The Effect of ZrO₂ as Different Components of Ni-Based Catalysts for CO₂ Reforming of Methane and Combined Steam and CO₂ Reforming of Methane on Catalytic Performance with Coke Formation

Abstract

The role of ZrO₂ as different components in Ni-based catalysts for CO₂ reforming of methane (CRM) has been investigated. The 10 wt.% Ni supported catalysts were prepared with ZrO₂ as a support using a co-impregnation method. As a promoter (1 wt.% ZrO₂) and a coactive component (10 wt.% ZrO₂), the catalysts with ZrO₂ were synthesized using a co-impregnation method. To evaluate the effect of the interaction, the Ni catalyst with ZrO₂ as a coactive component was prepared by a sequential impregnation method. The results revealed that the activity, the selectivity, and the anti-coking ability of the catalyst depend upon the ZrO₂ content, the Ni-ZrO₂ interaction, basicity, and oxygen mobility of each catalyst resulting in different Ni dispersion and oxygen transfer pathway from ZrO₂ to Ni. According to the characterization and catalytic activation results, the Ni catalyst with low ZrO₂ content (as a promoter) presented highest selectivity toward CO owning to the high number of weak and moderate basic sites that enhance the CO₂ activation-dissociation. The lowest activity (CH₄ conversion ≈ 40% and CO₂ conversion ≈ 39%) with the relatively high quantity of total coke formation (the weight loss of the spent catalyst in TGA curve ≈ 22%) of the Ni catalyst with ZrO₂ as a support is ascribed to the lowest Ni dispersion due to the poor Ni-ZrO₂ interaction and less oxygen transfer from ZrO₂ to the deposited carbon on the Ni surface. The effect of a poor Ni-ZrO₂ interaction on the catalytic activity was deducted by decreasing ZrO₂ content to 10 wt.% (as a coactive component) and 1 wt.% (as a promoter). Although Ni catalysts with 1 wt.% and 10 wt.% ZrO₂ provided similar oxygen mobility, the lack of oxygen transfer to coke during CRM process on the Ni surface was still indicated by the growth of carbon filament when the catalyst was prepared by co-impregnation method. When the catalyst was prepared by a sequential impregnation, the intimate interaction of Ni and ZrO₂ for oxygen transfer was successfully developed through a ZrO₂-Al₂O₃ composite. The interaction in this catalyst enhanced the catalytic activity (CH₄ conversion ≈ 54% and CO₂ conversion ≈ 50%) and the oxygen transport for carbon oxidation (the weight loss of the spent catalyst in TGA curve ≈ 7%) for CRM process. The Ni supported catalysts with ZrO₂ as a promoter prepared by co-impregnation and with ZrO₂ as a coactive component prepared by a sequential impregnation were tested in combined steam and CO₂ reforming of methane (CSCRM). The results revealed that the ZrO₂ promoter provided a greater carbon resistance (coke = 1.213 mmol·g⁻¹) with the subtraction of CH₄ and CO₂ activities (CH₄ conversion ≈ 28% and CO₂ conversion ≈ %) due to the loss of active sites to the H₂O activation-dissociation. Thus, the H₂O activation-dissociation was promoted more efficiently on the basic sites than on the vacancy sites in CSCRM.

1. Introduction

Among the different routes for CO₂ utilization, CO₂ reforming of methane (CRM) is regarded as one of the most beneficial reactions for energy and environment. This reaction converts CO₂ and CH₄, the main greenhouse gases (GHGs), into a synthesis gas, a mixture of CO and H₂ (R1) [1,2]. The synthesis gas can be used as a fuel or a source of high value-added chemicals such as oxygenated derivatives (e.g., methanol) and synthetic fuels (via the Fischer–Tropsch (FT) process) [3,4,5]. Because CRM is a highly endothermic reaction, it is usually operated at temperature range of 650 °C to 1000 °C [6,7]. A high temperature causes the short life of Ni-based catalysts mainly due to the metal sintering and the carbon formation, a by-product from side reactions including the Boudouard reaction (R2) and CH₄ decomposition (R3) [8,9]. Additionally, reverse water gas shift (RWGS) reaction (R4) is an another endothermic side reaction that decrease the hydrogen yield [10,11,12]. To avoid catalyst deactivation via carbon deposition during CRM process, adding steam into the feed has been applied in order to raise the rate of coke oxidation [13]. Thus, the combined steam and CO₂ reforming of methane (CSCRM) (R5) has been considered as a better promising technique for a synthesis gas production. Therefore, CSCRM combines between CO₂ reforming and steam reforming of methane (SRM) (R6) in a single process. However, this process may cause an over expected energy consumption because of the evaporation of a high water quantity. Hence, the optimum steam content in the feed should be concerned in term of the coke resistance as well as the energy balance [14,15,16].

CH₄(g) + CO₂(g) → 2H₂(g) + 2CO(g)   ΔH°₂₉₈ = +247 kJ/mol

(R1)

2CO(g) → CO₂(g) + C(s)    ΔH°₂₉₈ = −173 kJ/mol

(R2)

CH₄(g) → 2H₂(g) + C(s)     ΔH°₂₉₈ = +75 kJ/mol

(R3)

CO₂(g) + H₂(g) → CO(g) + H₂O(g)   ΔH°₂₉₈ = +41 kJ/mol

(R4)

3CH₄(g) + 2H₂O(g) + CO₂(g) → 8H₂(g) + 4CO(g)  ΔH°₂₉₈ = +712 kJ/mol

(R5)

CH₄(g) + H₂O(g) → 3H₂(g) + CO(g)    ΔH°₂₉₈ = +206 kJ/mol

(R6)

In general, Ni-based catalysts are considered as the most practical catalysts for the CRM and CSCRM processes because of their cost-effective and high catalytic performance comparable to precious metal catalysts [10,16]. As it was mentioned, Ni-based catalysts are prone to deactivating by the sintering and the carbon deposition. Therefore, the development of Ni-based catalysts has become the main strategy for the commercialized CRM process. The well-dispersed Ni nanoparticles on the catalyst surface significantly resist the metal sintering and suppress the carbon deposition. An effective method for preventing surface carbon is to create oxygen mobility in Ni-based catalysts. Because the oxygen vacancies play an important role in generating the oxygen mobility of the catalyst, oxygen storage materials have been used as a component of Ni supported catalysts for this purpose. Zirconium oxide (ZrO₂) has been considered as the practical oxygen storage material due to its oxygen storage capacity, the acid-base bifunctional properties, the mechanical strength, and the high thermal stability [17,18,19]. The ZrO₂ promoter decreased the Ni-Al₂O₃ interaction in the Ni/Al₂O₃ catalyst [20,21,22]. Adding high ZrO₂ content, compared to low ZrO₂ content, enhanced the activity and stability of Ni-based catalysts since the oxygen vacancies of the ZrO₂ surface absorbed and stored more oxygen from the dissociation of CO₂ molecule, thereby improving the reforming activity and the oxidizing carbon deposits on the active surface [21,23]. The ZrO₂ support displayed a limitation in the specific surface area [24,25,26,27]. The different preparation methods of catalysts are widely examined in the literature. Such favourite methods as co-impregnation and sequential impregnation method were found that the ZrO₂-Ni/Al₂O₃ catalysts prepared via sequential impregnation demonstrated a greater catalytic performance and coke inhibition for a dry reforming reaction and CO methanation reaction [22,28]. In CSCRM process, several works reported that the ZrO₂ addition to Ni-based catalysts was employed to enhance the oxygen storage capacity of ZrO₂, the redox property, and the promotion of metal dispersion. Additionally, the ZrO₂ improves the adsorption of steam onto the surface and activates the gasification of carbon deposited, reflecting an increase in H₂ production with a protection against a deposited carbon on Ni-based catalysts [29,30,31]. Although certain studies reported the effect of ZrO₂ as a different component, the role of them derived from their physicochemical properties that related to the CRM and CSCRM performances of Ni catalysts are required to be more clarified under the precisely similar condition. These relations can fulfil the detail of catalyst design for methane reforming processes.

This research reveals the linkage of ZrO₂ role in various components of the Ni supported catalysts and catalytic performances based upon characterization results accompanied with the CRM and CSCRM reaction data under the similar condition. In this paper, components of ZrO₂ that are not a support have been assigned by its content. According to the literature [32], promoters are defined as a substance added to heterogeneous catalysts in amounts of a few percent to rise the catalytic performances of a catalyst, and promoter loading in many works was not greater than 5%. Consequently, 1 wt.% ZrO₂ is assigned as a promoter. Because CO₂, a reactant of CRM reaction, can be adsorbed and activated at vacancies on ZrO₂, ZrO₂ at similar loading to the main active species, it is thus considered as a coactive component. As a support, the ZrO₂ was synthesized using the precipitation method, and the 10 wt.% Ni/ZrO₂ catalyst was prepared using the impregnation method. As a promoter, 1 wt.% ZrO₂ was co-impregnated with 10 wt.% Ni onto the Al₂O₃ support. As a coactive component, 10 wt.% ZrO₂ was loaded with 10 wt.% Ni onto the Al₂O₃ support using co-impregnation and the sequential impregnation method to generate the different interactions of Ni-ZrO₂ on the prepared catalysts. Physicochemical properties of these catalysts were comprehensively characterized by X-ray diffraction (XRD), N₂ adsorption–desorption, the temperature-programmed desorption of hydrogen (H₂-TPD), the temperature-programmed reduction of hydrogen (H₂-TPR), the temperature-programmed desorption of carbon dioxide (CO₂-TPD) and the temperature-programmed desorption of oxygen (O₂-TPD). The CRM activities of all catalysts were evaluated in a tubular fixed-bed reactor at 620 °C under atmospheric pressure for 10 h. Then, catalysts from this set were chosen for studying the effect of steam using CSCRM process that was tested in the similar condition with the additional steam into the feed. After the tests, the coke deposited on spent catalysts was determined employing thermal gravimetric analysis (TGA) for catalysts from CRM and the temperature-programmed oxidation (TPO) for catalysts from CSCRM. Then, this study identified the relation between the role of ZrO₂ as each component in the Ni supported catalysts and their reforming performances of methane. This study assists in further Ni-based catalysts development.

2. Results and Discussion

2.1. XRD Analysis and H₂-Temperature Programmed Desorption

The X-ray diffractograms of the reduced catalysts are displayed in Figure 1. The peaks of the tetragonal ZrO₂ phases at 2θ of 30.3°, 35.2°, 50.2°, and 60.2° (JCPDS 27-997) with the peaks of monoclinic ZrO₂ phases at 2θ of 24.1°, 28.3°, 31.5°, 34.2°, and 50.4° (JCPDS 37-1484) were only observed in the diffractograms of the reduced 10Ni/ZrO₂ catalyst [33]. The XRD patterns of the ZrO₂ were not detectable in the reduced 10Ni-xZrO₂/Al₂O₃ catalysts. In the diffractograms of the reduced 10Ni-xZrO₂/Al₂O₃ catalysts, three board peaks at 2θ of 37.0°, 43.1°, and 62.6° are ascribed to γ-Al₂O₃ phase (JCPDS No. 50-0741), overlapping with the peak pattern of the NiAl₂O₄ spinel structure (JCPDS No. 78-0552). The diffractograms of all catalysts showed peaks of the Ni crystalline phase at 2θ of 37.5°, 45.6°, and 66.6° (JCPDS No. 73-1519). For a quantitative comparison, the Ni particle size and the percentage metal dispersion calculated from the H₂-TPD results are included in

Table 1. Bulk and surface properties of Al₂O₃ support, ZrO₂ support, 10Ni/ZrO₂, 10Ni-1ZrO₂/Al₂O₃ (CI), 10Ni-10ZrO₂/Al₂O₃ (CI), and 10Ni-10ZrO₂/Al₂O₃ (SI).

Samples	N2 Adsorption-Desorption		H2-TPD
	Surface Area (m2·g−1)	Total Pore Volume (cm3·g−1)	Ni Dispersion (%)	Ni Particle Size (nm)
Al2O3 support	165	0.20	-	-
ZrO2 support	95	0.21	-	-
10Ni/ZrO2	75	0.16	3.0	21.5
10Ni-1ZrO2/Al2O3 (CI)	153	0.21	8.1	7.8
10Ni-10ZrO2/Al2O3 (CI)	134	0.20	7.1	8.9
10Ni-10ZrO2/Al2O3 (SI)	112	0.20	10.0	6.3

. The Ni particle sizes on 10Ni/ZrO₂, 10Ni-1ZrO₂/Al₂O₃ (CI), 10Ni-10ZrO₂/Al₂O₃ (CI), and 10Ni-10ZrO₂/Al₂O₃ (SI) catalysts were estimated as 21.5, 7.8, 8.9, and 6.3 nm, respectively. The largest Ni particle sizes with the lowest metal dispersion of 3.0% were found in the 10Ni/ZrO₂ catalyst, representing the weak metal-support interaction because of the poor interaction between Ni and ZrO₂ [34]. The 10Ni-xZrO₂/Al₂O₃ (CI) catalysts prepared by the co-impregnation method provided smaller Ni particle sizes with the greater metal dispersion compared to the 10Ni/ZrO₂ catalyst. The Ni particle sizes in the 10Ni-xZrO₂/Al₂O₃ (CI) catalysts diminished when the ZrO₂ content decrease, indicating a stronger metal-support interaction. Although the 10Ni-10ZrO₂/Al₂O₃ (SI) and 10Ni-10ZrO₂/Al₂O₃ (CI) catalyst have similar ZrO₂ proportion, the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst prepared by sequential impregnation exhibited the highest dispersion, resulting in the smallest Ni particle sizes and suggesting a strongest Ni-support interaction.

2.2. N₂ Adsorption-Desorption

N₂ adsorption-desorption isotherms and pore size distributions of calcined catalysts are presented in Figure 2 and the textural properties are reported in Table 1. All recorded isotherms (Figure 2a) depicted the type IV isotherm features of mesoporous materials with an H2-shaped hysteresis loop. This shaped hysteresis loop is usually found on mesoporous materials containing cylindrical-shaped channels with non-uniform pores. The pore size distribution curves (Figure 2b) were interpreted from the desorption branches of N₂ isotherm based on Barret-Joyner-Halenda (BJH) approach. Support and catalyst samples illustrated a pore size distribution smaller than 10 nm. The ZrO₂ support displayed the board peak centered at 6.0 nm of pore distribution with the surface area of 95 m²·g⁻¹. The pore distribution was shifted downward the smaller pore size (4.0 nm) and a smaller surface area (75 m²·g⁻¹) when the 10 wt.% Ni was impregnated onto the ZrO₂ support (the 10Ni/ZrO₂ catalyst). The downward shift is explained by the limitation of a specific surface area of ZrO₂ that decreases rapidly due to the blockage of pores by the large Ni particle size on the ZrO₂ support [19,21,35]. As is evident in Figure 2b, the Al₂O₃ support and the 10Ni-xZrO₂/Al₂O₃ (CI) catalysts possessed a narrower pore size distribution peak centered at approximately 2.4 to 2.7 nm. Because of the blocking effect, the co-impregnated 10 wt.% Ni-1 wt.% ZrO₂ and the 10 wt.% Ni-10 wt.% ZrO₂ onto the Al₂O₃ support decreased the surface area from 165 m²·g⁻¹ to 153 m²·g⁻¹ and 134 m²·g⁻¹, respectively. Less surface area was detected from the 10Ni-10ZrO₂/Al₂O₃ (CI) by the reason of the large NiO particles from the low NiO dispersion in the catalyst for a high ZrO₂ content confirmed with Ni dispersion in the H₂-TPD results. Compared to the 10Ni-10ZrO₂/Al₂O₃ (CI) catalyst, the character of the surface area (112 m²·g⁻¹) and the pore size distribution for the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst shifted toward the pore character of ZrO₂ support reflect the greater interaction of the ZrO₂-Al₂O₃.

2.3. H₂-Temperature Programmed Reduction

Three temperature ranges of reduction peaks were detected in the H₂-TPR profiles from 50 °C to 800 °C of all calcined catalysts (Figure 3). Peaks at low temperatures (350 °C to 500 °C) coincide with the reduction of isolated NiO at the catalyst surface as well as the reduction of NiO weakly interacting with the support [36]. The higher temperature peaks at approximately 500 °C to 650 °C are assigned to the reduction of NiO having a medium interaction with the support. Reduction peaks at temperatures over 650 °C reflect the reduction of the Ni²⁺ species in the strong metal-support interaction (SMSI) oxides (such as the NiAl₂O₄ spinel phases) [35,37]. The H₂-TPR profile of the 10Ni/ZrO₂ catalyst (ZrO₂ as support) displayed the peak shoulder centered at 425 °C and the large reduction peak centered at 545 °C. These overlapping peaks refer to the reduction of the NiO species in a weak and a medium interaction with the support [33]. The reduction at the low and medium temperatures with the greatest amount of H₂ consumption in the TPR profile of the 10Ni/ZrO₂ represents the reduction of large NiO particles with the weakest metal-support interaction among all samples. The reason could be ascribed to the low surface area of ZrO₂ support simultaneously with the poor NiO-ZrO₂ interaction. The H₂-TPR profiles of the different ZrO₂ contents on the surface of the 10Ni-xZrO₂/Al₂O₃ (CI) catalysts established two temperature peaks. The small peaks at low temperatures (centered at 405 °C to 410 °C) are correlated to the reduction of the NiO in the weak metal-support interaction [34]. The broad reduction peaks at 550 °C to 800 °C are attributed to the reduction of the Ni²⁺ species in a medium associated with a strong metal-support interaction [33,36,38]. The peak in H₂-TPR profile of 10Ni-10ZrO₂/Al₂O₃ (SI) starting at 550 °C indicates only the reduction of the NiO species in a medium and a strong metal-support interaction. When compared to 10Ni-10ZrO₂/Al₂O₃ (CI), peaks in the H₂-TPR profile of 10Ni-10ZrO₂/Al₂O₃ (SI) shifted toward higher temperatures. This characteristic corresponds to a stronger metal-support interaction representing the smaller and well-dispersed NiO species in 10Ni-10ZrO₂/Al₂O₃ (SI). Considering the result of the Ni dispersion, BJH pore distribution, and H₂-TPR that exhibited the stronger interaction in the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst, it suggests the formation of the ZrO₂-Al₂O₃ composite where the ZrO₂ probably incorporated into the Al₂O₃ at the first step of the sequential impregnation [35,39,40] before the Ni solution was loaded. The composite enhances the interaction between the Ni and the ZrO₂ in the ZrO₂-Al₂O₃ composite phase compared to the Ni and the ZrO₂ in the ZrO₂ phase.

2.4. CO₂-Temperature Programmed Desorption

The basicity of the reduced catalyst relates to the catalytic ability of CO₂ adsorption and activation. Peaks as a function of temperature in CO₂-TPD profiles of reduced catalysts (Figure 4) demonstrated the distribution of weak basic sites at 50 °C to 200 °C, moderate basic sites at 200 °C to 550 °C, and strong basic sites at temperature higher than 550 °C. The calculated number of basic sites in an individual type and in total are summarized in

Table 2. Deconvolution of the CO₂-TPD profiles and O₂-TPD profiles for all catalysts.

Samples	CO2-TPD Deconvolution (mmol·g−1)				O2-TPD Deconvolution (mmol·g−1)
	Weak	Medium	Strong	Total Basicity	α Oxygen	β Oxygen	γ Oxygen	Total O2 Desorption
10Ni/ZrO2	0.058	0.042	0.007	0.107	0.151	0.020	0.005	0.176
10Ni-1ZrO2/Al2O3 (CI)	0.191	0.225	0.034	0.450	0.009	0.260	0.028	0.297
10Ni-10ZrO2/Al2O3 (CI)	0.081	0.046	0.047	0.174	0.008	0.215	0.121	0.344
10Ni-10ZrO2/Al2O3 (SI)	0.069	0.112	0.030	0.211	0.017	0.316	0.096	0.429

. Among all types of basic sites, the weak and moderate basic sites are crucial for CO₂ adsorption and activation without the effect of too strong adsorption. As informed in Table 2, the 10Ni/ZrO₂ catalyst possessed the lowest basicity while the 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst displayed the highest number of the weak and moderate basic sites. The 10Ni-10ZrO₂/Al₂O₃ (CI) and 10Ni-10ZrO₂/Al₂O₃ (SI) catalysts presented a closing overall basic site. It implied that 10Ni-1ZrO₂/Al₂O₃ (CI) is the most active catalyst toward the CO₂ adsorption and activation in this work.

2.5. O₂-Temperature Programmed Desorption

The oxygen mobility is a property of a catalyst that relates to its carbon resistance. In order to determine the oxygen mobility properties of reduced catalysts, O₂-TPD measurement was performed. O₂-TPD profiles (Figure 5) from 50 °C to 850 °C of all samples depicted three temperature ranges of desorption peaks. Peaks at low temperature (before 250 °C), known as α oxygen, are attributed to the desorption of weakly chemisorbed oxygen molecules on the surface. The peaks at 250 °C to 550 °C, known as β oxygen, are ascribed to the desorption of the oxygen species nearby the vacancy sites. Desorption peaks at temperature higher than 550 °C, known as γ oxygen, are assigned to the release of bulk lattice oxygen [41,42,43]. The total number of O₂ desorption of 10Ni/ZrO₂, 10Ni-1ZrO₂/Al₂O₃ (CI), 10Ni-10ZrO₂/Al₂O₃ (CI), and 10Ni-10ZrO₂/Al₂O₃ (SI) catalysts were 0.176, 0.297, 0.344, and 0.429 mmol⋅gcat⁻¹ (Table 2), respectively. Although the O₂-TPD profile of the 10Ni/ZrO₂ catalyst (ZrO₂ as a support) demonstrated the minimum total O₂ desorption, it provided the highest numbers of α oxygen. When ZrO₂ located at the surface and interacted greater with Ni, the number of oxygen desorption increased with the larger amounts of β and γ oxygen. It suggests more oxygen desorption from vacancy sites and the bulk lattice. Regarding to γ oxygen among 10Ni-1ZrO₂/Al₂O₃ (CI), 10Ni-10ZrO₂/Al₂O₃ (CI), and 10Ni-10ZrO₂/Al₂O₃ (SI) catalysts, the most uniform γ oxygen species was observed in the O₂-TPD profile of the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst. This result implies the most uniform NiO-ZrO₂ interaction in the bulk.

2.6. Catalytic Activity

The results of CRM catalytic tests over the Ni-based catalysts with the ZrO₂ as a different component are illustrated in Figure 6. Considering the CRM results with the characterization results, the 10Ni/ZrO₂ catalyst (ZrO₂ as a support) represented the lowest catalytic activity. This low activity is explained by its weakest Ni-ZrO₂ interaction compared to other catalysts, eventuating in the low dispersion with the large Ni particle size [44]. The similarly high CH₄ and CO₂ conversions were observed over the catalysts with a ZrO₂ promoter and a ZrO₂ coactive component. These relatively high catalytic activities are subjected to (i) the stronger interaction of the Ni-Al₂O₃ and/or the ZrO₂-Al₂O₃ composite in the 10Ni-xZrO₂/Al₂O₃ catalysts, leading to the better dispersion with the smaller Ni particle size that increases the Ni surface area [33,40] and (ii) the ZrO₂ surface that improves the CO₂ adsorption and dissociation. Furthermore, the reactant conversions of 10Ni-10ZrO₂/Al₂O₃ catalyst increased for 4% approximately after 200 min of time-on-stream. It can be hypothesized that after 200 min, the effective coke oxidation was involved in the CRM process due to the strongest metal-support interaction combined with the β oxygen species. The pathway of rapid oxygen transport from ZrO₂ through vacancies to Ni improves CO₂ activation-dissociation via vacancy sites of ZrO₂. This reason allowed CH₄ to be activated more efficiently on Ni. Figure 6c depicts that the H₂/CO ratio in the catalytic profile of all catalysts was lower than unity because of the reverse water-gas shift side reaction (R4). At the first 20 min of time-on-stream, H₂/CO ratio decreased for all the studied samples excluding 10Ni-1ZrO₂/Al₂O₃ (CI). It can be explained by the mechanism of CRM and CO₂-TPD results. Many studies reported that the CRM reaction involves CH₄ dissociation as well as CO₂ dissociation. The dissociative chemisorption of CH₄ was denoted as an initial step and other steps take place along the time. Because CH₄ is a source of H₂ and CO₂ can be counted as a source of CO in the syngas product, the decreasing trend of H₂/CO ratio of each catalyst was observed at the beginning. Differently, the stable lowest H₂/CO ratio was obtained from the 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst as it is the most active catalyst material for CO₂ activation-dissociation (discussed in CO₂-TPD part) that can drive CO₂ dissociation since the beginning. Moreover, the H₂/CO ratio of 10Ni-10ZrO₂/Al₂O₃ (SI) was also low due to the high CO₂ activation-dissociation via the vacancies. In addition, the catalytic activities in this work are in the same range of literature data (

Table 3. Comparison of the results of this work with those of literature.

Catalyst	Preparation Method	Reactor Type	Operating Temperature	% Conversion		Ref.
				CH4	CO2
2.5%Ni nanosheet/ZrO2	Wet impregnation	Fixed-bed quartz reactor	800 °C	46.0	60.0	[4]
10%Ni/m-ZrO2-Al2O3	Evaporation-induced self-assembly	Fixed-bed quartz reactor	700 °C	60.2	78.0	[45]
15%Ni-7%ZrO2/Al2O3	Sequential impregnation	Fixed-bed reactor	700 °C	42.5	n/a	[28]
3%Ni/ZrO2 treated with N2	Deposition-precipitation	Fixed-bed quartz microreactor	700 °C	62.5	63.0	[44]
10%Ni-10%ZrO2/Al2O3	Sequential impregnation	Fixed-bed reactor	620 °C	54.3	49.8	This work

).

According to CRM results, the 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst and the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst were more selective toward CO₂ activation-dissociation as a different behavior. Thus, these two catalysts were chosen here for investigation in a CSCRM test using a similar condition to CRM with the additional steam, and the results are shown in Figure 7. During the first 100 min of time-on-stream, the CH₄ conversion as well as the CO₂ conversion decreased to approximately constant, and the H₂/CO ratio became almost stable on both catalysts. It could be counted on the adjustment of such a complex reaction. After 100 min, the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst established lower CH₄ conversion, lower CO₂ conversion, and higher H₂/CO ratio than those obtained from the 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst. It implied that a larger number of active sites on the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst were occupied by the reaction loop of H₂O activation-dissociation compared to the 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst. Consequently, the H₂O activation-dissociation was promoted greater on the basic sites than on the vacancy sites in CSCRM for this catalyst set.

2.7. Coke Deposition

The thermogravimetric analysis (TGA) and the derivative thermogravimetric (DTG) curves of spent catalysts (Figure 8) were employed to evaluate the carbon deposition over the catalyst during the CRM tests. The weight loss in the TGA curve reflects the amount of deposited coke and the temperature at which different types of coke are oxidized, indicating the difficulty of their removal. The TGA/DTG profile of the spent 10Ni/ZrO₂ catalyst (Figure 8a) represented two temperature ranges for a total weight loss of 22% relating to the oxidation of the graphitic carbon (350 °C to 550 °C) and the filament carbon (550 °C to 700 °C) [16,46,47]. The high quantity of carbon formation was observed on the spent 10Ni/ZrO₂ catalyst according to the low dispersion of large Ni on the catalyst surface because of the poor Ni-ZrO₂ interaction as found in the H₂-TPD results and the lowest number of oxygen desorption as discussed in the O₂-TPD results. The weight loss of 17% and 32% was monitored in the TGA/DTG profiles of the spent 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst (Figure 8b) and the spent 10Ni-10ZrO₂/Al₂O₃ (CI) catalyst (Figure 8c), respectively. The coke combustion occurred in the similar range of 550 °C to 800 °C corresponding to the oxidation of the filament/graphitic carbon species [48], the main type of the catalyst deactivation growing on the poor Ni-ZrO₂ interaction. Although the catalysts prepared by the co-impregnation exhibited the similar oxygen mobility and type of coke, the amount of coke on a catalyst with a low content of ZrO₂ at the surface as promoter (the 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst) is less than the coke on the catalyst with a high content of ZrO₂ at the surface as a coactive species (the 10Ni-10ZrO₂/Al₂O₃ (CI) catalyst). Thereby, the carbon deposition amount reduced with a decrease of the ZrO₂ content due to the lower effect of a poor Ni-ZrO₂ interaction providing to the smaller Ni particle sizes. Moreover, compared to the spent 10Ni-10ZrO₂/Al₂O₃ (CI) catalyst, the coke proportion on the spent 10Ni/ZrO₂ catalyst was lower. It can be ascribed to the desorption of α oxygen on the 10Ni/ZrO₂ catalyst, which practically eliminates the amount of deposited coke and inhibits filament carbon. Finally, the continuous weight loss of 7% appeared mostly at a temperature below 300 °C on the spent 10Ni-10ZrO₂/Al₂O₃ (SI) (ZrO₂ as a coactive component) catalyst (Figure 8d), correlating to the oxidation of the amorphous carbon. The deposited coke on the spent 10Ni-10ZrO₂/Al₂O₃ (SI) catalyst was found not only in the lesser amount but also in an easier removal characteristic compared to other catalysts. Considering the O₂-TPD and the H₂-TPR results, the reason for this easier removal is attributed to the high amount of oxygen mobility accompanied with the greater interaction between Ni and the ZrO₂-Al₂O₃ composite that enhances both the Ni distribution (resulting in smallest Ni particle sizes) and pathway of the vacancy sites for the oxygen transfer from ZrO₂ in ZrO₂-Al₂O₃ through coke on the Ni surface in the CRM process [27,28,49].

The coke deposition on the spent catalysts of CSCRM was determined by TPO results (Figure 9). In the TPO profiles of the carbon deposition on the spent 10Ni-1ZrO₂/Al₂O₃ (CI) catalyst and 10Ni-10ZrO₂/Al₂O₃ (SI) catalysts used in CSCRM, two peaks of carbon removal depended on the carbon types were observed. Peaks at low temperature (50 °C to 350 °C) and peaks at high temperature (500 °C to 750 °C) in the profiles correlate to the amorphous carbon and the filament carbon, respectively. The quantity of carbon deposition over the spent catalysts was calculated directly from the peak area. As a result, the coke formation appeared mostly as the filament carbon. The total amount of coke on the spent 10Ni-1ZrO₂/Al₂O₃ (CI) and the spent 10Ni-10ZrO₂/Al₂O₃ (SI) from CSCRM were 1.213 mmol·g⁻¹ and 1.571 mmol·g⁻¹, respectively. For comparison purpose, in CSCRM process, the coke resistance/removal of 10Ni-1ZrO₂/Al₂O₃ (CI) is more effective than those of the 10Ni-10ZrO₂/Al₂O₃ (SI). It can be ascribed to more oxygen surface from H₂O activation-dissociation promoted via weak and moderate basic sites on the catalyst [50].

3. Materials and Methods

3.1. Catalyst Preparation

The ZrO₂ support was synthesized via the precipitation method. The solution 0.4 M zirconyl chloride (ZrOCl₂, Sigma Aldrich, St. Louis, MS, USA) was used as a precursor, and a 2 M ammonium hydroxide (NH₄OH, 65% PanReac AppliChem ITW Reagents) solution was used as a precipitant. These two aqueous solutions were added by drops into a beaker with a controlled pH of 9.3. Then, the mixture was stirred thoroughly for 30 min at 60 °C, and the precipitate was aged overnight at the ambient temperature. The precipitate powder was vacuum filtered out and rinsed with distilled water until neutral (with a pH of approximately 7). Afterwards, it was dried overnight at 70 °C and calcined in the air at 750 °C for 5 h. The 10 wt.% Ni/ZrO₂ catalyst was prepared by impregnating it with a 10 wt.% of nickel (II) nitrate hexahydrate solution (Ni(NO₃)₂·6H₂O, 98%, Acros OrganicsTM) onto the ZrO₂ support followed by drying at 50 °C overnight and calcination in the air at 650 °C for 6 h.

The Al₂O₃ support was synthesized by the sol–gel method using aluminium isopropoxide (C₉H₂₁AlO₃, 98%, Acros OrganicsTM) and nitric acid (65% CARLO ERBA Reagents) as precursors. The gel product was dried at 50 °C overnight and calcined at 650 °C for 6 h with a heating rate of 3 °C min⁻¹. The 10 wt.% Ni-x wt.% ZrO₂/Al₂O₃ catalysts (denoted as 10Ni-xZrO₂/Al₂O₃ (CI); x = 1 and 10) were prepared by co-impregnation method. The aqueous solution of the desired amount of zirconium (IV) oxynitrate hydrate (ZrO(NO₃)₂·H₂O, Sigma Aldrich) mixed with Ni(NO₃)₂·6H₂O was added dropwise to the Al₂O₃ support. The resulting wet powder was followed by drying and calcination using the same conditions used for the 10Ni/ZrO₂ catalyst. To generate the different interactions of Ni-ZrO₂, the sequential impregnation (SI) method was employed to prepare the 10 wt.% Ni-10 wt.% ZrO₂/Al₂O₃ catalyst (denoted as 10Ni-10Zr/Al₂O₃(SI)). First, Al₂O₃ support was impregnated with an aqueous solution of zirconium (IV) oxynitrate hydrate followed by drying and calcination, using the condition for 10Ni-xZr/Al₂O₃ (CI). Second, 10 wt.% ZrO₂/Al₂O₃ was impregnated with a Ni(NO₃)₂·6H₂O solution followed by drying and calcination, using similar condition as with 10Ni/ZrO₂.

3.2. Catalysts Characterization

The X-ray diffraction (XRD) patterns of the reduced catalysts were collected on a Bruker AXS Model D8-Discover (Bruker AXS, Billerica, MA, USA) with a Cu Kα radiation (λ = 0.154 nm), operating at 40 kV and 40 mA and with a scanning range between 20° to 70° with a step width of a scanning speed of 0.02° min⁻¹.

The metal particle size (d, nm) and the metal dispersion (D_m, %) were indirectly determined by the hydrogen temperature desorption (H₂-TPD) technique performed in a BELCAT-basic system (BEL JAPAN, INC., Osaka, Japan). At the beginning, the catalyst (0.05 g) was preactivated at 600 °C in pure H₂ flow for 2 h, followed by cooling to 100 °C in He flow. Then, the catalyst was isothermally chemisorbed on the surface with pure H₂ at 100 °C for 1 h and cooled to 40 °C in He flow. The quantity of H₂ desorbed was measured by passing He flow through the catalyst from 50 °C to 850 °C at the rate of 10 °C min⁻¹. The metal particle size (Equation (1)) and the metal dispersion (Equation (2)) were calculated based on the stoichiometry of the hydrogen monolayer chemisorption uptake.

$$\%D_m=\frac{\left(\frac{{\mathrm{V}}_{\mathrm{chem}}}{\mathrm{22,414}}\right)\times \mathrm{SF}\times \mathrm{MW}}{\left(\frac{\mathrm{m}\times \mathrm{w}}{100}\right)}\times 100$$

(1)

$$d=\frac{6000}{\frac{\left(\frac{{\mathrm{V}}_{\mathrm{chem}}}{\mathrm{22,414}}\right)\times \mathrm{SF}\times 6.02\times {10}^{23}\times \mathsf{\rho }\times {\mathsf{\sigma }}_{\mathrm{m}}\times {10}^{-18}}{\left(\frac{\mathrm{m}\times \mathrm{w}}{100}\right)}}$$

(2)

where V_chem is the amount of desorbed hydrogen (cm³), SF is a stoichiometry factor (Ni mol:H₂ mol in the H₂ chemisorption = 2), MW is an atomic weight of metal (g mol⁻¹), m is a sample weight (g), w is a wt.% of supported metal content, σ_m is a cross-sectional area of one Ni atom (the area occupied by Ni atom = 0.0649 nm²), and ρ is a density of metal (g cm⁻³).

The surface area and pore properties of the catalysts were determined using the N₂ adsorption–desorption isotherm at the boiling temperature of liquid N₂ (−196 °C) on a BELSORP-mini II. Prior (BEL JAPAN, INC., Osaka, Japan). Prior to the experiment, the catalyst was degassed at 350 °C under a N₂ flow for 4 h. The surface area and the pore volume were evaluated by the Brunauer–Emmett–Teller (BET) analysis method. The pore size distribution was calculated based on the Barett–Joyner–Halenda (BJH) method.

The reducibility of the catalysts was measured by the temperature-programmed reduction in hydrogen (H₂-TPR) employing a BELCAT-basic system apparatus (BEL JAPAN, INC., Osaka, Japan) equipped with a thermal conductivity detector (TCD). Prior to measurement, a 0.05 g catalyst was degassed at 220 °C for 1 h and cooled to 40°C in Ar. Then, the sample was reduced in a 5 vol.% H₂/Ar flow (30 mL min⁻¹) from 50 °C to 800 °C at the ramping rate of 10 °C min⁻¹.

The basicity of the reduced catalysts was characterized by the temperature-programmed desorption in carbon dioxide (CO₂-TPD) using a BELCAT-basic system (BEL JAPAN, INC., Osaka, Japan) apparatus. Prior to the TPD experiment, the catalyst (0.05 g) was pre-treated by heating at 220 °C for 1 h in He. Then, the catalyst was cooled to 40°C under He flow and reduced in situ at 620°C for 2 h with pure H₂ flow (30 mL min⁻¹) followed by cooling to 40°C in He flow. Subsequently, adsorption of CO₂ was performed in pure CO₂ flow (30 mL min⁻¹) at 40°C for 1 h. The physisorbed CO₂ was removed with He flow. Afterward, the sample was heated from 50 °C to 850 °C at a rate of 10 °C min⁻¹ in He flow, and the desorbed CO₂ was detected.

The oxygen mobility of the catalysts was evaluated by the temperature-programmed desorption in oxygen (O₂-TPD) on a BELCAT-basic system (BEL JAPAN, INC., Osaka, Japan). Before the O₂-TPD testing, the 0.05 g catalyst was reduced in-situ at 600 °C in a pure H₂ flow for 1.5 h and cooled to 200 °C in flowing Ar. Afterward, a pure O₂ gas was introduced to be adsorbed on the catalyst surface at 200 °C for 1.5 h. Then, the catalyst was purged in Ar flow to flush weakly adsorbed O₂ on the surface. After purging, the catalyst was heated from 50 °C to 850 °C in Ar flow (30 mL min⁻¹) at the constant ramping rate of 10 °C min⁻¹. The desorbed O₂ was monitored by TCD.

The quantity and types of deposited carbon over the spent catalysts from the CRM test were investigated by the mean of the thermogravimetric analysis (TGA) on a METTLER TOLEDO thermogravimetric analyzer (METTLER TOLEDO, Columbus, OH, USA). A 0.012 g sample of spent catalyst was combusted in an air stream from 40 °C to 800 °C at a heating rate of 10 °C min⁻¹.

The amount and types of the coke accumulated on the spent catalysts during the CSCRM test reaction were interpreted by temperature-programmed oxidation (TPO) via a BELCAT-basic system (BEL JAPAN, INC., Osaka, Japan). Before the TPO measurement, in Ar flow (30 mL min⁻¹), a 0.050 g of the spent catalyst was pretreated at 220 °C for 2 h followed by cooling to 40 °C. Afterwards, the spent catalyst was oxidized in a 5 vol.% O₂/Ar flow (30 mL min⁻¹) with a ramping rate of 10 °C min⁻¹ from 50 °C to 800 °C and the O₂ consumption was detected.

3.3. Catalytic Tests

The CRM reaction was performed in a tubular fixed-bed reactor at 620 °C for 10 h. Before the test, the catalyst (0.200 g) was in situ reduced in a pure H₂ (purity 99.999%) at 620 °C for 6 h. Then, the feed molar composition of CH₄/CO₂/N₂ = 3/5/4 at a total flow rate of 60 mL min⁻¹ was introduced to the reactor. The outlet gas was analyzed via on-line gas chromatograph (Agilent GC7890A Agilent, Santa Clara, CA, USA) equipped with the TCD. The CSCRM reaction was conducted for 6 h using the same system and operating conditions of CRM with the additional steam (the CH₄:CO₂:H₂O:N₂ ratio of 3:5:2.4:4 with overall flow rate of 72 mL min⁻¹). The catalyst performance was evaluated in terms of a CH₄ conversion (Equation (3)), CO₂ conversion (Equation (4)) and H₂/CO ratio (Equation (5)) as expressed in the following equations:

$${\mathrm{X}}_{{\mathrm{CH}}_4}(\%)=\frac{{({\mathrm{CH}}_4)}_{\mathrm{in}}-{({\mathrm{CH}}_4)}_{\mathrm{out}}}{{({\mathrm{CH}}_4)}_{\mathrm{in}}}\times 100$$

(3)

$${\mathrm{X}}_{{\mathrm{CO}}_2}(\%)=\frac{{({\mathrm{CO}}_2)}_{\mathrm{in}}-{({\mathrm{CO}}_2)}_{\mathrm{out}}}{{({\mathrm{CO}}_2)}_{\mathrm{in}}}\times 100$$

(4)

$${\mathrm{H}}_2/\mathrm{CO} \mathrm{ratio}=\frac{{\left({\mathrm{H}}_2\right)}_{\mathrm{out}}}{{\left(\mathrm{CO}\right)}_{\mathrm{out}}}$$

(5)

4. Conclusions

This research is an in-depth examination of the role of ZrO₂ as various components in the Ni-based catalysts for CRM. The 10 wt.% Ni/ZrO₂ catalyst (10Ni/ZrO₂) was prepared by the impregnation method. The 10 wt.% Ni/Al₂O₃ catalyst with a 1 wt.% ZrO₂ promoter was prepared by the co-impregnation method (10Ni-1Zr/Al₂O₃(CI)). The 10 wt.% Ni/Al₂O₃ catalysts with a 10 wt.% ZrO₂ coactive component were prepared by the co-impregnation (10Ni-10Zr/Al₂O₃(CI)) and sequential impregnation methods (10Ni-10Zr/Al₂O₃(SI)) in order to study the effect of the different interactions in the catalyst. All catalysts were characterized and tested for CRM at 620°C under atmospheric pressure. The influence of ZrO₂ on the catalytic activity and the coke resistance mainly relies on the ZrO₂ content, Ni-ZrO₂ interaction, basicity, and oxygen mobility of catalysts. The lowest H₂/CO ratio (highest CO selectivity) were obtained from 10Ni-1Zr/Al₂O₃(CI) due to the weak and medium basic sites that promote the CO₂ activation-dissociation. Normally, ZrO₂ has a poor interaction with Ni that causes the low Ni dispersion with the large Ni particle sizes. Therefore, the 10Ni/ZrO₂ catalyst exhibited the lowest activity (CH₄ conversion of 40% and CO₂ conversion of 39%) due to the large Ni particle sizes from the Ni-ZrO₂ poor interaction. Because of the smallest number of total oxygen mobility, the relatively high amount of carbon (the weight loss of 22% in TGA curve of the spent catalyst) was detected on this catalyst. The 10 wt.% Ni/Al₂O₃ catalysts with the lesser proportion of ZrO₂ (10Ni-1Zr/Al₂O₃(CI) and 10Ni-10Zr/Al₂O₃(CI)) illustrated the greater Ni dispersion resulting in higher activity (CH₄ conversion of 51% to 54% and CO₂ conversion of 45% to 50%). Although all catalysts provided the similar number of oxygen mobility. The high activity with a most practical carbon resistance (the weight loss of 7%in TGA curve of the spent catalyst) was successfully developed in 10Ni-10Zr/Al₂O₃(SI). The reason is explained by the great amount of oxygen mobility associated with the intimate interaction of the Ni and ZrO₂-Al₂O₃ composite in this catalyst enlarges the transferring oxygen from ZrO₂ to Ni, which increases the oxidation of carbon on the Ni surface via vacancies.

According to the basicity and oxygen mobility effect from CRM results, CSCRM at the similar condition of CRM with the addition of steam was performed on the 10Ni-1Zr/Al₂O₃(CI) catalyst and the 10Ni-10Zr/Al₂O₃(SI) catalyst. The higher H₂/CO ratio (2.1) with the lower carbon deposition (1.213 mmol·g⁻¹) was monitored on the 10Ni-1Zr/Al₂O₃(CI) catalyst. It demonstrated that the basic sites enhance the H₂O activation-dissociation greater than the oxygen vacancy. However, this catalytic behavior causes a loss of active sites for CH₄ activation-dissociation. Overall, the results of our study offer useful information for the further Ni-based catalyst development.