Nickel Catalysts Supported on SiO₂-CeO₂ Mixed Oxides for Methane Dry Reforming

Abstract

Nickel-supported catalysts over SiO₂-CeO₂ mixed oxides were investigated as catalysts for syngas production via dry reforming of methane. SiO₂-CeO₂ supports were optimized by varying the preparation method and ceria loading with the aim of stabilizing nickel nanoparticles, enhancing the catalytic performance, and improving the resistance to coke formation under high-temperature reforming conditions. To investigate the effect of support composition, SiO₂-CeO₂ mixed oxides with ceria contents ranging from 5 to 30 wt% were prepared using two synthesis routes: sol–gel and wetness impregnation methods. A nickel loading of 5 wt% was deposited on the resulting supports. The catalysts were characterized by XRD, N₂ physisorption, temperature-programmed reduction (TPR), and Raman spectroscopy. Catalytic activity tests were carried out over reduced catalysts in an H₂-He stream at 750 °C, using a feed mixture containing 15 vol% CH₄ and 15 vol% CO₂ in He. The effect of temperature on catalytic performance was evaluated in the range of 450–750 °C. Thermogravimetric, XRD and Raman analyses of spent catalysts were used to assess carbon deposition and the nature of crystalline phases. The results highlight the role of CeO₂ content and preparation method in determining nickel dispersion, reducibility, catalytic performance in DRM, and coke resistance.

1. Introduction

Greenhouse gas emissions from fossil fuel combustion and industrial processes have emerged as a primary driver of the current environmental crisis. The increasing atmospheric concentration of pollutants such as carbon dioxide and methane is resulting in pronounced climate change. Addressing this crisis requires a transition to low-carbon energy systems combined with coordinated international policy efforts to ensure long-term environmental sustainability [1]. This challenge has motivated the development of catalytic technologies designed to convert greenhouse gases into valuable chemical feedstocks. In this context, dry reforming of methane (DRM: CH₄ + CO₂ → 2H₂ + 2CO) has attracted growing attention because it simultaneously consumes CH₄ and CO₂ to produce synthesis gas (CO + H₂), a key intermediate for downstream processes such as Fischer–Tropsch synthesis, methanol synthesis, and hydrocarbon upgrading [2,3]. However, practical implementation of DRM faces significant challenges, including high temperatures required to activate stable reactants and catalyst deactivation through carbon deposition and metal sintering [4,5].

Nickel-based catalysts are widely regarded as the most economical and industrially viable materials for DRM, offering high intrinsic activity for C–H bond activation at a significantly lower cost than noble metals such as Ru, Rh, or Pt [6]. Nevertheless, their application remains limited by rapid deactivation caused by carbon formation [7], growth and agglomeration of Ni nanoparticles, and interactions leading to inactive phases. To mitigate these issues, the development of advanced materials that can stabilize Ni particles, improve reducibility, and enhance coke resistance has become a key research focus [8,9,10]. Among the reported catalytic systems for methane dry reforming, CeO₂ has emerged as a promising support for Ni active sites, owing to its excellent redox properties and oxygen storage capacity which promote the removal of carbon deposits, enhance catalyst stability, and mitigate coke formation [11,12,13]. The Ce³⁺/Ce⁴⁺ redox cycle enables lattice oxygen to participate in the oxidation of CH₄, generating oxygen-deficient CeO_2−x species that can be reoxidized by CO₂, thus preserving the catalytic performance and extending the lifetime of the Ni/CeO₂ catalytic system.

Hybrid SiO₂-CeO₂ mixed oxides represent a particularly attractive support for Ni-based catalysts [14,15,16]. Silica provides high surface area, thermal stability, and well-controlled porosity [17,18], while ceria contributes oxygen mobility and redox activity that facilitate carbon oxidation and CO₂ activation [19]. The combination of these two oxides can therefore enhance nickel dispersion, strengthen metal–support interactions, and improve resistance to carbon deposition during DRM [20,21]. Importantly, both the ceria content and the support synthesis method significantly influence the physicochemical properties of SiO₂-CeO₂ systems, affecting parameters such as surface area, degree of Ni–support interaction, oxygen vacancy concentration, and reducibility [22,23,24].

In this study, we systematically investigate the influence of CeO₂ content (5–30 wt%) and the support synthesis method (sol–gel vs. wetness impregnation) on the structural, textural, and catalytic properties of Ni/SiO₂-CeO₂ catalysts for methane dry reforming. SiO₂-CeO₂ composites were prepared using either the sol–gel route or wetness impregnation, followed by incorporation of nickel via wetness impregnation. Comprehensive physicochemical characterization, including X-ray diffraction (XRD), N₂ physisorption, and H₂ temperature-programmed reduction (H₂-TPR), was performed to elucidate structural evolution, porosity, and reducibility. Catalytic performance was evaluated between 450 and 800 °C using a CH₄/CO₂ mixture, and spent catalysts were examined by thermogravimetric analysis (TG) and XRD to evaluate carbon deposition and phase stability. By correlating preparation methods, ceria content, and reaction performance, this work provides new insights into the design of Ni-based catalysts with enhanced coke resistance and improved stability for DRM.

2. Results and Discussion

SiO₂-CeO₂ mixed oxides (denoted as SiCe) were developed as supports for Ni-based catalysts to be used in syngas production via methane reforming with carbon dioxide. These SiCe supports were designed to stabilize dispersed Ni nanoparticles, providing enhanced catalytic activity and improved resistance to coke formation during the high-temperature reforming process.

SiCe powders were synthesized using two distinct strategies: the sol–gel method (SiCe5-30-SG) and the wetness impregnation method (SiCe5,30-WI).

To investigate the influence of the support composition, four SiCe materials with ceria contents of 5, 10, 20, and 30 wt% were synthesized using the sol–gel method. A pure SiO₂ support was also prepared using both methods as a reference material, and as the reference support for the synthesis of two additional SiCe samples with CeO₂ contents of 5 and 30% by the wetness impregnation method (SiCe5-WI, SiCe30-WI).

The supported Ni catalysts (5 wt%), prepared by wetness impregnation of SiCe powders, were labeled as NiSi, NiSiCe5-SG, NiSiCe10-SG, NiSiCe20-SG, NiSiCe30-SG, NiSiCe5-WI, and NiSiCe30-WI.

Ni catalysts were analyzed by X-ray powder diffraction (Figure 1) to investigate their structural properties. The recorded patterns of most samples display the characteristic reflections of CeO₂ at 28.5°, 33.0°, 47.5°, and 56.3° 2θ, assigned to the (111), (200), (220), and (311) planes of its fluorite-type cubic structure (ICSD No. 28753, space group Fm–3m).

In contrast, when the sol–gel method was applied to prepare the material with the lowest CeO₂ content, NiSiCe5-SG, the corresponding XRD pattern shows barely visible CeO₂ diffractions, suggesting the presence of ceria as highly dispersed phase and fine nanometric domains embedded within the amorphous silica network. XRD patterns of the calcined catalysts clearly display NiO reflections consistent with ICSD reference No. 24018. The primary NiO peaks are detected at 37.2°, 43.3°, and 62.9° 2θ, corresponding to the (111), (200), and (220) planes, respectively. Interestingly, as the ceria content increases, these peaks become less intense and broader, indicating that NiO particles are more finely dispersed on the supports with higher CeO₂ loadings.

NiO mean crystal sizes were calculated using the Scherrer equation, and the values are listed in

Table 1. Specific surface area, pore volume, pore size, and crystal size of the supports and investigated catalysts.

Sample	SSA (m2/g)	Pore Volume (cm3g−1)	BJH Mean Pore Size (nm)	NiO Crystal Size 1 (nm)
Si-SG	306	0.68	10.4	-
NiSi	242	0.50	16.4	22.7
SiCe5-SG	359	0.30	2.8	-
NiSiCe5-SG	310	0.16	2.7	17.0
SiCe10-SG	304	0.20	3.3	-
NiSiCe10-SG	230	0.12	2.9	16.6
SiCe20-SG	290	0.15	4.0	-
NiSiCe20-SG	149	0.10	3.1	15.7
SiCe30-SG	263	0.15	2.7	-
NiSiCe30-SG	163	0.12	4.2	11.8
SiCe5-WI	275	0.57	16.2	-
NiSiCe5-WI	235	0.48	17.7	19.0
SiCe30-WI	240	0.53	14.4	-
NiSiCe30-WI	228	0.52	14.9	11.0

¹ Calculated by Scherrer equation.

. An effect of the catalyst composition on structural properties was observed. NiSi, without ceria, exhibits the largest NiO crystallite size (22.7 nm). With incorporation of CeO₂ into the silica matrix, a systematic reduction in NiO crystallite size was observed. At low ceria loadings (NiSiCe5-SG and NiSiCe10-SG), the mean NiO particle size decreases to about 17.0–16.6 nm and further to 15.7 nm for NiSiCe20-SG. The smallest NiO crystallites (11.8 nm) were obtained for NiSiCe30-SG. This progressive decrease in NiO crystallite size suggests that ceria effectively inhibits NiO crystal growth, likely because of strengthened metal–support interactions and enhanced dispersion of Ni species promoted by CeO₂. A similar effect was also abserved for the wetness-impregnated samples, NiSiCe5-WI, and NiSiCe30-WI, showing smaller and better dispersed NiO crystallites at higher CeO₂ loadings.

N₂ physisorption measurements were performed on both the supports and the investigated catalysts to evaluate their textural properties, including specific surface area (SSA), pore volume, and mean pore size (Table 1).

Specific surface area (SSA) of the catalysts differs significantly based on CeO₂ content and preparation method. The silica-supported catalyst (NiSi) has a high surface area of 242 m²/g, a pore volume of 0.50 cm³/g, and a mean pore diameter of 16.4 nm. When CeO₂ is incorporated using the sol–gel method, there is a notable impact on the textural properties compared to the pure silica oxide. At low CeO₂ content (SiCe5-SG), SSA increases to 359 m²/g, while pore volume decreases significantly to 0.30 cm³/g, and the mean pore diameter drops to 2.8 nm. This suggests the formation of a finer pore structure, likely due to the well-dispersed ceria within the silica matrix. When the CeO₂ content is increased further (SiCe10-30-SG), there is a gradual decrease in surface area and pore volume, accompanied by a slight increase in pore diameter ranging between 2.9 and 4.0 nm. This observation likely indicates a modification of the pore structure or aggregation of the CeO₂ particles. Upon Ni deposition, all the silica–ceria oxides prepared by the sol–gel method undergo a significant decrease in specific surface area and pore volume, reflecting partial pore filling and the formation of NiO crystallites on the surface.

Catalysts prepared by wetness impregnation (NiSiCe5-WI and NiSiCe30-WI) and their supports (SiCe5-WI and SiCe30-WI), while exhibiting slightly lower SSA values than pure silica and NiSi, retain higher pore volumes (0.48–0.57 cm³ g⁻¹) and larger pore sizes (14.4–17.7 nm) compared with the sol–gel samples. Moreover, it can be inferred that NiO particles interact preferentially with surface ceria rather than with silica, as the surface area of the WI catalysts is not strongly affected by nickel deposition.

Overall, based on the results reported so far, it is evident that the preparation method plays a significant role in the dispersion of CeO₂ and in the resulting textural properties of Ni/SiO₂–CeO₂ catalysts.

Hydrogen temperature-programmed reduction (H₂-TPR) analysis was performed on the Ni-supported catalysts to investigate the effect of ceria content on their reducibility and Ni-support interactions.

The observed features can be attributed to the reduction of Ni²⁺ to metallic Ni and of Ce⁴⁺ to Ce³⁺ (Figure 2), as confirmed by experimental H₂ consumption values (20–35 mL g⁻¹). These values are consistent with the theoretical hydrogen consumption required for complete reduction of Ni²⁺ to Ni⁰ (20.5 mL g⁻¹) and for the reduction of Ce⁴⁺ to Ce³⁺, which is almost complete at ceria loadings ≤10 wt%.

NiO particles with moderate interaction with the support are reduced in the range of 300–400 °C, whereas the reduction of NiO that is strongly interacting with the support occurs at higher temperatures. The TPR profile of NiSi shows broad asymmetric peaks centered at ~350 °C and ~600 °C, along with a further consumption at ~750 °C. The low-temperature reduction corresponds to large NiO crystallites with weak interaction with the support, more characteristic of bulk NiO, whereas the high-temperature features can be ascribed to the reduction in small NiO particles strongly interacting with silica [24,25].

CeO₂-containing catalysts display additional reduction features between 400 and 700 °C, which can be assigned to the reduction in surface and bulk ceria (Ce⁴⁺ → Ce³⁺). By comparing the reduction profiles of the NiSiCe-SG series with those of the NiSiCe-WI catalysts, it is evident that the high-temperature reduction peak is shifted to lower values. This finding supports the occurrence of a strong metal–support interaction in the SG series, which promotes the reduction in bulk ceria at lower temperatures.

The catalytic activity of NiSiCe(5-30)-SG catalysts was evaluated in the dry reforming of methane (DRM) reaction under temperature-gradient conditions ranging from 450 to 750 °C (Figure 3). Results indicate an increase in CH₄ conversion in the following order: NiSi > NiSiCe5-SG > NiSiCe20-SG > NiSiCe30-SG > NiSiCe10-SG.

The relatively high activity of NiSi can be attributed to the presence of NiO particles with weak interaction with the support. Among the NiSiCe-SG samples, the material with the lowest CeO₂ content, NiSiCe5-SG, performs slightly better than other catalysts. This could likely be attributed to the combination of well-dispersed ceria domains and moderately interacting NiO species.

As CeO₂ content increases beyond 10 wt%, catalytic activity decreases. This trend can be attributed to stronger Ni–support interactions induced by ceria, which lead to the formation of highly stabilized and less reactive NiO species. These factors can limit the amount of active Ni⁰ available on the surface of the catalyst under reaction conditions.

Overall, the results reported so far suggest that the catalytic performance of the NiSiCe(5-30)-SG series is influenced by the ceria content, which affects NiO particle size, reducibility and the extent of Ni–support interaction.

Figure 3 (inset) shows H₂/CO molecular ratio values as a function of temperature in the range 500–750 °C. H₂/CO values decreased with increasing temperature, dropping from above 2 at 500 °C to approximately 1–0.8 in the 650–750 °C range. The temperature-dependent evolution of the H₂/CO ratio can be attributed to side reactions during the DRM test [6]. Specifically, the exothermic Boudouard reaction (2CO ⇌ C + CO₂), which is favored at lower temperatures, leads to CO consumption, and thus contributes to the higher H₂/CO ratio. In contrast, endothermic reactions such as the reverse Boudouard and the reverse water–gas shift (RWGS) reaction (CO₂ + H₂ ⇌ CO + H₂O) enhance CO formation at higher temperatures, thereby reducing the H₂/CO molar ratio [6,25,26].

After the DRM reaction, spent catalysts were characterized by TGA and XRD analyses. The extent of carbon deposition on spent NiSi and NiSiCe (5-30)-SG catalysts was determined by thermogravimetric (TG) analysis, as shown in Figure 4a. The weight losses associated with carbon oxidation occurred between ~600–720 °C and ranged from approximately 18.5 wt% for NiSiCe5-SG to 2.8 wt% for the least active NiSiCe10-SG.

TG analysis of the spent NiSi catalyst revealed the coexistence of different carbon species (e.g., amorphous and graphitic carbon), as evidenced by its DTG profile showing two oxidation peaks centered at 586 and 703 °C.

The formation of graphitic carbon on NiSiCe (5-30)-SG catalysts after DRM was further confirmed by XRD analysis (Figure 4b), which shows a diffraction peak at 26.5° (2θ). In addition, Ni⁰ was clearly detected in spent samples, identified by the (111), (200), and (220) reflections at 2θ = 44.4°, 51.8°, and 76.3°, respectively. The calculated mean particle size of metallic nickel was in the range 25–14 nm, with the largest size for NiSi that decreased progressively with the ceria content. NiSiCe5-SG, NiSiCe10-SG and NiSiCe20-SG exhibited average particle size equal to 19–18 nm, while NiSiCe30-SG was characterized by the smallest size, 14 nm. This trend is in agreement with the previous one observed for NiO (see Table 1), with only a slight increase in the size of metallic Ni.

The absence of CeO₂ diffraction peaks after DRM was observed for all samples and has also been reported in the literature for Ni/SiO₂–CeO₂ catalysts [27,28]. This result was ascribed to the catalyst reduction pre-treatment at 750 °C under 5% H₂/He, carried out before the DRM test. The reducing atmosphere induced Ce⁴⁺ → Ce³⁺ reduction and the formation of oxygen vacancies, disrupting the long-range crystallinity of the fluorite structure.

A long-term stability test was conducted at 700 °C for 24 h over NiSi taken as a reference catalyst. As shown in Figure 5, both CH₄ and CO₂ conversions exhibit a slight increase during the initial hours on stream, followed by stable values over the entire 24 h time-on-stream test. CO₂ conversion values were higher than those recorded for CH₄. This behavior is consistent with the literature and can be explained by the occurrence of the reverse water–gas shift reaction (RWGS: CO₂ + H₂ ⇌ CO + H₂O) [25].

CO₂ conversion rapidly reaches approximately 70% and remains essentially constant thereafter, indicating steady catalytic performance. Likewise, CH₄ conversion increases to about 55% in the early stage of the reaction and subsequently stabilizes, with no evidence of deactivation. The absence of any progressive decline in conversion for either reactant points to good catalyst stability under investigated conditions and is consistent with the limited carbon deposition observed during the temperature-gradient test.

In order to verify the hypothesis that NiSiCe-SG catalysts exhibit lower activity than the corresponding NiSi sample due to the strong Ni-CeO₂ interaction, the catalyst with the highest ceria content, NiSiCe30-SG, was subjected to a redox treatment (TPR-TPO-TPR). Specifically, after the first TPR, the sample was oxidized from room temperature up to 600 °C (hold time 1 h) using a 5% O₂/He mixture (v/v, 30 mL/min), then cooled to room temperature under helium and, finally, a second TPR was performed. This procedure was carried out to investigate changes in the reducibility of the oxidized Ni species. In Figure 6a, the second reduction profile was compared with the first one, already shown in Figure 2. An evident change in the second reduction profile of NiSiCe30-SG was registered in comparison with the first TPR. After the redox treatment (TPR-TPO-TPR), the second H₂-TPR profile shifts the main reduction feature toward lower temperatures, with the maximum now centered at around 450–500 °C, and a narrower distribution of reduction peaks. This trend suggests a partial weakening of the metal–support interaction and an increased fraction of more easily reducible Ni²⁺ and Ce⁴⁺ species formed upon reoxidation. The enhanced reducibility, observed in the second TPR, indicates that the redox cycling likely promotes changes at the Ni-CeO₂ interface, which may affect the catalytic performance of the NiSiCe30-SG. Based on the results reported so far, a fresh portion of the NiSiCe30-SG catalyst was subjected to the first reduction at 750 °C for 1 h under 5% H₂/He, then, after cooling to room temperature under inert gas, the sample was oxidized at 600 °C for 1 h, and again reduced at 750 °C for 1 h. The long-term stability test was then carried out at 700 °C for 24 h (Figure 6b). Both reactants, CH₄ and CO₂, exhibit relatively high initial conversions, followed by a rapid decline during the first few hours on stream. CO₂ conversion decreases sharply from its initial value (68%) to approximately 40–45% within the first 2–3 h, whereas CH₄ conversion shows an even more pronounced early drop, falling from 70% to 30% over the same period.

The initially high activity indicates that the second reduction treatment is beneficial for the reorganization of Ni active sites, which are likely more dispersed and thus more accessible on the catalyst surface. However, the catalyst undergoes rapid deactivation within the first few hours on stream, probably due to modifications of the active sites, such as metal sintering, which affect CH₄ activation more severely than CO₂. After this transient period, both conversion curves gradually approach a quasi-steady state. Between approximately 6 and 24 h on stream, CO₂ conversion stabilizes at around 23–27%, while CH₄ conversion levels off at lower values, ca. 14–17%. Throughout the entire test, CO₂ conversion remains consistently higher than that of CH₄, in agreement with previous observations (see Figure 5). In Figure 6b (inset), the CH₄ and CO₂ conversion profiles recorded at 700 °C over the same NiSiCe30-SG catalyst after a single reduction treatment at 750 °C are shown during the first 5 h of the reaction. This observation confirms the poor catalytic activity previously observed under temperature-gradient conditions (see Figure 3). In this case, CO₂ conversion decreases modestly from ~15% to ~10% within the first hour and then remains nearly constant, whereas CH₄ conversion stays close to ~5% over the entire 5 h interval.

The influence of the support synthesis method on catalytic performance was evaluated for NiSiCe5-WI and NiSiCe30-WI catalysts by performing temperature-gradient tests between 450 °C and 750 °C. For reference, the catalytic results of the corresponding sol–gel (SG) samples (Figure 3) are included in Figure 7a, providing a comprehensive comparison of all catalysts prepared from SiCe supports with 5 wt% and 30 wt% ceria. Both NiSiCe5-WI and NiSiCe30-WI exhibit higher CH₄ conversions than NiSi and NiSiCe(5-30)-SG samples, with NiSiCe5-WI achieving the highest conversion. H₂/CO ratios for both NiSiCe-WI catalysts lie between 0.7 and 0.8 in the 500–700 °C range, remaining lower than those of the SG samples. At 750 °C, the curves intersect those of NiSi and NiSiCe-SG samples at a H₂/CO ratio equal to 1 (see inset, Figure 7a).

Spent NiSiCe5-WI and NiSiCe30-WI catalysts were subsequently characterized by XRD and TG analyses, which revealed the presence of carbon deposits (Figure 7b and Figure 8b). XRD patterns displayed main diffraction peaks of graphitic carbon partially overlapping with those of metallic Ni, especially for the spent NiSiCe5-WI, while CeO₂ diffraction peaks were not detected, in agreement with previous findings. The average particle size of metallic Ni was estimated equal to 15 nm in the case of spent NiSiCe30-WI.

Based on these findings, NiSiCe5-WI and NiSiCe30-WI were tested in DRM at 700 °C for 24 h (Figure 8a) using a feed of 45% CH₄, 45% CO₂ and 10% N₂ in order to assess the catalysts stability in realistic conditions. NiSiCe5-WI exhibited the highest activity, achieving CH₄ and CO₂ conversions of 80% and 85%, respectively, and a H₂/CO ratio of 0.7. CH₄ conversion values for NiSiCe5-WI exceed those obtained under temperature-gradient conditions by 14% (Figure 7a), indicating that NiSiCe5-WI performs efficiently at 700 °C in long-term conditions. During the long-run DRM test, NiSiCe30-WI retained almost the same CH₄ conversion as the corresponding catalytic test performed under temperature-gradient conditions. Both catalysts showed a carbon balance higher than 95% during all the analyzed time range.

Raman spectra of the as-prepared and spent NiSiCe5,30-SG and NiSiCe5,30-WI catalysts were recorded to investigate additional structural features (Figure 9), complementing XRD findings. For all samples, the resulting spectra exhibit a peak centered at 483 cm⁻¹, which is ascribed to the stretching vibration of Si-O-Si bonds [29].

For spent catalysts, the intensity of this peak is markedly reduced, suggesting partial coverage or blockage of the silica framework. This effect could likely be caused by carbon deposition formed during the DRM reaction, which can hinder the detection of Si–O–Si vibrations and may also affect the structural integrity and accessibility of the active sites. The F_2g vibration mode of the CeO₂ fluorite structure was clearly detected at ~462 cm⁻¹ for the as-prepared NiSiCe5-WI and NiSiCe30-SG. The sharper peak observed for the as-prepared NiSiCe5-WI is consistent with its higher crystallinity [30], which can be attributed to the synthesis method. In contrast, the F_2g peak was not detected in the NiSiCe5-SG analog, reflecting the highly dispersed and poorly crystalline nature of CeO₂ resulting from the sol–gel synthesis of the SiCe5-SG support.

Consistently with XRD analysis, the characteristic F₂g Raman mode of CeO₂ is not clearly detected in spent catalysts after DRM, which may be attributed to the combined effects of carbon deposition and a reduction in the long-range order of ceria domains during the reaction. The D₁ bands observed in the 500–600 cm⁻¹ range are ascribed bulk oxygen vacancies (O_v−b) in the ceria fluorite structure [31]. These bands reflect the presence of local structural defects in CeO₂, which are likely related to the presence of Ce³⁺ species. Moreover, the D₁ band of the spent catalysts appears more pronounced than in the corresponding as-prepared samples, indicating that DRM conditions induce additional oxygen vacancies and structural disorder. Qualitatively, while the F_2g band of CeO₂ reflects the long-range fluorite order, the relative prominence of the D₁ band in the spent samples suggests an increase in defect density. Among the catalysts studied, NiSiCe5 WI exhibits the strongest D₁ signal, consistent with a higher concentration of bulk oxygen vacancies, which may contribute to its superior catalytic performance [32].

Raman spectroscopy was also employed to elucidate the nature and structural order of carbon deposits on spent catalysts after DRM, focusing on the 1260–1680 cm⁻¹ Raman shift range (Figure 10). Spectra of the spent catalysts exhibit the characteristic G band at ~1588 cm⁻¹ associated with the E_2g vibration of sp² carbon, and the D band at ~1309–1352 cm⁻¹ which becomes active in the presence of defects or structural disorder. The I_D/I_G ratio was calculated to assess the degree of disorder versus graphitization in the deposited carbon, as it is widely used to quantify the structural quality of sp² carbon materials [33,34].

Higher values denote a more defective carbon network, whereas lower ratios indicate increased ordering. The catalysts recovered after the temperature-gradient test exhibit I_D/I_G values of NiSiCe5-WI (1.21) > NiSiCe30-SG (1.17) > NiSiCe5-SG (1.14) > NiSiCe30-WI (1.10), indicating that NiSiCe30-WI accumulates the most graphitized carbon. After the long-term stability test, the I_D/I_G values converge (1.06 for NiSiCe5-WI and 1.05 for NiSiCe30-WI), suggesting that extended operation promotes the formation of similarly ordered carbon species.

3. Materials and Methods

3.1. Synthesis Procedures of Si-SG, SiCe(5-30)-SG, SiCe(5-30)-WI and the Corresponding Ni-Based Powders

Pure SiO₂ (Si-SG) and SiO₂-CeO₂ mixed oxides (SiCe(5-30)-SG) were prepared via the sol–gel method. For the synthesis of pure SiO₂, TEOS (11 mL) was first dissolved in 8.5 mL of ethanol under magnetic stirring at room temperature. To this solution, 8 mL of an acetic acid solution (pH = 6) was added dropwise under continuous stirring. The resulting mixture was then heated to 45 °C for 1 h and subsequently to 80 °C for an additional hour. After heating, the sol was cooled to room temperature and kept overnight to promote gelation. The resulting gel was manually crushed at 80 °C and dried at 100 °C overnight. The dried material was then ground and calcined at 450 °C for 4 h (2 °C min⁻¹).

SiCe(5-30)-SG materials were prepared following the same procedure, adding the appropriate amounts of cerium (III) nitrate (Ce(NO₃)₃·6H₂O) as the cerium oxide precursor to the silica sol solution to obtain final ceria loadings of 5%, 10%, 20%, and 30% (w/w).

SiCe5-WI and SiCe30-WI were prepared by sequential wetness impregnation of the pure Si-SG with an aqueous solution of cerium nitrate. The resulting materials were dried at 100 °C and calcined at 450 °C with a heating rate of 2 °C min⁻¹ for 4 h.

All Ni-containing catalysts, based on both the unmodified and Ce-modified supports, were synthesized by sequential wetness impregnation using an aqueous solution of Ni(NO₃)₂·6H₂O. Nominal nickel loading was 5 wt%. Catalyst samples were dried at 100 °C and calcined under static conditions at 600 °C with a heating rate of 2 °C min⁻¹ for 4 h (as-prepared samples).

3.2. Characterization of the Materials

Specific surface areas and pore volumes were determined from N₂ adsorption–desorption isotherms measured at −196 °C using an ASAP 2020 instrument (Micromeritics, Norcross, GA, USA). Prior to analysis, the samples were degassed at 250 °C for 2 h. Surface areas were calculated using the Brunauer–Emmett–Teller (BET) method over the relative pressure range 0.05–0.30 (P/P₀). Pore volumes and pore size distributions were obtained from the desorption branch using the Barrett–Joyner–Halenda (BJH) method.

Powder X-ray diffraction (XRD) patterns were collected on a D5000 diffractometer (Bruker AXS, Karlsruhe, Germany) using Ni-filtered Cu Kα radiation. Data were recorded over the 2θ range 10–90° with a step size of 0.05°. Phase identification was performed using the ICSD database (Inorganic Crystal Structure Database, version 5.5.0), and crystallite sizes were calculated using the Scherrer equation applied to the main reflections [35].

Hydrogen temperature-programmed reduction (H₂-TPR) was conducted using an Autochem 2910 system (Micromeritics, Norcross, GA, USA) equipped with a TCD detector. Samples (0.10 g) were pretreated in 5% O₂/He (30 mL min⁻¹) by heating to 400 °C at 10 °C min⁻¹ and holding for 30 min. After cooling to room temperature under He flow (30 mL min⁻¹), reduction was carried out in 5% H₂/Ar (30 mL min⁻¹) up to 900 °C at 10 °C min⁻¹. H₂ consumption was quantified by integration of the TPR peaks using a calibration curve obtained from different H₂/Ar gas mixtures. The uncertainty in hydrogen consumption measurements was ±2%. Consecutive temperature-programmed reduction/temperature-programmed oxidation (TPR/TPO) and subsequent TPR experiments were conducted by oxidizing the previously reduced sample at 600 °C for 1 h with 5% O₂/He (30 mL min⁻¹), followed by cooling to room temperature under He flow. A subsequent TPR run was then performed under identical conditions as above described.

Raman spectra of the powders were recorded using a Thermo Fisher Scientific (Waltham, MA, USA) DXR3 Raman spectrometer with a 532 nm excitation laser and a laser power of 0.2 mW.

3.3. DRM Catalytic Tests

The dry reforming of the methane (DRM) reaction was performed in a fixed-bed quartz reactor (inner diameter: 12 mm) operating at atmospheric pressure. Catalytic tests were conducted as a function of the temperature, in the temperature range 450–750 °C, using 50 °C increments. Tests were also carried out at 700 °C for 24 h of time-on-stream (TOS) to evaluate the stability in catalytic conversion.

Typically, for each catalytic, 100 mg of test catalyst (sieved to 180–250 µm) was diluted to 1:2 with inert SiC and loaded into the reactor. Prior to reaction, each catalyst was treated in 5% O₂/He (v/v, 50 mL min⁻¹) at 350 °C for 30 min to clean the surface and ensure reproducible starting conditions. After cooling to room temperature, the sample was reduced at 750 °C (heating rate of 10 °C min⁻¹) under 5% H₂/He (v/v, 30 mL min⁻¹) and held at this temperature for 1 h. The feed mixture consisted of 10% CH₄, 10% CO₂, and N₂ as balance, introduced at a total flow rate of 100 mL min⁻¹ (STP), corresponding to a gas hourly space velocity (GHSV) of 60,000 mL g⁻¹ h⁻¹. For two selected catalysts, NiSiCe5-WI and NiSiCe30-WI, a concentrated feed containing 45% CH₄, 45% CO₂ and 10% N₂ was used for performing 24 h long-run DRM tests at 700 °C to assess the catalysts stability in realistic conditions.

The inlet and outlet streams were analyzed online by an Agilent 7890B gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) equipped with a DB-1 capillary column and a molecular sieve column, using flame ionization (FID) and thermal conductivity (TCD) detectors to monitor CH₄, CO₂, CO, and H₂. Water formed during reaction was condensed at the reactor outlet.

Methane, carbon dioxide conversions (XCH₄ and XCO₂) and carbon balance were calculated using the following equations, where CH₄ⁱⁿ, CO₂ⁱⁿ and CH₄^out, CO₂^out, H₂^out and CO^out denote inlet and outlet concentrations (ppm), respectively:

XCH₄ = 100 × (CH₄ⁱⁿ − CH₄^out)/CH₄ⁱⁿ

XCO₂ =100 × (CO₂ⁱⁿ − CO₂^out)/CO₂ⁱⁿ

Carbon balance = 100 × (CH₄^out + CO₂^out + CO^out)/(CH₄ⁱⁿ + CO₂ⁱⁿ)

4. Conclusions

Nickel catalysts supported on SiO₂-CeO₂ mixed oxides were investigated for methane dry reforming (DRM), focusing on effects of ceria loading and the support synthesis method on Ni dispersion, reducibility, catalytic activity, and carbon resistance. Results show that both support composition and preparation method strongly influence the structure–performance relationships of Ni-based catalysts. For sol–gel-derived supports, increasing CeO₂ (5–30 wt%) enhanced NiO dispersion and reduced metallic Ni crystallite size, indicating metal–support interactions. At high loadings, however, such effect was detrimental to the catalytic activity, likely due to the low reactivity of Ni species strongly stabilized by ceria.

The sol–gel-derived Ni/SiO₂ catalyst under temperature-gradient tests performed slightly better than NiSiCe10-30-SG samples and exhibited good stability at 700 °C, maintaining constant CH₄ and CO₂ conversions over 24 h. This observation proves the benefit of weak-to-moderate metal–support interactions for stabilizing active Ni⁰ sites. Redox cycling of the high-ceria NiSiCe30-SG catalyst temporarily improved reducibility and initial activity but resulted in rapid deactivation, likely due to Ni sintering or structural changes. A comparison of wetness-impregnated catalysts revealed a strong preparation effect: while these catalysts preserved silica texture and exhibited higher activity and very high initial conversions, they suffered from increased carbon deposition, particularly ordered graphitic carbon.

Overall, optimal Ni/SiO₂-CeO₂ catalysts for DRM require careful tuning of the ceria content and synthesis method. NiSiCe5-WI exhibits the best CH₄ and CO₂ conversion values at 700 °C for 24 h of time-on-stream thanks to the high concentration of bulk oxygen vacancies and a balanced metal–support interaction, highlighting the beneficial effect of the wetness impregnation method.