Effect of Ce/Zr Composition on Structure and Properties of Ce1−xZrxO2 Oxides and Related Ni/Ce1−xZrxO2 Catalysts for CO2 Methanation

Ce1−xZrxO2 oxides (x = 0.1, 0.25, 0.5) prepared via the Pechini route were investigated using XRD analysis, N2 physisorption, TEM, and TPR in combination with density functional theory calculations. The Ni/Ce1−xZrxO2 catalysts were characterized via XRD analysis, SEM-EDX, TEM-EDX, and CO chemisorption and tested in carbon dioxide methanation. The obtained Ce1−xZrxO2 materials were single-phase solid solutions. The increase in Zr content intensified crystal structure strains and favored the reducibility of the Ce1−xZrxO2 oxides but strongly affected their microstructure. The catalytic activity of the Ni/Ce1−xZrxO2 catalysts was found to depend on the composition of the Ce1−xZrxO2 supports. The detected negative effect of Zr content on the catalytic activity was attributed to the decrease in the dispersion of the Ni0 nanoparticles and the length of metal–support contacts due to the worsening microstructure of Ce1−xZrxO2 oxides. The improvement of the redox properties of the Ce1−xZrxO2 oxide supports through cation modification can be negated by changes in their microstructure and textural characteristics.


Introduction
The CO 2 methanation process has received great interest during the last few years as a promising method of the production of synthetic natural gas as a hydrogen storage approach [1][2][3]. Nickel-based catalysts exhibit high activity and selectivity in the methanation of carbon oxides and are less expensive than systems containing noble metals [3][4][5]. Because of this, nickel catalysts are widely investigated. The impact of a support on the dispersion of nickel particles and catalytic performance of nickel catalysts is significant [6][7][8]. CeO 2 was shown to be one of the most effective support materials [6][7][8][9][10][11][12][13][14][15][16]. Ceria works as both a support for nickel particles and a reaction promoter. The promoting effect is related to the easiness and reversibility of Ce 4+ -Ce 3+ transition associated with appearing or healing oxygen vacancies [17]. Oxygen vacancies on the support surface were reported to participate in the activation of CO 2 molecules [14,[18][19][20][21][22].
The modification of ceria by doping with foreign cations is a well-known strategy to improve its redox properties [17]. Nickel catalysts supported on mixed Ce 1−x Zr x O 2 oxides are promising systems in terms of catalytic activity and stability [12,[23][24][25][26]. Zirconium is recognized as a promoter of the reducibility and oxygen storage capacity of ceria [27][28][29]. Our previous study [26] showed that the formation of oxygen vacancies on the support surface is essential for activity of Ni/Ce 1−x Zr x O 2 catalysts in the methanation of carbon oxides. The reducibility of Ce 1−x Zr x O 2 oxides was experimentally shown to increase with the Zr content [29,30]. The regulation of the support redox properties through the variation of Zr content is of great interest. However, there have been few studies on the effect of the Ce/Zr ratio on the performance of Ni/Ce 1−x Zr x O 2 catalysts in CO 2 methanation. Ocampo et al. [23] studied CO 2 methanation over 5 wt.% Ni/Ce 1−x Zr x O 2 catalysts differing in Zr content (x = 0.28, 0.5, 0.86). The 5 wt.% Ni/Ce 0.5 Zr 0.5 O 2 catalyst showed the highest activity. Nie et al. [31] studied NiO-CeO 2 -ZrO 2 mixed oxides containing 40 wt.% Ni. The catalyst with a Ce/Zr molar ratio of 9:1 (Ce 0.9 Zr 0.1 O 2 ) exhibited the best catalytic properties. Atzori et al. [32] tested NiO-CeO 2 -ZrO 2 catalysts (30 wt.% Ni-Ce 1−x Zr x O 2 ) in CO 2 and CO co-methanation. The activity was the same for the catalysts with Zr content in x = 0-0.5 and decreased at a higher content. To summarize, a correlation between the activity of Ni/Ce 1−x Zr x O 2 catalysts and the Ce/Zr ratio has not been fully understood. All the reported studies indicated that nickel catalysts based on highly doped Ce 1−x Zr x O 2 oxides (x > 0.5) are less effective. The data concerning systems based on Ce 1−x Zr x O 2 oxides with a lower Zr content (x ≤ 0.5) are somewhat controversial.
The present work aims to provide further insight into the relation between the composition of the Ce

Computational Methods and Details
The DFT+U calculations were performed using the Vienna Ab initio Simulation Package (VASP) program [33]. The generalized gradient approximation (GGA) PBE96 functional was employed [34]. The core electrons were described by projector augmented-wave (PAW) potentials [35,36], and the valence electrons were described by a plane-wave basis set. The DFT+U method (U eff = 5 eV) [37] was used to correct the strong Coulomb repulsion of cerium and DFT-D3 [38] to take into account dispersion corrections. The cutoff energy was 400 eV. A 2 × 2 × 1 Monkhorst-Pack k-point set was used for the Brillouin-zone integration.
The face-centered cubic unit cell of a fluorite-type structure (space group: Fm 3 m) was used as the initial geometry in the calculations [39,40]. Ce 1−x Zr x O 2 supercells (x = 0.25, 0.50, 0.75) were built by replacing Ce atoms with Zr ones (Supplementary Materials, Figure S1). Ce 1−x Zr x O 2 (100) and (111) surfaces were modeled as p (2 × 2) and p (3 × 3) slab cells. The (100) and (111) Ce 1−x Zr x O 2 supercells contained four and three layers, respectively. A vacuum space of 15 • Å was set between neighboring slabs to keep the spurious interaction.
The energy of oxygen vacancy formation (E f ) was computed as: The catalysts containing 10 wt.% Ni were prepared by the impregnation of the obtained Ce 1−x Zr x O 2 oxide materials. Metal precursor Ni(CH 3 COO) 2 ·4H 2 O (99.0%) and ethylene glycol (99.5%) were dissolved in the distilled water under stirring at 70 • C for 20 min. Then, the support material was added into the solution. The EG:Ni molar ratio was set to 5.3. The suspension of the support and impregnating solution was stirred for 2 h at 70 • C and dried at 120 • C for 12 h in air. The dried samples were heated at a rate of 2 • C/min and calcined in air at 400 • C for 2 h. The catalysts were designated as Ni/Ce 1−x Zr x O 2 , with x indicating Zr content.

XRF Analysis
Elemental compositions of the catalysts were determined via X-ray fluorescent spectroscopy (XRF) using an ARL-Advant'x device (Thermo Fisher, Vienna, Austria). Measurements were carried out in a helium atmosphere using the Rh X-Ray tube. UniQuant software was used to calculate the element percentages.

BET Surface Area Analysis
The BET specific surface areas (S BET , m 2 /g) of the Ce 1−x Zr x O 2 oxides were determined by N 2 -physisorption at −196 • C. The experiments were carried out using an ASAP 2400 instrument (Micrometrics, Norcross, GA, USA).

XRD Analysis
X-ray diffraction (XRD) measurements were carried out using a D8 Advance diffractometer (Bruker, Germany) equipped with a Lynxeye linear detector. The measurements were carried out using the Cu Kα radiation (λ = 1.5418 Å) in the 2θ range of 10-83 • with a step of 0.05 • . XRD phase analysis was performed using the ICDD PDF-4+ database. The evaluation of substructure parameters of the Ce 1−x Zr x O 2 oxides was performed. The separation of the crystallite size (D XRD ) and microstrain (∆d/d) effects to the line broadening was performed by means of Williamson-Hall plots [42]. From the D XRD values, the crystallite surface area (S XRD ) values were calculated assuming the crystallites were quasi-spherical: where ρ is the theoretical density of the material (g/cm 3 ).
To estimate the degree of agglomeration of the Ce 1−x Zr x O 2 crystallites, an agglomeration coefficient (ξ) was calculated on the basis of the ratio of S BET and S XRD values: The average crystallite size of nickel-containing phases was estimated by the line broadening analysis according to the Scherrer equation [43]. The Rietveld refinement was carried out using the software package Topas v.4.2 (Bruker-AXS, Karlsruhe, Germany).

CO Pulse Chemisorption
The average sizes of metallic Ni 0 nanoparticles in catalysts after catalytic experiments were determined using the pulse chemisorption of CO on the assumption that each surface Ni atom adsorbed one CO molecule. The measurements were carried out using a Chemosorb analyzer (Modern Laboratory Equipment, Novosibirsk, Russia). An amount of 50 mg of each sample was placed inside a U-shape quartz reactor and reduced at 350 • C in H 2 flow (100 mL/min) for 30 min. The treated sample was subsequently cooled down to room temperature, followed by Ar purge. After that, pulses of CO were fed to the reactor (100 µL) until the amount of CO in the outlet stopped changing according to the thermal conductivity detector. The amount of chemisorbed CO was estimated. CO adsorption over the pure supports was negligible.

TEM-EDX
Transmission electron microscopy (TEM) studies were carried out using a JEM-2200FS (JEOL, Tokyo, Japan) and a Themis Z electron microscope (Thermo Fisher Scientific, Eindhoven, The Netherlands) operated at 200 kV. Images in Scanning-TEM (STEM) mode were acquired using a high-angle annular dark field (HAADF) detector. The local elemental composition of the samples was studied using a Thermo Fisher Scientific Super-X EDX spectrometer. The samples were ground, suspended in ethanol, and placed on a copper grid coated with a holey carbon film.

SEM-EDX
Scanning electron microscopy-energy-dispersive X-ray analysis (SEM-EDX) studies were carried out using a dual-beam scanning electron microscope, Tescan Solaris FE/SEM (Tescan, Brno, Czech Republic). The experiments were performed in secondary electron mode at an accelerating voltage of 20 kV. The microscope is equipped with an AztecLive EDX spectrometer (Oxford Instruments, High Wycombe, UK) with a Silicon Drift Detector and energy resolution of 128 eV. The cross-sections of catalyst granules in epoxy resin were prepared for the examination. The cross-sections with a given flatness of 0.25 µm were covered with a conductive carbon layer of 10-20 nm thickness.

H 2 -TPR
The temperature-programmed reduction (TPR) via hydrogen was performed with 40-60 mg of sample in a quartz reactor using a flow setup with a thermal conductivity detector. The gas mixture containing 10 vol.% of H 2 in Ar was fed at 40 mL/min. The rate of heating from 25 to 800 • C was 10 • C/min. The TPR curves were normalized per sample mass.

Tests of CO 2 Methanation Activity
The catalytic tests were performed in a U-shaped tubular continuous-flow reactor (i.d. = 3 mm) at ambient pressure and the temperature range from 200 to 400 • C. The temperature was controlled using a K-type thermocouple placed in the middle of the catalyst bed. The feed gas contained 4 vol.% CO 2 , 16 vol.% H 2 , and Ar as balance. The catalyst load was 150 mg, 0.2-0.5 mm fraction, and the flow rate-75 mL/min. Prior to the experiment, each catalyst was reduced in 10 vol.% H 2 in Ar flow (50 mL/min) at 400 • C for 1 h. The compositions of the inlet and outlet gas mixtures were determined using a gas chromatograph, KHROMOS-1000 (Khromos, Dzerzhinsk, Russia), equipped with a thermal conductivity detector (CaA molecular sieves column) and flame ionization detector (Porapak Q column) with a methanator characterized by sensitivity to CO, CH 4 , and CO 2 of~1 ppm. The separation of CO, CH 4 , and CO 2 on the column followed by the methanation of carbon oxides allowed the flame ionization detector to be used to analyze their concentration. The equilibrium compositions were calculated using equilibrium software HSC 7.0. It was assumed that equilibrium mixtures only contained gaseous substances (CH 4 , CO, CO 2 , H 2 , and H 2 O), i.e., no carbon deposition processes were taken into account. The catalytic properties of Ni/Ce 1−x Zr x O 2 catalysts were also compared with those of industrial catalyst NIAP-07-05 for the methanation of carbon oxides, which contains 38 wt.% NiO, 12 wt.% Cr 2 O 3 , and 50 wt.% Al 2 O 3 (further denoted as 38Ni−Cr−Al). The catalyst has a Ni 0 -specific surface area of 3.3 m 2 /g in a reduced state [26].

Calculation Results
According to DFT+U calculations, the parameter of the Ce 1−x Zr x O 2 lattice decreases with an increase in Zr content (Table 1)  The calculated values of the surface energy are presented in Table 1. The Ce 1−x Zr x O 2 (111) surface is more stable than the (100) surface. The CeO 2 (111) surface is known to be the most stable surface among the low-index surfaces (100), (110), and (111) [44,45]. Zirconium incorporation slightly increases the surface energy at 0 < x < 0.5, while a sharp increase in the energy is observed at x > 0.5.
The calculated energies of oxygen vacancy formation in Ce 1−x Zr x O 2 are also listed in Table 1. The E f values decrease considerably in the range of compositions 0 < x < 0.5, while at x > 0. 5, the E f values increase. The Ce 0.5 Zr 0.5 O 2 surfaces are characterized by the lowest energies of oxygen formation. It was found that Zr incorporation facilitates the generation of oxygen vacancy. The obtained result agrees with data of interatomic potential simulations reported by G. Balducci et al. [46,47], which showed that Ce 4+ /Ce 3+ reduction energy is reduced even by small amounts of zirconium incorporated into ceria. Based on DFT+U calculations, Yang et al. [48,49] also found that zirconium addition leads to lowering energy of the oxygen vacancy formation. The structure distortion induced by Zr cations can be responsible for the decrease in the reduction energy. It was assumed that the smaller Zr 4+ cations counterbalance steric strains arising at formation Ce 3+ cations, which are larger than Ce 4+ cations [46,49].

Ce 1−x Zr x O 2 Support Materials
The powder XRD patterns of the Ce 1−x Zr x O 2 samples are shown in Figure 1. Only Bragg peaks related to oxide with a fluorite-type cubic crystal structure (S.G. Fm 3 m) are observed. Table 2 lists data on the structural parameters obtained by Rietveld refinement. The lattice parameters of all the oxides are lower than the value characteristic of CeO 2 (a = 5.411 Å, PDF# 00-028-0753). The lattice shrinkage indicates the formation of substitutional solid solution Ce 1−x Zr x O 2 . The gradual decrease in the lattice parameter with the increase in Zr content is observed. The composition of the Ce 1−x Zr x O 2 solid solutions was evaluated with use of Vegard's rule. The linear dependence of the lattice parameter on the Zr content is provided elsewhere [26]. Estimated values of zirconium content coincide with those set at the synthesis ( Table 2). This implies that obtained materials are single-phase Ce 1−x Zr x O 2 solid solutions. As can be seen from Table 2, the isotropic temperature factors of atoms increase with Zr content. This suggests the increase in crystal lattice distortion resulted from Zr incorporation.  Microstrain analysis also indicates the intensification of structure deformation with an increasing Zr concentration in Ce 1−x Zr x O 2 oxides (Table 3). The microstrain value for the Ce 0.5 Zr 0.5 O 2 sample is roughly two times higher than that for the Ce 0.9 Zr 0.1 O 2 sample. A difference in the radii of Ce 4+ and Zr 4+ cations is the main reason for the crystal lattice distortion. As mentioned above, the structure strains induced by Zr doping can be responsible for the improved reducibility of the Ce 1−x Zr x O 2 mixed oxides. Table 3. Average crystallite sizes and microstrain values according to XRD data, specific surface areas calculated from XRD-derived crystallite sizes and determined by BET method, agglomeration coefficients, and average crystallite sizes according to HRTEM data. All the oxides are highly dispersed; the D XRD are in the range of 5-7 nm. However, quite low S BET values were obtained. Calculated S XRD values significantly exceed S BET ones (Table 3). Such a difference between S XRD and S BET values can be explained by the agglomeration of primary crystallites. The analysis of estimated agglomeration coefficients (Table 3) showed that an increase in Zr content in the Ce 1−x Zr x O 2 oxides is accompanied with an increase in the agglomeration of crystallites. TEM data confirmed these results. The TEM images shown in Figure 2a  The reduction behavior of the Ce 1−x Zr x O 2 oxides was studied. The H 2 -TPR profiles are reported in Figure 3. The H 2 -TPR profile of the low-doped Ce 0.9 Zr 0.1 O 2 oxide exhibits two broad peaks at 500 • C and 820 • C. Such a two-peak pattern is characteristic of CeO 2 oxide and reflects the stepwise reduction in the surface and bulk [50,51]. The TPR profiles of the Ce 0.75 Zr 0.25 O 2 and Ce 0.5 Zr 0.5 O 2 oxides exhibit asymmetric peaks with significantly increased intensities. The second high-temperature reduction peak is not observed. These data indicate that surface and bulk reduction processes occur almost simultaneously. The observed intensive asymmetric peak reflects the co-reduction in surface and bulk. This suggests the improvement of the reducibility of the mixed Ce 1−x Zr x O 2 oxides with an increase in Zr content in full agreement with previous TPR studies of Ce 1−x Zr x O 2 oxides [28][29][30].  Table 4 summarizes the average size characteristics of nickel species in the as-prepared and used Ni/Ce 1−x Zr x O 2 catalysts according to the XRD and CO chemisorption data. The Ni/Ce 0.9 Zr 0.1 O 2 catalyst is characterized by a higher dispersion of initial NiO crystallites as well as Ni 0 crystallites formed upon reduction. Reaction conditions provoke the sintering of nickel species. The average size of Ni 0 crystallites in the aged catalysts is larger than the size of NiO crystallites in the as-prepared catalysts. The Ni/Ce 0.9 Zr 0.1 O 2 catalyst is characterized by a higher resistance of Ni 0 crystallites to sintering. The average size of Ni 0 crystallites in the Ni/Ce 0.9 Zr 0.1 O 2 catalyst is about two times smaller than in other catalysts. CO chemisorption results (Table 4)    The HAADF STEM images and EDX mapping patterns are presented in Figure 5. The analysis of EDX data revealed that the spatial distribution of Ce and Zr in the catalysts is homogeneous. No Ce-rich or Zr-rich areas are observed. This confirms that Ce 1−x Zr x O 2 oxides used as supports are single-phase substitutional solid solutions. The analysis of Ni distribution via EDX allows us to see Ni-rich nanoparticles of 5-10 nm in size in the catalysts. These particles in all the studied samples were shown to be NiO particles (Supplementary material Figure S2). Individual NiO nanoparticles in contact with support particles and ones assembled into large aggregates are observed. The Ni/Ce 1−x Zr x O 2 catalysts were found to differ in the amount of non-agglomerated NiO nanoparticles and the possibility of their fixation as single particles. Thus, a large number of single NiO particles being in contact with support particles are observed in the images of the Ni/Ce 0.9 Zr 0.1 O 2 catalyst (Figure 5b). More developed contacts between NiO particles and Ce 0.9 Zr 0.1 O 2 support are likely responsible for the higher resistance of nickel particles to sintering under conditions of CO 2 methanation. In the case of the Ni/Ce 0.5 Zr 0.5 O 2 catalyst, agglomerated NiO nanoparticles with slight contact with the support are observed (Figure 5h). The Ni particles undergo reduction and sintering during the reaction, as revealed by XRD analysis. However, the highly dispersed particles fixed on the support are retained (Supplementary material Figure S3). The observed differences in the dispersion of nickel species in the catalysts seem to be related to effect of the microstructure of the Ce 1−x Zr x O 2 oxides on the formation of supported nanoparticles. As noted above, the Ce 1−x Zr x O 2 support materials differed in particle organization and the agglomeration of crystallites. The microstructure features of the catalysts were additionally studied via SEM coupled with EDX analysis. SEM analysis showed quite different internal organization and porous structure in grains of the Ni/Ce 0.9 Zr 0.1 O 2 and Ni/Ce 0.5 Zr 0.5 O 2 catalysts. In the Ni/Ce 0.5 Zr 0.5 O 2 catalyst, coarse elongated support particles are aggregated with the formation of large, unevenly distributed voids (Figure 6a,b). Large NiO particles are visualized in the big voids, while the main part of the support material is weakly covered with nickel compounds according to EDX mapping (Figure 6c,f). In the Ni/Ce 0.9 Zr 0.1 O 2 catalyst, aggregates of smaller and thinner support particles have a more uniform distribution of narrow interparticle voids (Figure 7a,b). EDX mapping showed the more homogeneous distribution of nickel over the Ce 0.9 Zr 0.1 O 2 support (Figure 7c,f) than over the Ce 0.5 Zr 0.5 O 2 support (Figure 6c,f). It appears that a network of narrow channels provided a more uniform distribution of the nickel compounds over the support surface in the impregnation step. As a result, the support is effectively covered with nickel compounds and there is no pronounced gradient in the size of formed NiO particles. Thus, the SEM-EDX results allowed one to explain the observed differences in Ni/Ce 1−x Zr x O 2 catalysts in the dispersion of nickel species. The microstructure features of the Ce 1−x Zr x O 2 oxides affected the dispersion of nickel compounds as well as their spatial distribution over the supports. The microstructure of the Ce 0.9 Zr 0.1 O 2 support is more beneficial for the uniform distribution of the nickel compounds without their segregation.
The formation of larger particles in the Ce 0.5 Zr 0.5 O 2 oxide seems to be caused by the considerable aggregation of the constituent crystallites (Table 3, Figure 2). On the one hand, greater aggregation is likely related to the higher dispersion of crystallites. Ultrafine crystallites assemble to lower the surface energy [52,53]. On the other hand, differences in the microstructure can be related with specifics of the formation of Ce 1−x Zr x O 2 oxides at their preparation via the Pechini route. Thus, the intensity of combustion processes when burning the organic polymer matrix depends on the relative contents of Ce and Zr atoms. During the preparation of Ce 0.9 Zr 0.1 O 2 oxide, a higher content of easily oxidized Ce cations intensifies combustion processes. The explosive formation of gaseous products favors the formation of loose particles with the low agglomeration of crystallites. The catalytic performance of the Ni/Ce 1−x Zr x O 2 catalysts in CO 2 methanation was studied. Figure 8 shows light-off curves for the CO 2 methanation. For all the catalysts, the CH 4 concentration increases with temperature, reaches a maximum, and then decreases coinciding with the equilibrium values. All the Ni/Ce 1−x Zr x O 2 catalysts had comparable or higher activity in comparison with the industrial 38Ni-Cr-Al catalyst containing a significantly higher amount of nickel (38 wt.% vs. 10 wt.%). In a previous work, we reported that the synergism of redox properties of the Ni/Ce 1−x Zr x O 2 system enhances the catalytic performance in the methanation of carbon oxides [26].   (Table 4). It was also revealed that the most active Ni/Ce 0.9 Zr 0.1 O 2 catalyst is characterized by more developed contacts between nickel particles and the support (Figure 5b). It is well accepted that high Ni 0 dispersion and a developed metal-support interface area are highly important for the catalytic activity for CO 2 methanation [6,8,23,54,55].
It was recently suggested that zirconium addition impacts the dispersion of nickel nanoparticles in Ni/Ce 1−x Zr x O 2 catalysts as well as the degree of metal-support interaction [23,31]. The results of this study imply that the cation composition of the Ce 1−x Zr x O 2 oxides can affect their microstructure and texture features and, consequently, the dispersion of supported nickel nanoparticles. This influence is likely related to the synthesis technique used. The Ce 1−x Zr x O 2 oxides under study were prepared via the Pechini method. A similar effect was observed by Iglesias et al. [24] in the study of Ni/Ce 1−x Zr x O 2 catalysts with supports prepared using the coprecipitation method. It was shown that an increase in Zr content diminishes the specific surface area of the Ce 1−x Zr x O 2 supports with a decrease in the dispersion of supported Ni 0 nanoparticles. It was also reported that the increase in Zr content reduces the specific surface area of the Ce 1−x Zr x O 2 oxides prepared via the sol-gel technique [56] and the citrate complexation route [57].

Conclusions
In this study, Ce 1−x Zr x O 2 mixed oxides with different compositions were prepared using the Pechini method and used as the supports for Ni/Ce 1−x Zr x O 2 catalysts. It was demonstrated that an increase in Zr content enhances the distortion of the crystal structure of Ce 1−x Zr x O 2 oxides and leads to improvements in their redox properties. It was also found that microstructure features of Ce 1−x Zr x O 2 oxides change with cation composition.
The effect of the Ce 1−x Zr x O 2 composition on the structure and catalytic activity of the Ni/Ce 1−x Zr x O 2 catalysts in CO 2 methanation was investigated. The activity of the Ni/Ce 1−x Zr x O 2 catalysts decreased in the order: Ni/Ce 0.9 Zr 0.1 O 2 >> Ni/Ce 0.75 Zr 0.25 O 2 > Ni/Ce 0.5 Zr 0.5 O 2 . The drop in the activity correlated with the decrease in the dispersion of metallic Ni 0 nanoparticles. It was revealed that differences in the microstructural characteristics of the Ce 1−x Zr x O 2 supports are responsible for differences in the dispersion of supported Ni 0 nanoparticles and the length of the metal-support interface.
It was shown that improving the redox properties of Ce 1−x Zr x O 2 oxides, which are important for catalysis, through cation modification can be counterbalanced by worsening their microstructure characteristics, which determine the dispersion of supported Ni 0 nanoparticles.