed, and dried in turn. Finally, H-Y CeO2 was obtained by subsequent calcination in air at 500 °C for 2 h.
H/I-Y-doped CeO2 was synthesized using R4: the Ce/Y precursor synthesized in R3 was impregnated into a saturated solution of Y 3+ ions at room temperature for 24 h. After filtration and drying, the H/I-Y-doped CeO2 was obtained by subsequent calcination in air at 500 °C for 2 h.

Characterization
The crystallographic phases of the samples were characterized by X-ray diffraction (XRD, DX-2700). The surface composition and binding energy of the CeO2 samples were determined by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Scientific, Waltham, MA, USA). The oxygen vacancy (VO) defects of the CeO2 samples were characterized using a Raman spectrometer (LabRAM Aramis, Horiba Jobin-Yvon, Paris, France) with a He-Cd laser of 325 nm, and the exposure time for the measurement set was 60 s.

Characterization
The crystallographic phases of the samples were characterized by X-ray diffraction (XRD, DX-2700). The surface composition and binding energy of the CeO 2 samples were determined by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Scientific, Waltham, MA, USA). The oxygen vacancy (V O ) defects of the CeO 2 samples were characterized using a Raman spectrometer (LabRAM Aramis, Horiba Jobin-Yvon, Paris, France) with a He-Cd laser of 325 nm, and the exposure time for the measurement set was 60 s.

OSC
The OSC of CeO 2 was estimated using O 2 temperature-programmed desorption (O 2 -TPD) measurements, which were carried out in a plug-flow microreactor system (TP5000) with a thermal conductivity detector, and the amount of O 2 desorption during the process was measured by the thermal conductivity detector. About 0.1 g of CeO 2 powder was activated using an air stream at 400 • C for 30 min, then moved into He and cooled, then exposed to an air stream for 30 min at 120 • C, followed by purging with a He stream to remove the excess O 2 . Finally, the surface oxygen desorption was conducted at a flow rate of He (10 mL/min) while the temperature was raised to~900 • C (10 • C/min).

Results and Discussion
XRD was employed to characterize the phase composition of the as-synthesized undoped and Y-doped samples. Figure 2 shows the XRD patterns of the Undoped CeO 2 and H-Y-, I-Y-, and H/I-Y-doped CeO 2 powders. For the XRD pattern of the Undoped CeO 2 , eight well-resolved peaks were observed, which could be indexed to the (111), (200), (220), (311), (222), (400), (331), and (420) planes of cubic CeO 2 (JCPDS no. 34-0394; fluorite). No additional phases were detected, suggesting pure CeO 2 had been obtained by the hydrothermal process using route R1 in Figure 1. After the introduction of Y cations in the synthesis process, the XRD patterns of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 samples exhibited a similar profile to that of the Undoped CeO 2 . However, no peaks for impurity phases such as Y 2 O 3 were detected, which could be explained as follows: the impurities in Y-doped CeO 2 samples might exist as highly dispersed or amorphous surface species, or the amount of the Y impurity was low. Another possibility is that the Y cations partially substituted the Ce ions to form a solid solution. The inset in Figure 2 shows that the (111) reflection shifts toward lower 2θ values with the incorporation of Y ions, and it can be found that the shift exhibited by the H-Y-doped CeO 2 sample was the greatest. Moreover, the lattice parameters of CeO 2 were estimated using Bragg's equation and summarized in Table 1. It was found that the calculated lattice parameters for H-Y-(5.4242 Å), I-Y-(5.4190 Å), and H/I-Y-doped (5.4227 Å) CeO 2 were greater than that of the undoped sample (5.4117 Å). These findings implied that the large Y ions (1.02 Å) partially substituted the Ce 4+ ions (0.97 Å [32]) to form a CeO 2 -based solid solution while holding the cubic fluorite structure of CeO 2 . In order to probe the possibility of the presence of the Y element in CeO 2 , as well as the chemical state of its presence, XPS was employed to study the Undoped CeO 2 and Y-doped CeO 2 . Figure 3a shows the wide-scan XPS spectra of the Undoped CeO 2 and H-Y-, I-Y-, and H/I-Y-doped CeO 2 powders. As observed in Figure 3a, the XPS profiles of all the samples were similar, dominated by the signals of Ce, O, and C elements, in accordance with a previous report for pure and Y-doped CeO 2 [33]. However, there was no sign of the Y element at first sight from the wide-scan XPS spectra of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 . To ascertain whether the CeO 2 contained the Y cations and the Y-doping was real, the corresponding Y 3d XPS regions of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 were recorded. For the Y 3d XPS regions of the H-Y-and H/I-Y-doped CeO 2, we could cleanly identify the Y 3d signal and its unique contour, which is assigned to the trivalent Y ions. However, the Y signal was weak from the Y 3d XPS regions of the I-Y-doped CeO 2 , yet its signal peak of Y 3d could still be identified in graphing alone (inset in Figure 3b).
found that the shift exhibited by the H-Y-doped CeO2 sample was the greatest. Moreover, the lattice parameters of CeO2 were estimated using Bragg's equation and summarized in Table 1. It was found that the calculated lattice parameters for H-Y-(5.4242 Å ), I-Y-(5.4190 Å ), and H/I-Y-doped (5.4227 Å ) CeO2 were greater than that of the undoped sample (5.4117 Å ). These findings implied that the large Y ions (1.02 Å ) partially substituted the Ce 4+ ions (0.97 Å [32]) to form a CeO2-based solid solution while holding the cubic fluorite structure of CeO2. In order to probe the possibility of the presence of the Y element in CeO2, as well as the chemical state of its presence, XPS was employed to study the Undoped CeO2 and Ydoped CeO2. Figure 3a shows the wide-scan XPS spectra of the Undoped CeO2 and H-Y-, , we could clea identify the Y 3d signal and its unique contour, which is assigned to the trivalent Y i However, the Y signal was weak from the Y 3d XPS regions of the I-Y-doped CeO2, ye signal peak of Y 3d could still be identified in graphing alone (inset in Figure 3b). In order to understand the effect of Y-doping on Ce ions in CeO2 crystals, the Ce XPS regions of the Undoped CeO2 and H-Y-, I-Y-, and H/I-Y-doped CeO2 were recor and fitted, as shown in Figure 4. The Ce 3d core-level XPS of all CeO2 samples could fitted into eight peaks, referring to the 3d5/2 and 3d3/2 spin-orbit doublet of Ce cations cluding Ce 3+ and Ce 4+ ions). The bonds labeled as v2 and u2 belong to the spin-doublet t of the Ce 3+ state, and the bands labeled as v4, v3, and v1 (and those for u) are due to the of the Ce 4+ state [34]. A quantitative analysis of the concentration of Ce 3+ ions based on measured Ce 3d XPS spectra, labeled as [Ce 3+ ]XPS, could be performed using Equation and the results are summarized in Table 1. The [Ce 3+ ]XPS values of the H-Y-, I-Y-, and Y-doped CeO2 were 12.60%, 8.95%, and 11.37%, respectively. These were higher than of the Undoped CeO2 (6.54%), indicating that pure CeO2 crystal itself contains a cer number of Ce 3+ ions and that Y-doping could promote the formation of Ce 3+ species, e cially H-Y-doped CeO2, which exhibited the highest [Ce 3+ ]XPS values. The Ce 3+ specie pure CeO2 contributed to the OSC of CeO2 through the oxidation/reduction cycle c posed of Ce 3+ and Ce 4+ states (Ce 3+ Ce 4+ ).
where Ai is the integrated area of the ith fitting peak from Ce 3d XPS spectra. In order to understand the effect of Y-doping on Ce ions in CeO 2 crystals, the Ce 3d XPS regions of the Undoped CeO 2 and H-Y-, I-Y-, and H/I-Y-doped CeO 2 were recorded and fitted, as shown in Figure 4. The Ce 3d core-level XPS of all CeO 2 samples could be fitted into eight peaks, referring to the 3d 5/2 and 3d 3/2 spin-orbit doublet of Ce cations (including Ce 3+ and Ce 4+ ions). The bonds labeled as v 2 and u 2 belong to the spin-doublet term of the Ce 3+ state, and the bands labeled as v 4 , v 3 , and v 1 (and those for u) are due to the case of the Ce 4+ state [34]. A quantitative analysis of the concentration of Ce 3+ ions based on the measured Ce 3d XPS spectra, labeled as [Ce 3+ ] XPS , could be performed using Equation (1), and the results are summarized in Table 1. The [Ce 3+ ] XPS values of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 were 12.60%, 8.95%, and 11.37%, respectively. These were higher than that of the Undoped CeO 2 (6.54%), indicating that pure CeO 2 crystal itself contains a certain number of Ce 3+ ions and that Y-doping could promote the formation of Ce 3+ species, especially H-Y-doped CeO 2 , which exhibited the highest [Ce 3+ ] XPS values. The Ce 3+ species in pure CeO 2 contributed to the OSC of CeO 2 through the oxidation/reduction cycle composed of Ce 3+ and Ce 4+ states (Ce 3+ ↔Ce 4+ ).
where A i is the integrated area of the ith fitting peak from Ce 3d XPS spectra.  In order to understand the effect of Y-doping on oxygen ions in CeO2 crystals, the O 1s XPS regions of the Undoped CeO2 and H-Y-, I-Y-, and H/I-Y-doped CeO2 were recorded and fitted, as shown in Figure 5. The O 1s XPS spectrum of the Undoped CeO2 was curvefitted into two peaks: one peak, labeled as α, at ~529.2 eV, could be attributable to the lattice oxygen species; the other peak, labeled as β, at ~531.4 eV, could be attributable to the chemisorbed oxygen species and/or weakly bonded oxygen species related to the oxygen vacancy (VO) defects. For the O 1s spectra of the H-Y-, I-Y-, and H/I-Y-doped CeO2, a new peak labeled as γ, at~528.5 eV, was curve-fitted, which could be assigned to the O-Y species. In addition, the relative oxygen vacancies ratio (labeled as [VO]XPS) could be quantified using Equation (2), and the results are shown in Table 1.
where Ai is the integrated area of the ith fitting peak from O 1s XPS spectra. As observed in Table 1  In order to understand the effect of Y-doping on oxygen ions in CeO 2 crystals, the O 1s XPS regions of the Undoped CeO 2 and H-Y-, I-Y-, and H/I-Y-doped CeO 2 were recorded and fitted, as shown in Figure 5. The O 1s XPS spectrum of the Undoped CeO 2 was curvefitted into two peaks: one peak, labeled as α, at~529.2 eV, could be attributable to the lattice oxygen species; the other peak, labeled as β, at~531.4 eV, could be attributable to the chemisorbed oxygen species and/or weakly bonded oxygen species related to the oxygen vacancy (V O ) defects. For the O 1s spectra of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 , a new peak labeled as γ, at~528.5 eV, was curve-fitted, which could be assigned to the O-Y species. In addition, the relative oxygen vacancies ratio (labeled as [V O ] XPS ) could be quantified using Equation (2), and the results are shown in Table 1.
where A i is the integrated area of the ith fitting peak from O 1s XPS spectra. As observed in To further investigate the V O defects, Raman spectra of CeO 2 were obtained. Raman spectroscopy is a powerful tool for the structural characterization of metal oxides due to its sensitivity to structural changes, such as V O defects. Figure 6 shows the Raman spectra of the Undoped and H-Y-, I-Y-, and H/I-Y-doped CeO 2 . For the Undoped CeO 2 , the spectral envelope in the 200~1000 cm −1 range displayed a strong band at 464 cm −1 associated with the triply degenerate F 2g vibrational mode of CeO 2 [35,36], while the band located at 592 cm −1 was associated with the optical LO mode of substoichiometric CeO 2−x units, underscoring an increase in V O defects [37,38]. Upon Y-doping, increases in the intensity of the bands located at 464 and 592 cm −1 were observed for the H-Y-, I-Y-, and H/I-Y-doped CeO 2 , which were associated with the presence of substoichiometric CeO 2-x , underscoring an increase in V O defects. However, the Y 3+ -doping into the CeO 2 lattice Materials 2022, 15, 8971 7 of 12 increased its lattice distortion and hence interfered with the vibrations of CeO 2 . Therefore, the Raman band intensities of the Y-doped CeO 2 were clearly affected by the incorporation of Y 3+ into the CeO 2 lattice. Consequently, an alternative approach to estimate the relative concentration of V O defects can be adopted by calculating the ratio of the integrated area of the Raman band at 592 cm −1 to that of 464 cm −1 (labeled [V O ] Raman ) [39,40]. The values of A 592 /A 464 , that is, the relative concentration of oxygen vacancies ([V O ] Raman ), are shown in Table 1 To further investigate the VO defects, Raman spectra of CeO2 were obtained. Raman spectroscopy is a powerful tool for the structural characterization of metal oxides due to its sensitivity to structural changes, such as VO defects. Figure 6 shows the Raman spectra of the Undoped and H-Y-, I-Y-, and H/I-Y-doped CeO2. For the Undoped CeO2, the spectral envelope in the 200~1000 cm −1 range displayed a strong band at 464 cm −1 associated with the triply degenerate F2g vibrational mode of CeO2 [35,36], while the band located at 592 cm −1 was associated with the optical LO mode of substoichiometric CeO2−x units, underscoring an increase in VO defects [37,38]. Upon Y-doping, increases in the intensity of the bands located at 464 and 592 cm −1 were observed for the H-Y-, I-Y-, and H/I-Y-doped CeO2, which were associated with the presence of substoichiometric CeO2-x, underscoring an increase in VO defects. However, the Y 3+ -doping into the CeO2 lattice increased its lattice distortion and hence interfered with the vibrations of CeO2. Therefore, the Raman band intensities of the Y-doped CeO2 were clearly affected by the incorporation of Y 3+ into the CeO2 lattice. Consequently, an alternative approach to estimate the relative concentration of VO defects can be adopted by calculating the ratio of the integrated area of the Raman band at 592 cm −1 to that of 464 cm −1 (labeled [VO]Raman) [39,40]. The values of A592/A464, that is, the relative concentration of oxygen vacancies ([VO]Raman), are shown in Table 1  According to the Ce 3d XPS analyses in Figure 4, it can be seen that pure CeO2 had a certain number of Ce 3+ ions, contributing to the OSC of CeO2 with the formation and filling of intrinsic VO defects, which could be expressed by Equation (3) and written using Kroger and Vink notations as in Equation (4). In the synthesis of CeO2, Y ions were introduced and doped into the CeO2 lattice, and a substoichiometric CeO2-x unit was formed with an increase in VO defects. The creation of extrinsic VO defects could be expressed by Equations (5) and (6). The vacancy compensation mechanism has been suggested for the increased According to the Ce 3d XPS analyses in Figure 4, it can be seen that pure CeO 2 had a certain number of Ce 3+ ions, contributing to the OSC of CeO 2 with the formation and filling of intrinsic V O defects, which could be expressed by Equation (3) and written using Kroger and Vink notations as in Equation (4). In the synthesis of CeO 2 , Y ions were introduced and doped into the CeO 2 lattice, and a substoichiometric CeO 2-x unit was formed with an increase in V O defects. The creation of extrinsic V O defects could be expressed by Equations (5) and (6). The vacancy compensation mechanism has been suggested for the increased concentration of V O for Y-doping into CeO 2 . As shown in Equations (3) and (4), besides the intrinsic V O in CeO 2 , there are two additional kinds of V O : one V O is created to balance the charge when two adjacent Ce 4+ cations are substituted by two Y 3+ cations, as shown in Equation (5); and substitution of one Ce 4+ by one Y 3+ gives rise to the formation of one V O with the adjacent Ce 4+ reduced to Ce 3+ , as shown in Equation (6).
where Y Ce represents a Y 3+ cation occupying the site of a Ce 4+ cation and Ce Ce represents a Ce 3+ cation occupying the site of a Ce 4+ cation. V O and V •• O represent an oxygen vacancy defect and one with two positive charges, respectively, which are produced via the vacancy compensation mechanism; and O × O is a lattice oxygen atom. From the O 1s XPS analysis in Figure 5 and the Raman analysis in Figure 6, it can be seen that pure CeO 2 crystal itself contained a certain number of V O species (namely intrinsic V O defects), which exhibited a large deviation from stoichiometry in the atmosphere, forming nonstoichiometric oxide CeO 2−x . After doping with Y 3+ cations, CeO 2 could still retain its fluorite crystal structure (see Figure 2), accompanied by the loss of oxygen from its lattice and the consequent formation of a large number of extrinsic V O defects (see Table 1, Figures 5 and 6).  [41][42][43]. By comparing these data of undoped and Y-doped CeO 2 , we could find that the lattice expansion occurred upon the incorporation of Y 3+ into the CeO 2 lattice, accompanied by the presence of Ce 3+ ions and more V O defects, which were consistent with our results, despite different methods of quantification. However, the relative content for Ce 3+ ions ([Ce 3+ ] XPS ) decreased with the doping of Y 3+ in the report [44], and the authors attributed this to the substitution of Y 3+ for Ce 3+ .

Authors Lattice Parameters (Å) [Ce 3+ ] XPS (%) [V O ] XPS (%) [V O ] Raman
Xu et al. [ OSC is the basic characteristic of CeO 2 and the premise of numerous applications. Therefore, the O 2 -TPD experiment was employed to evaluate the OSC of CeO 2 . Figure 7 shows the O 2 -TPD spectra of the Undoped and H-Y-, I-Y-, and H/I-Y-doped CeO 2 powders. For the Undoped CeO 2 , the asymmetrical peak of either the low temperature at~170 • C or the high temperature at~600 • C (light yellow area) implied the existence of at least two kinds of oxygen species at various coordination environments. The oxygen desorption at low temperatures could be attributed to the release of surface/subsurface lattice oxygen, while the oxygen desorption at high temperatures could be ascribed to the release of bulk lattice oxygen, which was consistent with the reported results [45]. Moreover, it could be clearly observed that the oxygen desorption was rapid at the early stages of the process (120~170 • C), suggesting that there were large amounts of adsorbed oxygen on CeO 2 , which could emigrate quickly at low temperatures. After the temperature reached~170 • C, the oxygen desorption started to decrease quickly until a temperature of~350 • C was reached, and then basically maintained a steady release of oxygen until 570 • C. Subsequently, the Undoped CeO 2 experienced the second oxygen release from 570 to 820 • C. Remarkably, the O 2 -TPD curve coincided with the baseline after 820 • C, suggesting that there was little release of oxygen for Undoped CeO 2 after 820 • C. ever, the relative content for Ce 3+ ions ([Ce 3+ ]XPS) decreased with the doping of Y 3+ in the report [44], and the authors attributed this to the substitution of Y 3+ for Ce 3+ .
OSC is the basic characteristic of CeO2 and the premise of numerous applications. Therefore, the O2-TPD experiment was employed to evaluate the OSC of CeO2. Figure 7 shows the O2-TPD spectra of the Undoped and H-Y-, I-Y-, and H/I-Y-doped CeO2 powders. For the Undoped CeO2, the asymmetrical peak of either the low temperature at ~170 °C or the high temperature at ~600 °C (light yellow area) implied the existence of at least two kinds of oxygen species at various coordination environments. The oxygen desorption at low temperatures could be attributed to the release of surface/subsurface lattice oxygen, while the oxygen desorption at high temperatures could be ascribed to the release of bulk lattice oxygen, which was consistent with the reported results [45]. Moreover, it could be clearly observed that the oxygen desorption was rapid at the early stages of the process (120~170 °C ), suggesting that there were large amounts of adsorbed oxygen on CeO2, which could emigrate quickly at low temperatures. After the temperature reached ~170 °C , the oxygen desorption started to decrease quickly until a temperature of ~350 °C was reached, and then basically maintained a steady release of oxygen until ~570 °C . Subsequently, the Undoped CeO2 experienced the second oxygen release from 570 to 820 °C. Remarkably, the O2-TPD curve coincided with the baseline after 820 °C, suggesting that there was little release of oxygen for Undoped CeO2 after 820 °C. For the H-Y-doped CeO2, the O2-TPD profile was similar to that of the Undoped CeO2. However, the desorption bands at high temperatures occurred at higher temperatures, indicating that the partial substitution of Ce 4+ ions (0.97 Å ) with the large Y ions (1.02 Å ) improved the stability of the CeO2 lattice. For the O2-TPD profiles of I-Y-and H/I-Y-doped CeO2, the asymmetrical peak at high temperatures was displaced by a smooth descent peak, indicating that the oxygen desorption mainly occurred on the surface and subsurface of the I-Y-and H/I-Ydoped CeO2 samples. Furthermore, for the O2-TPD profiles of the H-Y-, I-Y-, and H/I-Ydoped CeO2 after 820 °C, there were still some distances to the baseline (light green area), indicating sustaining oxygen desorption, which could be attributed to the formation of extrinsic VO defects in the interior of the CeO2 lattice caused by Y-doping.   Table 1. Compared with the OSC value of the Undoped CeO 2 , all of the Y-doped CeO 2 samples were enhanced, and the sequence of [OSC] was as follows: H-Y-doped CeO 2 (0.372 mmol/g) > H/I-Y-doped CeO 2 (0.353 mmol/g) > I-Y-doped CeO 2 (0.248 mmol/g) > Undoped CeO 2 (0.153 mmol/g). The OSC of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 were increased by 143.1%, 62.1%, and 130.7%, respectively, compared with that of the Undoped CeO 2 . The enhanced OSC of all the Y-doped CeO 2 could be explained as follows: when Y 3+ ions were doped into the CeO 2 lattice to substitute Ce 4+ ions, the extrinsic V O defects were formed to keep the electric neutrality of their fluorite structure, accompanied by the increase in the number of oxidation/reduction cycles composed of Ce 3+ and Ce 4+ states (Ce 3+ ↔Ce 4+ ). Y-doping of CeO 2 possesses both intrinsic and extrinsic V O defects, as well as the oxidation/reduction cycle of Ce 3+ ↔Ce 4+ , which could determine the transfer of oxygen ions and OSC.

Conclusions
In summary, three various routes were adopted to successfully synthesize Y-doped CeO 2 solid solutions. The large Y cations were incorporated into the CeO 2 lattice with normal trivalence and formed a Y-doped CeO 2 solid solution while holding the cubic fluorite structure of CeO 2 . The results of O 1s XPS and Raman spectroscopy indicated that pure CeO 2 crystal itself contained a certain number of intrinsic V O defects. With the substitution of Ce 4+ ions with Y 3+ ions in the CeO 2 lattice, local lattice expansion of CeO 2 crystal occurred, extrinsic V O defects were formed, and there was an increase in the number of oxidation/reduction cycles composed of Ce 3+ and Ce 4+ states. Moreover, the relative concentrations of V O defects were quantified by the A 592 /A 464 values obtained from Raman spectra, which were 1.47, 0.93, and 1.16 for the H-Y-, I-Y-, and H/I-Y-doped CeO 2 , respectively, and were higher than that of the undoped one (0.67). There were large amounts of adsorbed oxygen on CeO 2 , which could emigrate quickly at low temperatures, and the OSCs of the H-Y-, I-Y-, and H/I-Y-doped CeO 2 were increased by 143.1%, 62.1%, and 130.7%, respectively, compared with that of the Undoped CeO 2 (0.153 mmol O 2 /g CeO 2 ). Both the intrinsic and extrinsic V O defects, as well as the oxidation/reduction cycle of Ce 3+ ↔Ce 4+ , could determine the enhanced OSC of Y-doped CeO 2 . The CeO 2 with doping during only the hydrothermal process exhibited the maximum values of OSC, suggesting the effectiveness of the doping.