Functional Improvement of NiO x /CeO 2 Model Catalyst Active in Dry Methane Reforming via Optimization of Nickel Content

: The valorization of greenhouse gases, especially when focused on carbon dioxide, currently belongs to the main challenges of pro-environmental chemical processes. One of the important technologies in this field is dry methane reforming (DMR), leading to the so-called synthesis gas (CO + H 2 ). However, to be efficient and economically viable, an active and stable catalyst is required. Ni-based systems can be recommended in this regard. This research aimed to investigate how nickel content can influence the activity of model NiO x /CeO 2 catalysts in DMR. A series of NiO x /CeO 2 samples of various nickel loadings (0–10 wt.%) were prepared through dry impregnation. The obtained samples were characterized through XRD, RS, N 2 -BET, DRIFT, SEM, UV/Vis-DR, and XPS. Nonlinear changes in surface properties of the investigated samples with increasing nickel concentration were found. The observed changes are mirrored both in the determined nickel speciation and in the corresponding catalytic activity. The highest activity was found for the catalyst containing 3 wt.%. of nickel.


Introduction
Over the last 100 years, energy consumption has rapidly grown as the world population has skyrocketed, and technological and economic progress have significantly intensified.Moreover, the common use of cars and planes along with engine-powered trains and ships has made us extremely dependent on hydrocarbon-based fuels, such as gasoline, diesel, and jet fuels.The combustion of solid, liquid, and gaseous fossil fuels in various mobile and stationary energy-generated systems entails many environmental problems, including those related to the emission of greenhouse gases (GHGs).An increase in global environmental awareness boosted a common trend to reduce the emission of GHGs, especially CO 2 [1][2][3].
One of the most interesting technological ideas of CO 2 utilization is its conversion into a mixture of hydrogen and carbon monoxide, commonly called synthesis gas.This concept was originally described in 1928 by F. Fischer and H. Tropsch, who proposed various transition metals as catalysts, including Ni-and Co-based systems [4].Synthesis gas can be obtained in the reaction (1) described in the literature as dry reforming of methane (DMR): CH 4(g) + CO 2(g) = 2CO (g) + 2H 2(g) , (1) Due to quite demanding process conditions, there was no technological interest in the industrial implementation of this process for several decades, but, nowadays, environmental reasons and stricter regulations imposed on emitters make methane reforming relatively little attention was paid in the previously published papers to the effect of active metal concentration on its catalytic performance in DMR.In the literature devoted to dry methane reforming, only a few researchers have investigated this issue in more detail, as it can be inferred from the papers [39][40][41][42][43].Despite this fact, it is clearly seen from the literature mentioned above that the loading of active metal may be decisive in obtaining a catalyst showing both high activity and satisfactory stability.
This work was focused on the investigation of the nickel loading effect on the structure and catalytic properties of a series of model NiO x /CeO 2 catalysts for dry methane reforming.The reported results may constitute useful guidelines for the structural optimization of the supported Ni-based DMR catalysts.

Materials and Methods
The CeO 2 support was synthesized through the Pecchini method from cerium(III) nitrate hexahydrate (Aldrich, St. Louis, MO, USA), citric acid monohydrate (POCh, Gliwice, Poland), and ethylene glycol (Fisher, Hampton, NH, USA) in a molar ratio of 1:1.1:2.All reactants used for synthesis were of per analysis purity.Cerium(III) nitrate was dissolved in a minimal amount of water, and then citric acid was added to the mixture and the stirring was continued.After the total dissolving of citric acid, ethylene glycol was added.The resulting solution was mixed and heated on a magnetic stirrer until a gel was formed.The obtained residue was dried overnight at 120 • C and then calcined in air at 800 • C for 6 h.Final catalysts were prepared through impregnation of ceria support with aqueous solutions of nickel nitrate (Sigma-Aldrich, St. Louis, MO, USA) containing appropriate concentrations to obtain nickel contents in the final preparations ranging from 0 to 10 wt.%.A sample with nickel loading of x wt.% can be labeled hereafter as xNiO x /CeO 2 .After impregnation, all samples were dried overnight at 120 • C and then calcined at 400 • C in the air for 3 h to decompose the precursors.Such a procedure is similar to those applied in [44][45][46].The NiO reference sample was synthesized through precipitation from a nickel(II) nitrate solution using a 25% aqueous solution of ammonia, filtered, dried overnight at 120 • C, and then calcined in air at 400 • C for 3 h.
The structure of the obtained samples was verified through X-ray diffraction (XRD).For data collection, a PANalytical (Malvern, UK) X'pert Pro diffractometer equipped with a PW3050/60 goniometer and CuKα lamp (λ = 1.5406Å) was used.The anode voltage was set to 40 kV, while the anode current was set to 30 mA.The diffraction patterns were measured from 10 • to 90 • with a step of 0.026 • .The size of crystallites was estimated based on the Scherrer equation [47]: where K is a dimensionless shape factor (0.93); λ is the X-ray wavelength (0.15406 nm); β is the line broadening at half the maximum intensity (FWHM); and θ is the Bragg angle.Further studies of the structure were performed through micro-Raman spectroscopy.A Renishaw (Wotton-under-Edge, UK) InVia spectrometer equipped with a Leica DMLM confocal microscope and a CCD detector, with an excitation wavelength of 785 nm, was used to record RS spectra under ambient conditions.The power of the laser was 1.5 mW (0.5% of the total power) with a magnification of 20 times.Data collection was performed in the range of 150 to 800 cm −1 .At least 9 scans (15 s each) were recorded.The Brunauer-Emmett-Teller (BET) method was applied for the surface area measurements of the investigated samples.To do so, 3Flex chemisorption by Micrometrics (Unterschleißheim, Germany) using nitrogen adsorption (−196 • C) was used for the measurements.The mean particle size estimated from BET (d BET ) was calculated from the Equation (5): where ρ is the theoretical density of CeO 2 and S BET is the surface area estimated by the BET method [48].A JEOL (Peabody, MA, USA) FEG 6700F microscope working at 9 kV accelerating voltage was employed for scanning electron microscopy (SEM) images.A Bruker (Billerica, MA, USA) Tensor 27 spectrometer equipped with a Specac (Orpington, UK) DRIFT Environmental Chamber, a temperature controlling unit, and a He-Ne laser (λ = 633 nm) was used to collect Diffuse Reflectance Fourier Transform IR (DRIFT) spectra.The measurements were performed in a vacuum (10 −6 bar) in the frequency range of 4000-600 cm −1 with a resolution of 2 cm −1 .Each spectrum was measured with 256 accumulations to obtain sufficiently intense spectra.Nickel speciation was investigated through UV/Vis diffuse reflectance spectroscopy employing a Perkin-Elmer (Waltham, MA, USA) Lambda 650 UV/Vis spectrophotometer with Praying Mantis (Harrick, Pleasantville, NY, USA).Each spectrum was recorded in the range of 200-900 nm with a resolution of 1 nm under ambient conditions, and the data were transformed according to the Kubelka-Munk equation.Deconvolution of the spectra was carried out using the Fityk software (v.0.8.0).
X-ray photoelectron spectroscopy (XPS) was employed for further surface investigation.The XPS system PHI (Lake Drive East Chanhassen, MN, USA) VersaProbeII was used for data collection.This scanning XPS system uses monochromatic Al K α (1486.6 eV) X-rays focused on a 100 µm spot.The scan is performed over the sample area of 400 µm × 400 µm.The photoelectron take-off angle was set to 45 • , and the pass energy in the analyzer was set to 23.50 eV to obtain high-energy resolution spectra for the C 1s, O 1s, Ni 2p, and Ce 3d regions.To maintain a constant sample surface potential regardless of the sample conductivity, a dual beam charge compensation with 7 eV Ar + ions and 1 eV electrons was used.The C1s' peak of unfunctionalized, saturated carbon (C-C) at 284.8 eV was used to charge-reference all XPS spectra.The pressure in the analytical chamber was below 4•10 −9 mbar.PHI MultiPak software (v.9.7.0.1) was employed for the deconvolution of spectra.The Shirley method was used for spectrum background subtraction.
Catalytic activity in DMR was measured as follows: a mixture of gases with a total flow of 100 mL/min (5% CO 2 /He = 50 mL/min, 5% CH 4 /He = 50 mL/min) was introduced into a quartz reactor containing 100 mg of a catalyst.Reagents were analyzed using a gas chromatograph (Agilent Technologies, Santa Clara, CA, USA, 7890A GC System).In the first step, a catalyst was reduced at 850 • C with 5% H 2 /Ar flow (50 mL/min) for 120 min.After reaching a temperature of ca.550 • C, the H 2 flow was switched off, and other gaseous reactants were turned on.Gas chromatographic measurements were performed from 600 • C every 50 • C up to 800 • C and one back at 650 • C. The conversion of CO 2 and CH 4 , as well as the H 2 /CO ratio, were calculated by comparing the concentrations of these reactants before and after the catalytic reactor, similarly to [49][50][51].

Structural Characterization
As it can be seen in Figure 1a, the structure of the synthesized CeO 2 support was crystalline, giving rise to a regular phase with characteristic Bragg reflections (JCPDS card no.34-0394), and its impregnation with aqueous solutions of nickel(II) nitrate of various concentrations followed by calcination at 400 • C did not influence the phase composition of the studied support.The average size of the support crystallites estimated using the Scherer method was ca.43 nm.The lack of a broadening visible in the recorded XRD patterns suggested that independently of its nominal content, nickel was not incorporated into the bulk of CeO 2 and formed separate surface entities, which can be identified as NiO nanocrystals (JCPDS Card No. 00-004-0835), giving rise to the maxima at 37.2 • , 43.3 • , and 63.2 • (more visible in Figure 1b), attributed to (021), (202), and (024), respectively, which are reflections characteristic of nickel(II) oxide.They can be observed in the XRD patterns recorded for samples containing more than 3 wt.% of nickel, and their intensity increased with nickel loading.nanocrystals (JCPDS Card No. 00-004-0835), giving rise to the maxima at 37.2°, 43.3°, and 63.2° (more visible in Figure 1b), attributed to (021), (202), and (024), respectively, which are reflections characteristic of nickel(II) oxide.They can be observed in the XRD patterns recorded for samples containing more than 3 wt.% of nickel, and their intensity increased with nickel loading.Any changes in the recorded Raman spectra as a consequence of nickel deposition on ceria cannot be identified (Figure 1c).Only an intense band at about 460 cm −1 , attributed to the characteristic vibrations of F2g symmetry in the CeO2 regular phase [52], dominated in all recorded spectra.The lack of a broad band at ca. 590 cm −1 , usually attributed to the vibrations associated with oxygen vacancies induced in the ceria structure by nickel incorporation [53], suggested that nickel was stabilized mainly on the surface.That remains in agreement with the XRD results reported previously.Bulk NiO was not detected through Raman spectroscopy, which can be explained by a good dispersion as a possible result of relatively strong interactions between the active phase and the support [53].

Morphology
The specific surface of CeO2 support determined through the N2-BET method was found to be rather low and equal to 5.37 m 2 /g.It slightly increased after the deposition of nickel to reach specific surface areas (SSA), from 6.02 m 2 /g for 3NiOx/CeO2 to 8.42 m 2 /g for 10NiOx/CeO2.Partial segregation of the NiO phase at the highest loading is thus possible.The particle sizes of the investigated samples estimated based on the BET results were distinctly higher than those attributed to crystallites of diameters determined based on the XRD results (155 nm from BET and 42 nm from Scherrer for CeO2).This suggests that one particle was built from around 60 crystallites.Although all surface areas reported for the investigated supported samples increased with growing nickel contents, their micropore areas changed in a rather nonlinear way.As such, the micropores took 30% of the CeO2 support area (1.60 m 2 /g) and 15% of the area in the case of the 10NiOx/CeO2 sample (1.30 m 2 /g), whereas they took only 11% of 3NiOx/CeO2 area (0.66 m 2 /g).At the same time, the external surface areas consequently increased from 3.77 m 2 /g (for CeO2) to 5.35 m 2 /g (for 3NiOx/CeO2) to 7.12 m 2 /g (for 10NiOx/CeO2).Based on that, it can be stated that during impregnation, some of the pores became partially filled, and the segregation of NiO led to an increment in the surface area and was responsible for the formation of new pores within NiO nanocrystals.In any case, the volume of the micropores was relatively low and equal to 0.00103 cm 3 /g (for CeO2), 0.00036 cm 3 /g (for 3NiOx/CeO2), and 0.00046 cm 3 /g (for 10NiOx/CeO2).This remains in agreement with Figure 2, in which only narrow hystereses on the type V isotherms can be observed [54].Any changes in the recorded Raman spectra as a consequence of nickel deposition on ceria cannot be identified (Figure 1c).Only an intense band at about 460 cm −1 , attributed to the characteristic vibrations of F 2g symmetry in the CeO 2 regular phase [52], dominated in all recorded spectra.The lack of a broad band at ca. 590 cm −1 , usually attributed to the vibrations associated with oxygen vacancies induced in the ceria structure by nickel incorporation [53], suggested that nickel was stabilized mainly on the surface.That remains in agreement with the XRD results reported previously.Bulk NiO was not detected through Raman spectroscopy, which can be explained by a good dispersion as a possible result of relatively strong interactions between the active phase and the support [53].

Morphology
The specific surface of CeO 2 support determined through the N 2 -BET method was found to be rather low and equal to 5.37 m 2 /g.It slightly increased after the deposition of nickel to reach specific surface areas (SSA), from 6.02 m 2 /g for 3NiO x /CeO 2 to 8.42 m 2 /g for 10NiO x /CeO 2 .Partial segregation of the NiO phase at the highest loading is thus possible.The particle sizes of the investigated samples estimated based on the BET results were distinctly higher than those attributed to crystallites of diameters determined based on the XRD results (155 nm from BET and 42 nm from Scherrer for CeO 2 ).This suggests that one particle was built from around 60 crystallites.Although all surface areas reported for the investigated supported samples increased with growing nickel contents, their micropore areas changed in a rather nonlinear way.As such, the micropores took 30% of the CeO 2 support area (1.60 m 2 /g) and 15% of the area in the case of the 10NiO x /CeO 2 sample (1.30 m 2 /g), whereas they took only 11% of 3NiO x /CeO 2 area (0.66 m 2 /g).At the same time, the external surface areas consequently increased from 3.77 m 2 /g (for CeO 2 ) to 5.35 m 2 /g (for 3NiO x /CeO 2 ) to 7.12 m 2 /g (for 10NiO x /CeO 2 ).Based on that, it can be stated that during impregnation, some of the pores became partially filled, and the segregation of NiO led to an increment in the surface area and was responsible for the formation of new pores within NiO nanocrystals.In any case, the volume of the micropores was relatively low and equal to 0.00103 cm 3 /g (for CeO 2 ), 0.00036 cm 3 /g (for 3NiO x /CeO 2 ), and 0.00046 cm 3 /g (for 10NiO x /CeO 2 ).This remains in agreement with Figure 2, in which only narrow hystereses on the type V isotherms can be observed [54].SEM images (Figure 3) of the investigated catalysts showed a wide range of grain sizes.Generally, grain diameters measured just a few micrometers; however, larger aggregates with dimensions higher than 20 µm were also observed.Any notable changes with increasing nickel loading can be observed.

Surface Characterization
The analysis of hydroxyls that presented on the surface of the investigated samples can be seen in Figure 4. Regarding CeO2 support, four hydroxyl bands can be detected: a single bonded hydroxyl group (type I) as a shoulder around 3700 cm −1 and two types of hydroxyls bicoordinated (type IIA) to Ce(III) centered at 3690 cm −1 and Ce(IV) centered at 3655 cm −1 .The last peak can be ascribed to hydroxyl coordinated to two cations localized in the vicinity of an oxygen vacancy (type IIB) to Ce(IV) centered at 3620 cm −1 [55,56].It is noteworthy that even small amounts of nickel led to the disappearance of the bands attributed to hydroxyl groups mentioned above, with the simultaneous appearance of the band at 3680 cm −1 , which is typical of hydroxyls bonded to nickel(II) centers.Such an interpretation was confirmed by our reference DRIFT measurement carried out for the NiO SEM images (Figure 3) of the investigated catalysts showed a wide range of grain sizes.Generally, grain diameters measured just a few micrometers; however, larger aggregates with dimensions higher than 20 µm were also observed.Any notable changes with increasing nickel loading can be observed.SEM images (Figure 3) of the investigated catalysts showed a wide range of grain sizes.Generally, grain diameters measured just a few micrometers; however, larger aggregates with dimensions higher than 20 µm were also observed.Any notable changes with increasing nickel loading can be observed.

Surface Characterization
The analysis of hydroxyls that presented on the surface of the investigated samples can be seen in Figure 4. Regarding CeO2 support, four hydroxyl bands can be detected: a single bonded hydroxyl group (type I) as a shoulder around 3700 cm −1 and two types of hydroxyls bicoordinated (type IIA) to Ce(III) centered at 3690 cm −1 and Ce(IV) centered at 3655 cm −1 .The last peak can be ascribed to hydroxyl coordinated to two cations localized in the vicinity of an oxygen vacancy (type IIB) to Ce(IV) centered at 3620 cm −1 [55,56].It is noteworthy that even small amounts of nickel led to the disappearance of the bands attributed to hydroxyl groups mentioned above, with the simultaneous appearance of the band at 3680 cm −1 , which is typical of hydroxyls bonded to nickel(II) centers.Such an interpretation was confirmed by our reference DRIFT measurement carried out for the NiO

Surface Characterization
The analysis of hydroxyls that presented on the surface of the investigated samples can be seen in Figure 4. Regarding CeO 2 support, four hydroxyl bands can be detected: a single bonded hydroxyl group (type I) as a shoulder around 3700 cm −1 and two types of hydroxyls bicoordinated (type IIA) to Ce(III) centered at 3690 cm −1 and Ce(IV) centered at 3655 cm −1 .The last peak can be ascribed to hydroxyl coordinated to two cations localized in the vicinity of an oxygen vacancy (type IIB) to Ce(IV) centered at 3620 cm −1 [55,56].It is noteworthy that even small amounts of nickel led to the disappearance of the bands attributed to hydroxyl groups mentioned above, with the simultaneous appearance of the band at 3680 cm −1 , which is typical of hydroxyls bonded to nickel(II) centers.Such an interpretation was confirmed by our reference DRIFT measurement carried out for the NiO sample as well as by data from the literature [57].A similar situation can be observed in the case of carboxyl groups from adsorbed CO 2 from the air.Some of the bands visible in the DRIFT spectra of bare CeO 2 disappeared even at small nickel loading.Instead of carboxyl groups bonded to cerium centers, the doublet band at 1275-1260 cm −1 appeared, which can be ascribed to the vibration of the C-O bond [57].Those DRIFT results partially confirm strong interactions between CeO 2 support and the nickel-containing deposited phase.Additionally, the heterogeneity of surface hydroxylation has been confirmed.sample as well as by data from the literature [57].A similar situation can be observed in the case of carboxyl groups from adsorbed CO2 from the air.Some of the bands visible in the DRIFT spectra of bare CeO2 disappeared even at small nickel loading.Instead of carboxyl groups bonded to cerium centers, the doublet band at 1275-1260 cm −1 appeared, which can be ascribed to the vibration of the C-O bond [57].Those DRIFT results partially confirm strong interactions between CeO2 support and the nickel-containing deposited phase.Additionally, the heterogeneity of surface hydroxylation has been confirmed.Deconvoluted UV/Vis-DR spectra recorded for both CeO2 and NiOx/CeO2 samples in ambient conditions are shown in Figure 5.The spectrum of CeO2 support (Figure 5a) was deconvoluted into three peaks with maxima at 221, 276, and 343 nm.The first maximum can be attributed to the ligand-to-metal charge transfer (LMCT) O 2−  Ce 3+ transitions, while the second, LMCT O 2−  Ce 4+ transitions, and the last one-to the interband transitions [58,59].The addition of a small amount of nickel (0.2%) resulted in a UV/Vis-DR spectrum quite similar to the spectrum of the undoped sample (Figure 5b).Moreover, an expanded broad band at higher wavelengths can be easily visible.The presence of this kind of band can be attributed to the transition characteristic of bulk NiO [60].A further increase in nickel loading led to the increment of the intensity of the bands characteristic of nickel oxides (Figure 5c-f).Deconvoluted UV/Vis-DR spectra recorded for both CeO 2 and NiO x /CeO 2 samples in ambient conditions are shown in Figure 5.The spectrum of CeO 2 support (Figure 5a) was deconvoluted into three peaks with maxima at 221, 276, and 343 nm.The first maximum can be attributed to the ligand-to-metal charge transfer (LMCT) O 2− → Ce 3+ transitions, while the second, LMCT O 2− → Ce 4+ transitions, and the last one-to the interband transitions [58,59].The addition of a small amount of nickel (0.2%) resulted in a UV/Vis-DR spectrum quite similar to the spectrum of the undoped sample (Figure 5b).Moreover, an expanded broad band at higher wavelengths can be easily visible.The presence of this kind of band can be attributed to the transition characteristic of bulk NiO [60].A further increase in nickel loading led to the increment of the intensity of the bands characteristic of nickel oxides (Figure 5c-f).
The surface analysis performed using XPS in the characteristic Ni 2p region (Figure 6a) showed that both metallic and oxidized nickel coexisted on the surface of all the impregnated samples, while Ni 2+ was stabilized in the form of hydroxide Ni(OH) 2 and oxyhydroxide NiOOH, as it can be inferred from the spectra after the fitting procedure [61].The ratio of metallic nickel species to the oxidized one depended on nickel loadings.This ratio as a function of nickel concentration changed in a nonlinear way (Figure 6c).The analysis of the Ce 3d region (Figure 6b) showed that both Ce 3+ and Ce 4+ were present on the surface of all investigated samples.The Ce 3+ /Ce 4+ concentration ratio apparently changed with nickel loading (Figure 6d)-a fast decrement in this ratio can be observed up to 1 wt.% of nickel, and, above this value, the observed changes were smaller.The surface analysis performed using XPS in the characteristic Ni 2p region (Figure 6a) showed that both metallic and oxidized nickel coexisted on the surface of all the impregnated samples, while Ni 2+ was stabilized in the form of hydroxide Ni(OH)2 and oxyhydroxide NiOOH, as it can be inferred from the spectra after the fitting procedure [61].The ratio of metallic nickel species to the oxidized one depended on nickel loadings.This ratio as a function of nickel concentration changed in a nonlinear way (Figure 6c).The analysis of the Ce 3d region (Figure 6b) showed that both Ce 3+ and Ce 4+ were present on the surface of all investigated samples.The Ce 3+ /Ce 4+ concentration ratio apparently changed with nickel loading (Figure 6d)-a fast decrement in this ratio can be observed up to 1 wt.% of nickel, and, above this value, the observed changes were smaller.The overall surface composition of the investigated samples is collected in Table 1.As it can be seen, the concentrations of nickel were higher than the nominal value, which was expected due to the manner of its introduction-impregnated samples tend to localize nickel on the surface rather than in the bulk.However, the observed increment in the surface concentrations of nickel as a function of its nominal value is asymptotic, which suggests that nickel can be partially incorporated into the subsurface layers or, what is more probable, at higher loadings, nickel formed bigger crystals with higher loading (which remains in agreement with the reported XRD results).The cerium concentration decreased with nickel loading.This observation most probably suggests that cerium was covered by nickel on the surface.The concentration of oxygen remained constant in the investigated range.The overall surface composition of the investigated samples is collected in Table 1.As it can be seen, the concentrations of nickel were higher than the nominal value, which was expected due to the manner of its introduction-impregnated samples tend to localize nickel on the surface rather than in the bulk.However, the observed increment in the surface concentrations of nickel as a function of its nominal value is asymptotic, which suggests that nickel can be partially incorporated into the subsurface layers or, what is more probable, at higher loadings, nickel formed bigger crystals with higher loading (which remains in agreement with the reported XRD results).The cerium concentration decreased with nickel loading.This observation most probably suggests that cerium was covered by nickel on the surface.The concentration of oxygen remained constant in the investigated range.

Catalytic Activity
At temperatures above 550 • C, all NiO x /CeO 2 samples containing at least 1 wt.% of nickel showed catalytic activity in dry methane reforming at temperatures above 550 • C. The observed catalytic conversions generally increased with temperature (Figure 7a,b).However, a rapid deactivation was observed for the catalyst of the highest (10 wt.%) investigated loading of nickel.This can be explained by the formation of larger aggregates of nickel species, which is adverse to activity because of a higher amount of hidden nickel inside those aggregates.In this case, the conversion of CO 2 was higher than the corresponding conversion of CH 4 (maximal values above 90% and 60%, respectively).This effect can be explained by the associated occurrence of the reverse water-gas shift reaction (RWGS) (6) H which is the main side reaction competing with DMR.As a result of RWGS, the maximal ratio of the DMR products, hydrogen and carbon monoxide, was slightly lower than the stoichiometric ratio, i.e., it remained below unity for all investigated catalysts.Generally, the H 2 /CO concentration ratio seems to be somehow associated with the activity of the investigated catalysts.As it was confirmed by the current studies, both vital catalytic parameters, i.e., activity and selectivity, nonlinearly depend on nickel loading (Figure 7d-f).Maximal values of the activity were observed for 3 wt.% of nickel content, suggesting that the vital concentration-dependent factors (degree of Ni reduction, active-phase dispersion, as well as nickel speciation) are the crucial parameters for activity in dry methane reforming.Moreover, those parameters can be controlled by the appropriate optimization of nickel loading.Similar results were found for Ni/Al 2 O 3 -CeO 2 catalysts, where the influence of nickel loading was investigated in the range of 2.5 wt% to 15 wt% and the optimal concentration was found for 10 wt of Ni [62].An experiment to investigate the deactivation progress of the investigated catalysts required a return measurement of activity at 650 • C. A comparison of activity at the beginning of the catalytic test and after raising the reaction temperature to 800 • C is shown in Figure 8.It can be noticed that even if the optimal nickel concentration for DMR is 3 wt.%, taking into account the catalytic activity, the stability of the investigated catalysts distinctly increases with nickel loading, and thus the 5NiO x /CeO 2 catalyst showed better catalytic performance.The lower catalytic activity of 5NiO x /CeO 2 in comparison to 3NiO x /CeO 2 resulted in a lower rate of carbon deposit production and thus better stability.Moreover, the 5 wt.% sample possibly had a higher number of active centers (however, they were less active).SEM images of the investigated catalysts after the catalytic tests in DMR are shown in Figure 9.It can be stated that there are no significant changes in the grain morphology of supports; however, for the most active samples, 3% and 5% of nickel, characteristic carbon whiskers [63,64] can be observed.This clearly shows the reason for the observed activity decrement.Careful inspection of the images recorded at the higher magnification showed nickel/nickel oxide grains of diameters ranging several hundred nanometers.Such structures can be analyzed as a part of agglomerates/aggregates of dimensions from one to a few micrometers.Their formation can be a result of nickel reduction in the course of the DMR reaction.Images recorded for the fresh samples presented in Figure 3 did not reveal structures like those.For 3NiO x /CeO 2 , as well as for 5NiO x /CeO 3 samples, carbon whiskers of diameters comparable with nickel/nickel oxide grains can be observed.Based on the data from the literature [65], it can be stated that spherical morphology is less favorable for activity than other ones.

Conclusions
The described structural and functional investigations of a series of NiOx/CeO2 catalysts of various nickel loadings showed that both catalytic activities of our samples as well as their surface properties distinctly depend on nickel concentration.The observed character of this dependence is nonlinear.The metallic to oxidized nickel content ratio as well as the activity of the investigated samples in dry methane reforming initially increased

Figure 1 .
Figure 1.Diffraction patterns (a) and their zooming in the range of 35-45° (b), and Raman spectra (c) of the studied NiOx/CeO2 samples of various concentrations of nickel (from 0 to 10 wt.%).

Figure 1 .
Figure 1.Diffraction patterns (a) and their zooming in the range of 35-45 • (b), and Raman spectra (c) of the studied NiO x /CeO 2 samples of various concentrations of nickel (from 0 to 10 wt.%).

Figure 4 .
Figure 4. DRIFT spectra of the investigated samples in the diagnostic hydroxyl (a) and structural and carboxyl (b) regions.

Figure 4 .
Figure 4. DRIFT spectra of the investigated samples in the diagnostic hydroxyl (a) and structural and carboxyl (b) regions.

Figure 7 .
Figure 7. Conversion of CO2 (a) and CH4 (b) together with H2/CO (c) product content ratio as a function of temperature, as well as the conversion of CO2 (d) and CH4 (e) together with H2/CO (f) product content ratio as a function of the nominal nickel concentration in the series of NiOx/CeO2 catalysts (1-10% Ni) tested in dry methane reforming.

Figure 7 .
Figure 7. Conversion of CO 2 (a) and CH 4 (b) together with H 2 /CO (c) product content ratio as a function of temperature, as well as the conversion of CO 2 (d) and CH 4 (e) together with H 2 /CO (f) product content ratio as a function of the nominal nickel concentration in the series of NiO x /CeO 2 catalysts (1-10% Ni) tested in dry methane reforming.

Figure 7 .
Figure 7. Conversion of CO2 (a) and CH4 (b) together with H2/CO (c) product content ratio as a function of temperature, as well as the conversion of CO2 (d) and CH4 (e) together with H2/CO (f) product content ratio as a function of the nominal nickel concentration in the series of NiOx/CeO2 catalysts (1-10% Ni) tested in dry methane reforming.

Figure 8 .
Figure 8. Conversion of CO 2 (a) and CH 4 (b) in 650 • C (initial and return (to the back point) experiments) for NiO x /CeO 2 catalysts (1-5% Ni) tested in dry methane reforming.

Table 1 .
Surface compositions of the investigated samples through XPS.

Table 1 .
Surface compositions of the investigated samples through XPS.