Modification Strategies of Ni-Based Catalysts with Metal Oxides for Dry Reforming of Methane

Abstract

Syngas generated from the catalytic dry reforming of methane (DRM) enables the downstream production of H₂ fuel and value-added chemicals. Ni-based catalysts with metal oxides, as both supports and promoters, are widely applied in the DRM reaction. In this review, four types of metal oxides with support confinement effect, metal-support interaction, oxygen defects, and surface acidity/basicity are introduced based on their impacts on the activity, selectivity, and stability of the Ni-based catalyst. Moreover, the structure–performance relationships are discussed in-depth. Finally, conclusive remarks and prospects are proposed.

1. Introduction

CO₂ and CH₄ are two main sources of greenhouse gases released from industry and human activities, which adversely influences the natural environment and biological diversity [1,2,3,4]. A promising strategy to address the above issues is conversion of both gases into valuable products, such as syngas—a mixture of CO and H₂, which can be further transformed into pure H₂ via separation membranes or value-added chemicals (e.g., acetic acid, methanol, oxyalcohol, dimethyl ether, and long-chain hydrocarbons) via combination reactions [5,6,7,8,9,10].

Due to the strong endothermic forward reaction, a high temperature (600–800 °C) is usually needed to activate the reactant molecules. The utilization of catalysts certainly lowers the activation barrier and changes the reaction pathway. Noble metals exhibit a high conversion and anti-coking property; however, the limited reserve and high cost hinder their large-scale applications. In comparison, Ni-based catalysts are low cost and comparable catalytic activities, thus presenting a competitive application potential in the DRM reaction [11,12,13]; however, owing to the intensive reaction conditions, metal sintering is more likely to occur than in mild conditions, resulting in a loss of surface area and an activity drop. Additionally, coke formation from CH₄ decomposition and CO disproportionation easily covers the active sites and blocks the reactor [14,15,16,17]; therefore, modifications are necessary to develop a highly active and robust Ni-based catalyst.

Since CO₂ can be adsorbed and activated at the metal and oxygen ions, metal oxides (e.g., basic oxides, rare earth metal oxides, transition metal oxides, and mixed oxides) with oxygen defects and surface basicity are widely applied as supports or promoters [18,19,20,21,22]. In particular, strong basicity enhances the CO₂ conversion efficiency and facilitates the dissociation of CO₂ into CO and O radicals, leading to immediate carbon removal [23,24,25]. On the other hand, the existence of oxygen defects accelerates the surface oxygen mobility and lattice oxygen migration, thus realizing an effective coke elimination [26,27,28,29,30]. In addition to the CO₂ conversion, CH₄ prefers to adsorb onto the Ni surface and undergoes the activation to produce CH_x and H atoms. Thus, a highly dispersed Ni particle with a small size and large exposed area favors a fast CH₄ conversion [31,32,33]. To ensure a well-distributed and unchanged particle size, metal oxides can be added to interact strongly with Ni sites by forming the solid solution or spinel phase. Moreover, Ni nanoparticle migration could be retarded by the physical steric hindrance provided by the ordered porous or hierarchical metal oxide structures. Based on the above discussion, metal oxides play a crucial role in affecting the physicochemical properties and catalytic performances of Ni-based catalysts in the DRM reaction.

In recent years, reviews on the catalysts for the DRM reaction mainly focus on specific metal or support materials, such as Ni-based catalysts [34], transition metal catalysts [35], Ni/Al₂O₃ catalysts [36], silica-based catalysts [37], Lanthanoid-containing Ni-based catalysts [38], metal carbides [39] and alloy catalysts [40]. Very few works are focused on the applications of metal oxides in Ni-based catalysts relating to the modification impacts on the size, morphology, surface, and interface properties that are based on the catalytic activities and anti-deactivation behaviors in the DRM reaction; therefore, this review summarizes the state-of-art developments of Ni-based catalysts regarding the modification strategies (support confinement, metal–support interaction, oxygen defects, and surface acidity/basicity) of metal oxides (basic oxides, rare earth metal oxides, transition metal oxides, and mixed oxides) on the activity, selectivity, stability, and deactivation resistance in the DRM reaction. Moreover, the reaction and deactivation mechanisms of Ni catalysts in the DRM reaction are illustrated in detail. In addition, the structure–performance relationships are critically discussed in depth. Finally, conclusive remarks and prospects are proposed.

2. Reaction and Deactivation Mechanisms

2.1. Catalytic Reaction Mechanism

In the DRM reaction, an equal amount of CO₂ and CH₄ are transformed into syngas. Due to the strong C-H bonds in CH₄, and the highest valence state of C in CO₂ [41,42], the DRM reaction possesses a highly endothermic nature, as presented below [43]:

$${\mathrm{CH}}_4+{\mathrm{CO}}_2\to 2\mathrm{CO}+2{\mathrm{H}}_2\mathsf{\Delta }{\mathrm{H}}_{298}^0=247.3\mathrm{kJ}/\mathrm{mol}$$

(1)

As for the reaction mechanisms, mono-functional (on metals only) and bi-functional mechanisms (on metals and supports) are proposed as follows [44]. In the mono-functional pathway (Equations (2)–(10)), the dissociation of CH₄ and CO₂ takes place simultaneously, producing CO, O, H, and CH_x species. Subsequently, CH_x combines with O atoms to form CO and H atoms, whereas two H atoms combine to generate H₂ molecules. In the meantime, hydroxyl groups are produced by the combination of O and H atoms. After hydrogenation, -OH is converted to H₂O as a side product.

$${\mathrm{CH}}_4\leftrightarrow {\mathrm{CH}}_{\mathrm{x}\left(\mathrm{Ni}\right)}+\left(4-\mathrm{x}\right){\mathrm{H}}_{\left(\mathrm{Ni}\right)}$$

(2)

$$2{\mathrm{H}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_{2\left(\mathrm{gas}\right)}$$

(3)

$${\mathrm{CO}}_{2\left(\mathrm{gas}\right)}\leftrightarrow {\mathrm{CO}}_{2\left(\mathrm{Ni}\right)}$$

(4)

$${\mathrm{CO}}_{2\left(\mathrm{gas}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{Ni}\right)}+{\mathrm{O}}_{\left(\mathrm{Ni}\right)}$$

(5)

$${\mathrm{O}}_{\left(\mathrm{Ni}\right)}+{\mathrm{CH}}_{\mathrm{x}\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{Ni}\right)}+{\mathrm{xH}}_{\left(\mathrm{Ni}\right)}$$

(6)

$${\mathrm{CO}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{gas}\right)}$$

(7)

$${\mathrm{H}}_{\left(\mathrm{Ni}\right)}+{\mathrm{O}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{HO}}_{\left(\mathrm{Ni}\right)}$$

(8)

$${\mathrm{H}}_{\left(\mathrm{Ni}\right)}+{\mathrm{HO}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{gas}\right)}$$

(9)

$${\mathrm{OH}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{HO}}_{\left(\mathrm{support}\right)}$$

(10)

In the bi-functional mechanism (Equations (11)–(26)) [45], the activation of CO₂ occurs on the support, whereas that of CH₄ takes place at the Ni surface. In addition, O atoms can be generated from both the -OH decomposition at the Ni sites and the reaction between CO₂ and O²⁻ ions at the support. Subsequently, CO can be produced from both the oxidation of CH_x at the metal surface and the dissociation of HCO₂⁻ from the support derived from the hydrogenation of CO₃²⁻ and HCO₃⁻. Instead of the combination of -OH and H atoms in the mono-functional mechanism, two -OH groups react with each other to form the side product H₂O.

CH₄ activation at Ni sites:

$${\mathrm{CH}}_4\leftrightarrow {\mathrm{CH}}_{\mathrm{x}\left(\mathrm{Ni}\right)}+\left(4-\mathrm{x}\right){\mathrm{H}}_{\left(\mathrm{Ni}\right)}$$

(11)

$$2{\mathrm{H}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_{2\left(\mathrm{gas}\right)}$$

(12)

$${\mathrm{OH}}_{\left(\mathrm{support}\right)}\leftrightarrow {\mathrm{OH}}_{\left(\mathrm{Ni}\right)}$$

(13)

$${\mathrm{OH}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_{\left(\mathrm{Ni}\right)}+{\mathrm{O}}_{\left(\mathrm{Ni}\right)}$$

(14)

$${\mathrm{OH}}_{\left(\mathrm{Ni}\right)}+{\mathrm{H}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{gas}\right)}$$

(15)

$${\mathrm{O}}_{\left(\mathrm{Ni}\right)}+{\mathrm{CH}}_{\mathrm{x}\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{Ni}\right)}+{\mathrm{xH}}_{\left(\mathrm{Ni}\right)}$$

(16)

$${\mathrm{CO}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{gas}\right)}$$

(17)

CO₂ activation on support sites:

$${\mathrm{CO}}_{2\left(\mathrm{gas}\right)}\leftrightarrow {\mathrm{CO}}_{2\left(\mathrm{support}\right)}$$

(18)

$${\mathrm{CO}}_{2\left(\mathrm{support}\right)}+{\mathrm{O}}_{\left(\mathrm{support}\right)}^{2-}\leftrightarrow {\mathrm{CO}}_{3\left(\mathrm{support}\right)}^{2-}+{\mathrm{O}}_{\left(\mathrm{Ni}\right)}$$

(19)

$${\mathrm{CO}}_{2\left(\mathrm{support}\right)}+{\mathrm{OH}}_{\left(\mathrm{support}\right)}^{-}\leftrightarrow {\mathrm{HCO}}_{3\left(\mathrm{support}\right)}^{-}$$

(20)

$${\mathrm{H}}_{\left(\mathrm{Ni}\right)}\leftrightarrow {\mathrm{H}}_{\left(\mathrm{support}\right)}$$

(21)

$${\mathrm{CO}}_{3\left(\mathrm{support}\right)}^{2-}+2{\mathrm{H}}_{\left(\mathrm{support}\right)}\leftrightarrow {\mathrm{HCO}}_{2\left(\mathrm{support}\right)}^{-}+{\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{support}\right)}$$

(22)

$${\mathrm{HCO}}_{2\left(\mathrm{support}\right)}^{-}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{support}\right)}+{\mathrm{OH}}_{\left(\mathrm{support}\right)}^{-}$$

(23)

$${\mathrm{CO}}_{\left(\mathrm{support}\right)}\leftrightarrow {\mathrm{CO}}_{\left(\mathrm{gas}\right)}$$

(24)

$$2{\mathrm{OH}}_{\left(\mathrm{support}\right)}^{-}\leftrightarrow {\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{support}\right)}+{\mathrm{O}}_{\left(\mathrm{support}\right)}^{2-}$$

(25)

$${\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{support}\right)}\leftrightarrow {\mathrm{H}}_2{\mathrm{O}}_{\left(\mathrm{gas}\right)}$$

(26)

2.2. Deactivation Mechanisms

2.2.1. Coking

In the DRM reaction, coke formation can be caused by CH₄ cracking (Equation (27)) and CO disproportionation (Equation (28)) [46,47], which covers the Ni surface and blocks catalyst pore structures [46]. The impacts of carbon deposits depend on the structure. α-carbon is usually amorphous and mostly formed at the Ni sites with a small size and high dispersion, which is easily gasified during the reaction. In comparison, large Ni particles favor the formation of β-carbon, which is not active as α-carbon, and is possibly transformed into γ-carbon. As the most inert and ordered phase, γ-carbon is in the form of graphite and exerts the most detrimental effect on catalytic activity and stability [48,49].

$${\mathrm{CH}}_4\leftrightarrow \mathrm{C}+2{\mathrm{H}}_2\mathsf{\Delta }{\mathrm{H}}_{298}^0=74.9\mathrm{kJ}/\mathrm{mol}$$

(27)

$$2\mathrm{CO}\leftrightarrow \mathrm{C}+{\mathrm{CO}}_2\mathsf{\Delta }{\mathrm{H}}_{298}^0=-172.2\mathrm{kJ}/\mathrm{mol}$$

(28)

The factors determining the coke formation include the particle size, metal–support interaction (MSI), temperature, space-time, and the surface property of catalysts. For example, a large particle size favors the carbon deposition, especially in the form of inert graphitic carbons encapsulating the active sites. As for the MSI, a strong MSI mostly facilitates the coke removal, owing to the interface synergy; however, an excessively strong MSI might lead to the coverage of carbon nanotubes on the active centers. Moreover, a high reaction temperature probably inhibits the carbon formation because the CO disproportionation is greatly prevented; however, the catalyst should be carefully designed to gasify the carbon derived from methane decomposition. Moreover, at a low space-time, a serious deposition of filamentous and encapsulated carbons occurs; at a high space-time, more encapsulated cokes are eliminated, especially with the temperature increase. Finally, a basic and oxygen-deficient catalyst surface enables high CO₂ adsorption and activation with a rapid oxygen migration, which promotes the coke gasification.

2.2.2. Sintering

In high temperature conditions, a large Ni particle is generated following either Ostwald ripening or particle migration mechanism. In Ostwald ripening (atomic migration), which occurs at a relatively higher temperature and has a longer duration, the particle emitted from the metal is captured by another one to form a larger size. In comparison, in particle migration, which prefers lower temperatures, two particles move on the surface of the support and combine to produce a larger particle [50]. Apart from the metal sintering, support sintering is possibly caused by phase transformation or the evaporation/condensation of volatile molecules/atoms [46]. For example, θ-phase Al₂O₃ was transformed into α-phase when the temperature increased from 1000 to 1125 °C, resulting in a reduction in surface area [51]. According to the authors, the reduction of surface area was due to the micropore collapse and dense hcp phase formation. Notably, the transition temperature change initiated by the spinel formation might also be a driving force of the phase transformation.

In addition to the temperature effect, an ordered mesoporous support or a hierarchical structure offers a confinement effect, so as to hinder the random movement of metal particles; however, if the precursor concentration is very high, some metal ions fail to enter the pores and accumulate on the support surface, which easily agglomerate with each other during the subsequent high temperature treatment. Moreover, a fast flow rate of the reducing stream may eliminate the heat generated on the catalyst bed, thus enhancing the metal dispersion.

2.2.3. Poisoning

Due to the existence of impurities (e.g., sulfur species) in the raw feed, metallic Ni is easily converted to sulfides via a reaction with H₂S (Equation (29)) [52]. As a result, the adsorption and activation of CH₄ are inhibited. In addition to the sulfidation mechanism, side reactions or carbon formation might be favored under the coverage of sulfur species.

$$\mathrm{Ni}+{\mathrm{H}}_2\mathrm{S}\leftrightharpoons \mathrm{Ni}-\mathrm{S}+{\mathrm{H}}_2$$

(29)

To address the poisoning issues, the increase in reaction temperature is effective to break the Ni-S bond; in other reports, however, a high temperature causes irreversible S layers to form on the active sites, whereas a low temperature favors polysulfide formation, which is easily removed by H₂. As well as the temperature, the introduction of O₂ or steam could alleviate this poisoning effect; however, the oxidation of Ni metals into oxides or sulfates leads to the loss of active sites [50,52]. Moreover, the doping of noble metals may protect the Ni active sites from being poisoned.

3. Impacts of Metal Oxides

To promote the adsorption and activation of reactant molecules and prevent the deactivation of Ni catalysts, metal oxides can be added as either a support or a modifier so as to confine the Ni particles within the pores or core–shell structures, to anchor the Ni metals via a strong metal-support interaction (MSI), to improve the CO₂ adsorption at the abundant basic sites, and to facilitate the oxygen mobility so as to oxidize the carbon and sulfides. Commonly used metal oxides include basic oxides, rare earth metal oxides, transition metal oxides, and mixed oxides. In the following sections, the impacts of metal oxides on the physicochemical properties and catalytic performances of Ni catalysts in the DRM reaction will be discussed in four categories (support confinement, metal–support interaction, oxygen defects, and surface acidity/basicity).

3.1. Support Confinement

Many oxides with a porous structure exhibit the support confinement effect on the dispersion of metal sites. For example, ordered mesoporous silica materials provide an abundant confined space for inhibiting the metal migration. Benefiting from the ordered channels of SBA-15, negligible metal sintering occurred over a 40 h reaction [53]; however, the other two silica supports (MCM-41 and KIT-6) suffer from the poor structure stability and micropore blocking, resulting in a low activity but high coke formation [54,55]. Similarly, with the silica materials, a metal oxide with ordered porous structures can effectively accommodate the Ni species and prevent the migration of Ni particles, thus improving the anti-sintering property and maintaining the activation of CH₄. Compared with non-porous Al₂O₃, the 2D hexagonal Al₂O₃ support prepared by the “one-pot method” possessed abundant mesopores, offering a confined space for accommodating Ni–Fe alloy particles. The resultant highly dispersed Ni–Fe active sites stabilized the conversion efficiency with negligible metal growth or carbon deposition after 13 h during the DRM reaction at 700 °C (

Table 1. A summary of representative Ni-based catalysts modified with metal oxides for DRM reaction.

Catalyst	Temperature (°C)	CH4/CO2	CH4 Conversion (%)	CO2 Conversion (%)	H2/CO	Remark	Ref
Fe5%Ni5%Al2O3	700	1.8:1	50	89	1.1	NiFe alloy particles were confined within the ordered mesoporous Al2O3 frameworks.	[33]
Ni/Al2O3	850	1:1	99	96	0.89	Cubic and mesoporous Al2O3 confined Ni particles and exhibited strong sintering resistance over 210 h.	[31]
Ni–CeO2/SiO2	700	1:1	77	85	0.95	High Ni dispersion on CeO2 and abundant Ni–CeO2 interfaces enhanced the coke resistance.	[58]
Ni/La2O3	650	1:1	30	68	0.9	Ni agglomeration was alleviated over 50 h due to the La2O3 mesopore confinement.	[59]
Ni/Y2O3–ZrO2	700	1:1	67	71	0.85	Ni particle size was reduced over 8 h due to re-dispersion and strong MSI with Y2O3 doping.	[60]
LaNi0.34Co0.33Mn0.33O3	800	1:1.05	94	92.5	1.15	MnO enhanced the interaction between the metal and La2O3 support.	[61]
Ni/Al2O3–La2O3CO3	650	1:1	61	65	0.85	La2O2CO3 increased the number of Ni active sites by inhibiting the NiAl2O4 formation.	[32]
Ni/MgO–ZrO2	800	1:1	68	75	0.89	ZrO2 tuned the MSI in Ni/MgO and enhanced the reducibility.	[62]
1.5CeO2−x–NSNT	750	1:1	82	88	0.91	Ni silicate nanotubes (NSNTs) reacted with CeO2 to produce Ce3+ and oxygen defects, inhibiting the coke formation.	[28]
Ce0.70La0.20Ni0.10O2−δ	750	1:1	73	84	0.88	Oxygen defects and La2O2CO3 contributed to the improved coke resistance.	[29]
La(Co0.1Ni0.9)0.5Fe0.5O3	750	1:1	70	80	0.89	Co partial substitution generated oxygen vacancies and enhanced the amount of surface oxygen species.	[30]
La0.4Ce0.6Ni0.5Fe0.5O3	750	1:1	62	72	0.91	Reversible redox reaction and undercoordinated B-site cations increased oxygen defect concentration.	[63]
Co–Ni/Sc-SBA–15	700	1:1	72.5	79	0.91	More basic sites were generated with the Sc doping, reducing the inert carbon amount.	[19]
SmCoO3	800	1:1	93	90	1.1	Co activated CH4 and Sm2O2CO3 removed carbon intermediates.	[24]
Ni/Al2O3–MgO	800	1:1	40	52	0.7	MgO enhanced the concentration of medium and strong basic sites, thus alleviating the encapsulated carbon formation.	[25]
Y-doped Ni–Mg–Al double-layered hydroxides	700	1:1	76.2	80.8	0.92	Weak and medium basic sites were introduced by Y2O3, promoting reversible CO2 adsorption and desorption.	[64]

) [33]. With CeO₂ doped into the framework, the CeO₂–Al₂O₃ mesoporous support structures prevented Ni nanoparticles from migration and agglomeration, keeping the active sites exposed to the reactant molecules. At 700 °C, the CH₄ conversion was as high as 78% over 80 h for the CeO₂-modified catalyst [56]. As well as the 2D morphology, by carefully controlling the pH value in synthesis, the formed mesoporous Al₂O₃ with a cubic phase provided steric hindrance to confine the Ni metals within the support matrix. Owing to the strong resistance against coking and sintering, an excellent conversion of CO₂ and CH₄ were obtained (97% and 99%, respectively), with only 5% coke formation over a 210 h DRM reaction at 700 °C (Table 1) [31]. To simultaneously improve the Ni dispersion and mass diffusion, a hierarchically porous Al₂O₃ structure with bimodal pore distribution (macropore structure and mesoporous channels) confined the Ni nanoparticles, and allowed the fast diffusion of intermediates and products, thus facilitating the CH₄ activation and carbon removal [57].

Apart from Al₂O₃, CeO₂ can provide a confinement effect on the Ni particles. Via the one-step colloidal solution combustion method, highly dispersed Ni nanoparticles were formed due to the spatial confinement by CeO₂. During the reaction, Ni migration and agglomeration were greatly inhibited, producing a particle size of less than 5 nm. Owing to the small Ni metal size and abundant Ni–CeO₂ interfaces, only 1.8% carbon deposition was observed over a 20 h DRM reaction at 700 °C (Table 1) [58]. Similarly, mesoporous La₂O₃ was synthesized with SBA-15 as the hard template. As shown in Figure 1, large Ni particles were formed with an average size of 13.7 nm; in comparison, much smaller Ni nanoparticles with a size of 4.6 nm were highly dispersed and confined within the mesopores of La₂O₃. After 50 h of the DRM reaction at 650 °C, serious Ni agglomeration took place in the 5Ni/La₂O₃-n catalyst, such that the particle size was increased to 17.1 nm, whereas for the mesoporous La₂O₃ supported Ni catalyst, a lower degree of metal growth was presented, in that the spent Ni size was 5.5 nm. As well as the abundant active sites for CH₄ activation, a higher dispersion and smaller size of Ni provided more Ni–La₂O₃ interfaces, where CO₂ adsorption, bidentate carbonate formation, and coke removal were facilitated (Table 1) [59]. Different from a mesoporous structure, as a protective layer, ZrO₂ was proven effective in confining the Ni particles. By coating a porous ZrO₂ shell onto Ni, the metal phase agglomeration was significantly prevented, leading to a much smaller size (6 vs. 170 nm) over 20 h during the DRM reaction at 700 °C [21].

3.2. Metal–Support Interaction

Apart from the physical steric hindrance from the support via confinement effect, the chemical interaction between the Ni and support strengthens the contact at the interface and generates a synergy, thus alleviating the Ni agglomeration and promoting the activation of reactants. For example, unlike other reports where spinel phases are prone to deactivating the catalysts, in Jabbour’s study, the NiAl₂O₄ spinel phase was formed under air calcination at 700 °C, which enhanced the interaction between Ni and Al₂O₃, producing a small Ni size of only 5 nm. As mentioned earlier, amorphous carbon species prefer to form at small metal sites, which are easily removed during the reaction; therefore, the carbon deposition was as low as 3.8% over the 20 h DRM reaction at 700 °C. Moreover, the CO₂ and CH₄ conversions were both enhanced up to 85.4% and 77.6%, respectively, owing to abundant Ni active sites and large specific surface area [33]. In another work, an interesting finding was presented, stating that Ni metals were partially extracted from the NiAl₂O₄ spinel phase, and the resulting Ni active sites were likely anchored by the deficient Ni_xAl_yO_z structure (defective Al₂O₃–NiO solid solution) with multiple Ni²⁺ defects, and thus they favored the strong MSI and coke resistance. Compared with 37% carbon deposition in the controlled Ni/Al₂O₃, the partially reduced NiAl₂O₄ exhibited a much lower coke formation (8%) over the 100 h DRM reaction at 700 °C [65]. In addition to the NiAl₂O₄ spinel formation, the MSI could be improved when Y₂O₃ was doped in the Ni/ZrO₂. In particular, the reduction temperature (β-peak) shifted to higher temperature regions and the total peak intensity decreased compared with the undoped Ni/ZrO₂ (0.37 vs. 0.62 mmol H₂/g), which indicated that the enhanced MSI derived from the formation of the NiO-ZrO₂ or NiO-Y₂O₃ solid solution. Owing to the strong MSI and Ni re-dispersion, the Ni particle size in the Y₂O₃-promoted Ni/ZrO₂ catalyst decreased from 16 nm to 10 nm, whereas the pristine Ni/ZrO₂ catalyst exhibited a severe metal growth from 12 nm to 24 nm over the 8 h DRM reaction at 700 °C. Moreover, the coke formation of the Y₂O₃-modified sample was only 1.0%, much smaller than that of the unmodified counterpart (3.7%) (Table 1) [60]. As well as Y₂O₃, the MSI was strengthened when the Ni/C catalyst was doped with CeO₂. The average Ni particle size was reduced from 31.1 nm to 27.6 nm and negligible metal growth was presented after the reaction (27.6 nm vs. 27.9 nm) [66]. The enhanced MSI was also observed when CeO₂ was added into Ni/SiO₂, leading to the formation of smaller Ni particles [67]. The promotional effect of MSI was discussed in depth based on the Ni/La₂O₃ catalyst where the CO₂ activation and coke removal were improved [59]. In detail, as shown in Figure 2, the adsorption energy of CO₂ on the pure La₂O₃ support to form monodentate carbonate was similar to the adsorption energy generated when producing bidentate carbonate (−0.94 eV vs. −1.05 eV). In comparison, the CO₂ adsorption energy produced when forming bidentate carbonate at the Ni–La₂O₃ interface was −2.64 eV, which is much lower than that generated when forming monodentate carbonate (−0.12 eV), thus suggesting a greater potential of producing bidentate carbonate upon CO₂ activation at the interface, rather than the support. Based on the in situ DRIFTS measurement during the adsorption of CH₄, which reacted with the surface carbonates, the intensity of monodentate carbonate was almost unchanged, whereas that of bidentate carbonate showed a gradual decrease, indicating the consumption of carbon intermediates and suppression of cokes at the bidentate carbonate sites. Since bidentate carbonate prefers to form at the Ni–La₂O₃ interface, a strong MSI in Ni/La₂O₃ favored the generation of abundant interfaces, thus potentially reducing the coke formation [59].

In addition to the rare earth metal oxides, the in situ formed MnO in the Co,Mn-co-doped LaNi_0.34Co_0.33Mn_0.33O₃ perovskite structure, strengthened the interaction with both Ni and La₂O₃, thus promoting the coke gasification at the Ni surface. On the contrary, the controlled sample without MnO possessed a relatively weak MSI and suffered from the Ni detachment by the carbon species, subsequently retarding the coke elimination with the nearby O atoms (Figure 3a). In addition, compared with co-doped and unmodified counterparts, LaNi_0.34Co_0.33Mn_0.33O₃ exhibited a higher H₂/CO ratio over the 14 h DRM reaction at 800 °C, which could be attributed to the rapid CO₂ conversion and inhibited RWGS reaction at the surface (Figure 3b) (Table 1) [61].

However, an excessively strong MSI may decrease the active site concentration and intensify the metal agglomeration due to the poor reducibility and too high reduction temperatures [68]. To address this issue, CaO was added to Al₂O₃ to form calcium aluminate, thus weakening the NiO–Al₂O₃ interaction in the spinel phase. Owing to the moderate MSI, abundant Ni active sites were produced for a highly active CH₄ conversion. With the continued increase of CaO loading, however, the interaction between the Ni and support became too weak to anchor the Ni nanoparticles, thus causing Ni agglomeration and surface area reduction; moreover, higher electron density at the Ni surface limited the CH₄ activation, thus deteriorating both the activity and stability [22]. In addition to CaO, rare earth metal oxide La₂O₃ can be utilized as a modifier to adjust the MSI in Ni/Al₂O₃. For example, the monolayer of La₂O₃ strengthened the Al₂O₃–La₂O₃ interaction and inhibited the Ni migration; however, the coverage on the Ni sites reduced the effective exposure of active centers to the reactants. Under CO₂ treatment, La₂O₂CO₃ was formed rather than La₂O₃ on the Al₂O₃ surface, which increased the amount of active Ni sites by preventing NiAl₂O₄ spinel phase formation. At 650 °C, 61% CH₄, and 65% CO₂ conversions were achieved with only 4% carbon formation (Table 1) [32]. La₂O₃ was also used to tune the MSI of Ni/ZrO₂, that is, a moderate MSI was provided, and the reducibility was enhanced, which produced more active sites for the CO₂ and CH₄ activation, thus resulting in an admirable conversion of CO₂ (>70%) and CH₄ (>60%) at 700 °C; however, CH₄ direct decomposition was promoted by La₂O₃ and carbon filaments and fibers at the Ni surface, which gradually caused the activities to deteriorate over the 67 h DRM reaction at 700 °C [69]. Moreover, La₂O₃, ZrO₂, and CeO₂ were compared in terms of their adjustment of MSI in the Ni/MgO catalyst. In detail, a higher reducibility was achieved with ZrO₂, and the amount of Ni active sites was much higher than that of Ni/CeO₂–MgO (8.12 × 10⁻⁷ vs. 3.93 × 10⁻⁷ g_mol/g_cat). As a consequence, more CH₄ molecules were activated with ZrO₂, and more H₂ gas was generated based on the higher H₂/CO ratio of 0.89, compared with that of Ni/CeO₂–MgO (0.78) (Table 1) [62].

3.3. Oxygen Defects

Metal oxides with redox property and oxygen defects accelerate the lattice and surface oxygen migration and enhance the surface oxygen concentration. In both mono-functional and bi-functional mechanisms, the dissociated CH_x intermediates will be effectively gasified by O radicals to produce CO and H adsorbed on the metal sites, thus inhibiting the carbon deposition. Moreover, the high oxygen concentration and fast oxygen mobility benefit the S removal and facilitate the sulfide conversion. For example, the CO₂ adsorption and dissociation were promoted on the ZrO₂ surface with a redox property, thus inhibiting the coke formation [70]. Moreover, H₂-treatment generated more oxygen vacancies in ZrO₂ compared with N₂ and O₂ calcination, which enhanced the CO₂ activation by forming monodentate and bidentate carbonates, thus releasing oxygen species to eliminate the carbon deposits [71]. Similarly, when TiO₂ was doped in Ni/Al₂O₃, the carbon deposition was alleviated owing to the redox ability of TiO₂; however, an excessive amount of TiO₂ might lead to Ni agglomeration, coverage of active sites, and titania structure shrinkage [72]. Apart from the transition metal oxides, CeO₂ presents an impressive oxygen storage capacity and a redox pair of Ce³⁺/Ce⁴⁺, which interacts with Ni via the d–d orbital electron transfer from CeO₂ to Ni. During the reaction, the adsorbed CO₂ will dissociate into CO and O radicals, which oxidizes the carbon species (e.g., CH_x) to produce CO and H atoms (combining to generate H₂); in the meantime, CO₂ reacts with Ce₂O₃ to generate CO and CeO₂. Subsequently, carbon species from CH₄ activation are oxidized by CeO₂ to produce Ce₂O₃ and CO, thus completing the redox cycle of Ce³⁺/Ce⁴⁺ and removing the carbon deposits [73]. Additionally, due to the oxygen defects and redox pair of Ce³⁺/Ce⁴⁺, the surface oxygen mobility is promoted, and the content of surface oxygen becomes higher, thus eliminating the carbon effectively [74]. Moreover, owing to the enhanced surface oxygen species in the presence of CeO₂, CO₂ is adsorbed on the surface in the form of bidentate carbonate, which easily combines with carbons so as to maintain the active site as being free from cokes, and to achieve a high conversion efficiency in the DRM reaction [12,75]. When CeO₂ was doped into Ni/Al₂O₃, the CeAlO₃/CeO₂ redox pair was formed, producing abundant oxygen defects and surface oxygen species. Due to the promoted oxygen migration to the interface between Ni and supports, carbon elimination was effectively facilitated, keeping the active sites exposed to the reactants. Moreover, surface carbonates were generated from the reaction between CO₂ and CeO₂, which improved the CH_x conversion and coke resistance. When the Ce loading was increased up to 15 wt%, only 0.29 g/g_cat carbon deposition was produced over a 250 h DRM reaction at 800 °C [76]. Similarly, when Ni silicate nanotubes (NSNTs) were doped with CeO₂, both Ce³⁺ and oxygen defects were generated from the interface reaction of CeO₂ and NSNTs. Benefiting from the oxygen migration from the adsorbed CO₂ and unidentate carbonate to the CH_x and C species at the nearby Ni sites, only 6% carbon deposition was produced after 100 h DRM reaction at 750 °C (Table 1) [28]. Despite the abundant oxygen vacancies and promoted oxygen mobility, a careful control of the CeO₂ concentration doped in the catalyst is required because the addition of CeO₂ may reduce the metal–support interaction and produce large metal crystal sites, which lowers the CH₄ activation and causes coke deposition.

To further enhance the oxygen defect concentration, ZrO₂ was integrated with CeO₂ to produce a homogeneous solid solution, CeZrO₂, based on the peak at 29.4° in Figure 4a. Given that there were more oxygen defects than CeO₂, as reflected by the higher oxygen storage capacity (165 vs. 125 µmol/g_cat oxygen uptake), the carbon deposition was alleviated according to the higher intensity ratio of the D-band at 1356 cm⁻¹ (amorphous carbon) to G-band at 1580 cm⁻¹ (ordered carbon) (Figure 4b). Moreover, the Ni oxidation was also inhibited since the oxygen species were consumed by the redox cycle and the oxygen vacancies of the mixed oxides [77]. Similar to CeZrO₂, another mixed oxide Ce_0.70La_0.20Ni_0.10O_2−δ was applied in the DRM reaction, owing to the oxygen defects formed when the lattice expansion and partial dissolution of La³⁺ occurred. Additionally, more oxygen vacancies were produced with the increase in reduction temperatures. Benefiting from the oxygen defect, lattice oxygen migration from CeO₂ was accelerated, oxidizing Ni–C to produce Ce³⁺ species, Ni metals and CO molecules. Moreover, carbon deposits could be converted to CO by reacting with La₂O₂CO₃. As a result, zero coke formations were observed over the 50 h DRM reaction at 750 °C (Table 1) [29].

In addition to CeO₂, other rare earth metal oxides are characterized with a good oxygen storage capacity and redox potential. For example, the addition of Y₂O₃ enhanced the surface oxygen concentration and accelerated the oxygen transfer to the carbon species, thus effectively eliminating the cokes and retaining the catalytic stability [78]. As well as the intrinsic oxygen defects, more oxygen vacancies could be generated due to the Ni²⁺/Sm³⁺ ion exchange. As a result, Sm₂O₃-modified Ni catalysts exhibited a low carbon formation and stabilized CH₄ activation [79]. Zhang et al. [80] compared a series of rare earth metal oxides with the promoter of Ni/ZrO₂, based on the impact on the oxygen defects. The Y-doped catalyst possessed the largest amount of surface oxygen species, followed by Ni/ZrO₂ doped with Sm, La, Ce, and no dopant. Although the activation of CH₄ and CO₂ were facilitated in the presence of oxygen species, the coke deposition and elimination did not follow the same trend. In particular, at a low reaction temperature (e.g., 600 °C), CH₄ activation at the surface oxygen sites dominated the reaction pathway so that the formed carbon species might not be immediately removed, thus producing carbon deposits. On the contrary, when the reaction temperature was high (e.g., 800 °C), CO₂ activation caught up with the CH₄ dissociation, effectively removing the carbon species.

Different from above cases, LaAlO₃ perovskite mixed oxides were formed when Ni/Al₂O₃ was doped with La₂O₃. Compared with pristine Ni/Al₂O₃, the NiO–Al₂O₃ interaction was weakened due to the oxygen defects of LaNiO₃, and more active sites were produced for a better C–H activation, delivering a 37.2% increase of the H₂ yield over the 20 h DRM reaction at 700 °C. Interestingly, both unmodified and modified catalysts exhibited the formation of carbon nanotubes (CNTs); however, unlike most of the reported works, the catalytic activity of LaNiO₃ was promoted by CNTs in this instance, since the Ni active sites were located at the tip of CNTs due to the weakened MSI and metal detachment from the support surface, thus being fully exposed to the reactant molecules. In comparison, the Ni metals were anchored on the Al₂O₃ support surface for the Ni/Al₂O₃ catalyst, which resulted in metal growth from the neck (34.9 vs. 16.1 nm after the 10 h reaction) and possible coverage by the carbon clusters (Figure 5a) [81].

As for the perovskite structures, B-site modification is proven effective to enhance the oxygen defects. In the Fe-doped LaNi_0.8Fe_0.2O₃ catalyst, Fe₃O₄ was formed during the Fe oxidation by CO₂. Subsequently, La₂O₃ reacted with Fe₃O₄ to form LaFeO_3−δ with lattice distortion. To compensate for the low valence state of Fe ions, oxygen defects were generated and promoted with the coke elimination [82]. To further modify the perovskite structure of LaNi_0.5Fe_0.5O₃, a series of Co was doped to substitute the B site ions. According to Figure 5b,c, O_β at 531.4 eV referred to the surface oxygen species (e.g., carbonate and hydroxyl groups), whereas O_α at 528.9 eV represented the lattice oxygen. A higher ratio of O_β/O_α indicated a higher percentage of surface oxygen, which could be ascribed to the in situ formed oxygen defects of LaFeO₃ perovskite mixed oxides and undercoordinated B-site cations [63,83]. Moreover, the partial substitution of Co ions (0.1 and 0.3) exhibited the largest amount of oxygen defects due to the multiple spin and oxidation states of Co species. Furthermore, the La₂O₃ derived from the reduction of LaNi_xCo_1−xFe_0.5O₃ reacted with adsorbed CO₂ to form La₂O₂CO₃, which also contributed to the oxygen species during the reaction [84]. Owing to the facilitated oxygen mobility and enhanced surface oxygen concentration, La(Co_0.1Ni_0.9)_0.5Fe_0.5O₃ and La(Co_0.3Ni_0.7)_0.5Fe_0.5O₃ exhibited a stable conversion of CH₄ (70%) and CO₂ (80%) over the 30 h DRM reaction at 750 °C (Table 1). More excitingly, the coke formation was as low as 0.8 and 1.5 mg_C/g_cat for La(Co_0.1Ni_0.9)_0.5Fe_0.5O₃ and La(Co_0.3Ni_0.7)_0.5Fe_0.5O₃, respectively [30]. Similar to the Co and Fe dopings, when Ni ions were partially replaced by Mn ions, the concentration of monoatomic oxygen defects (O⁻) was increased and higher oxygen mobility was achieved. Compared with undoped LaNiO₃, the peak of graphitic carbon at 1583 cm⁻¹ disappeared in the Raman spectra and the intensity of amorphous carbon at around 1300 cm⁻¹ was also considerably lowered [85]. In addition to the transition metals, the doping of Ru produced the perovskite structure Sr_0.92Y_0.08Ti_0.98Ru_0.02O_3+/−δ. Since the p-band of oxygen shifted to the Fermi level, the formed Ru–O bond was weaker than the Ti–O bond, reducing the formation energy of oxygen defects and increasing the surface oxygen concentration [86].

As well as the B-site substitution, partial replacement of La by Sr generated more oxygen defects and accelerated the surface oxygen mobility due to the lattice distortion, leading to an improved coke resistance; however, an excessive amount of Sr doping weakened the Ni–La₂O₃ MSI, thus intensifying the metal sintering and particle growth [87]. In another study where La₂Zr_1.44Ni_0.56O_7-d was modified with Ca and Sr in the A-site, more oxygen vacancies were generated in the Sr-substituted perovskite catalyst because of the lower ZrO₂ concentration at the surface, which exerted a stronger shielding effect in terms of blocking the oxygen defects. As a result of the abundant Ni active sites and more oxygen vacancies at the surface, trace amounts of soft filamentous carbons (0.024 g/g_cat) were produced over the 100 h DRM reaction at 800 °C [88]. Apart from the basic metal oxides, rare earth metal oxide CeO₂ also presented a promotional influence on the catalytic performance as an A-site modifier in the LaNi_0.5Fe_0.5O₃ perovskite. In particular, a reversible redox reaction occurred between (LaCe)(NiFe)O₃ and CeO₂, where oxygen defects were produced from the redox pair of Ce³⁺/Ce⁴⁺. Additionally, the reduction of B-site cations into undercoordinated states was activated by the Ce³⁺ ions, further increasing the oxygen defect concentration. Owing to the above merits, CH₄ conversion was enhanced due to the metallic-like Ni, whereas CO₂ activation was promoted at the oxygen vacancies. Moreover, the accelerated lattice oxygen migration favored the immediate gasification of carbon deposits, thus maintaining catalytic stability. As a consequence, the La_0.4Ce_0.6Ni_0.5Fe_0.5O₃ mixed oxide delivered a quite stable conversion of CH₄ (62%) and CO₂ (72%) over the 25 h DRM reaction at 750 °C (Table 1) [63].

Apart from the perovskite mixed oxides, the spinel phase possesses oxygen defects, especially when Ni is doped. For example, Al₂O₃ dissolution in the MgAl₂O₄ spinel phase was charge compensated by the cation and oxygen vacancies at the octahedral and tetrahedral sites. The partial replacement of Mg by Ni introduced more oxygen defects in the formed (Ni,Mg)Al₂O₄, which facilitated the CO₂ activation and coke removal during the DRM reaction. When the (Co_0.375Ni_0.375Mg_0.25)Al₂O₄ structure was developed using co-doping into the (Ni,Mg)Al₂O₄, 18% of oxygen ions were removed from the structure under reduction, and a highly oxygen-deficient spinel phase was produced, where the sulfur species at the corners, step edges, and facet edges of metal sites were easily oxidized by the oxygen that migrated from the lattice and surface. As a consequence, after 12 h of the DRM reaction at 850 °C, the activity only dropped by 4% under 20 ppm H₂S [89]; however, oxygen defects might be only a partial explanation of the origin of sulfur-tolerance. In another study, where 20–30 ppm sulfur was fed in the form of dimethyl sulfoxide (DMSO), only NiCo/CeZr exhibited a quite stable CH4 conversion rather than Ni/CeZr or NiCo/CeLa. Since CeLa also provided abundant oxygen vacancies, the origin of sulfur resistance might not be the oxygen defects or mobility. In this case, Ni–Co synergy could be a dominant factor. Moreover, the gradual decrease of H₂/CO ratio suggested the more rapid deactivation of DRM than the RWGS reaction under a sulfur environment [90].

3.4. Surface Acidity/Basicity

CH₄ prefers to be dissociated at the acidic sites at the surface, easily causing the carbon deposition. To address the issue, a stronger basicity and more basic sites usually enable a better adsorption and activation of CO₂, and the produced O radicals effectively oxidize the CH_x and C species. For instance, carbon deposition could be initiated by the acidic sites at the Al₂O₃ surface. To tune the surface acidity and basicity, La₂O₃ can be added to Al₂O₃ to form a La₂O₃-doped Al₂O₃ surface, thus balancing the CO₂ adsorption/activation, coke elimination, and CH₄ dissociation. As well as the O radicals from the activation of adsorbed CO₂ at the basic sites, the La₂O₂CO₃ generated from the reaction between La₂O₃ and CO₂ favored the coke removal, producing La₂O₃ and CO [32,91]. In another study when Ni/SiC-foam was doped with La₂O₃, the surface basicity was greatly enhanced based on the much higher peak intensity in the CO₂-TPD profile, owing to the formation of Ni–La₂O₃ nanocomposites (Figure 6a). During the 50 h DRM reaction at 850 °C, less carbon deposition was observed according to TEM (Figure 6b) and TG results (10.1% mass loss) [92]; however, Ni agglomeration might be intensified in the presence of La₂O₂CO₃, which adversely affected the catalytic stability in the DRM reaction. Apart from La₂O₃, Sc₂O₃ as a modifier was proven effective in enhancing the basicity of Ni–Co/SBA-15 catalyst. With the 5% and 10% Sc addition, the overall basicity was increased by 32% and 37%, respectively, especially for the medium basic sites. After the 40 h DRM reaction at 700 °C, the amount of inert carbon species was much less than the unmodified Ni–Co/SBA-15, suggesting the ready transformation of graphitic carbon to amorphous carbon with Sc doping was easily gasified to produce CO (Table 1) [19]. In another study where Ga₂O₃ was integrated with SiO₂, the Ga₂O₃-rich surface adsorbed CO₂ in the form of carbonate and bicarbonate species, which promoted the adsorption and activation of CO₂ compared with the pristine SiO₂, where CO₂ only physically or linearly interacted with the support. During the 10 h DRM reaction at 700 °C, the carbon formation was reduced by 53% [20].

As well as the enhanced basicity of Al₂O₃ with La₂O₃ addition, the doping of Sr into the A-site deficient La_0.8Cr_0.85Ni_0.15O₃ produced La_0.6Sr_0.2Cr_0.85Ni_0.15O₃, which enhanced the amount of both weak and strong basic sites. Owing to the improved CO₂ adsorption and activation, CO₂ conversions were increased, and more oxygen radicals were generated, which subsequently gasified the carbon species. In the meantime, CH₄ activation was promoted by the sufficiently small Ni nanoparticles. During the 24 h DRM reaction at 750 °C, both CO₂ and CH₄ conversions reached a high value of 89%; more crucially, zero coke formations were found with this condition [23]. In another study where SmCoO₃ perovskite mixed oxides were applied as the catalyst, both medium acid and basic sites were observed based on the CO₂- and NH₃-TPD characterizations, relating to the Sm₂O₃ and Co species, respectively. In particular, CH₄ preferred to be activated at the Co sites where CH_x intermediates and carbon deposits were formed. At the nearby basic Sm sites, Sm₂O₂CO₃ was produced from the combination of Sm₂O₃ and CO₂, which converted the carbon intermediates and deposits into CO with a simultaneous Sm₂O₃ regeneration. As a consequence, over 90% of the CH₄ and CO₂ conversions were approached during the 30 h DRM reaction at 800 °C (Table 1) [24].

The strength of basicity also determines the CO₂ adsorption and conversion as well as the carbon elimination. For example, only weak basic sites were present on the unmodified Ni/Al₂O₃ surface; when the MgO was added, the concentration of medium and strong basic sites was increased, which promoted the CO₂ adsorption at hydroxyl groups and activation into the oxygen radicals responsible for the coke removal. As a consequence, the CH₄ conversion was enhanced up to 1780 L_CH4 g_Ni⁻¹ h⁻¹, thus outperforming the pristine Ni/Al₂O₃ by 26%. Moreover, 70% carbon reduction was realized and most of the eliminated cokes were in the form of detrimental encapsulated carbons (Table 1) [25]. Although a higher basicity and CO₂ adsorption can be obtained with the addition of MgO, the NiO-MgO solid solution may cause the MSI to be too strong, thus retarding the reduction, limiting the active sites, and causing metal sintering due to the high reduction temperature, which lowers the surface area and deteriorates the lifespan of the catalysts. To balance the basicity and MSI, the ratio of MgO and Al₂O₃ was optimized to enable a basic surface where coking was retarded and the RWGS reaction was inhibited, leading to a stable conversion and high H₂ selectivity. In particular, with the Mg/Al ratio increasing from 0.1 to 0.24, the coke formation was reduced by 2/3 due to the enhanced basicity; however, an excessive amount of MgO (Mg/Al = 0.5) adversely affected the performance since the surface area and mesoporous structure were both diminished [93]. Owing to the strong basicity and well-retained mesopores, Ni-DS19 (Mg/Al = 0.24) exhibited a better stability (30 h without reactor plugging) and lower carbon deposition rate than its other two counterparts (Figure 7). In addition to MgO, Y₂O₃ was doped to adjust the basicity of Al₂O₃ by introducing more weak and medium basic sites. Benefiting from the reversible CO₂ adsorption and desorption, coke deposition was significantly reduced, and negligible activity degradation (0.8% for CO₂ and 1.1% for CH₄) was observed during the 10 h DRM reaction at 700 °C [64]. In another study, the total basic site concentration of Ni/ZrO₂ was increased from 73 to 100 µmol CO₂/g with the doping of Y₂O₃. More importantly, the percentage of weak and medium basic sites was enhanced from 79.9% to 87%, favoring the formation of active surface carbonate species and subsequent oxidation of cokes [60]. Similarly, weak and medium basic sites were introduced to Ni/Mg-Al double-layered hydroxides with the addition of Y₂O₃. Owing to the promoted CO₂ activation, about a 10% increase of CH₄ conversion was obtained in the 1.5 wt% Y₂O₃-promoted catalyst at 700 °C. Moreover, the conversion of CH₄ dropped by 4% over 10 h in the unmodified catalyst, whereas that of the Y₂O₃-doped catalyst was only 1% [64].

4. Conclusive Remarks and Prospect

In this review, four modification strategies to improve the catalytic performances of Ni-based catalysts in the DRM reaction are critically discussed based on four types of metal oxides (basic oxides, rare earth metal oxides, transition metal oxides, and mixed oxides). In the support confinement section, order porous support structures and hierarchical core-shell designs are proven effective in controlling Ni size and dispersion in high temperature conditions, thus maintaining the number of active sites for the activation of reactant molecules. As for the metal-support interaction, solid solution or spinel phase formation contributes to the highly distributed Ni crystals which strongly interact with the matrix via chemical bonding, thus inhibiting the coke formation; however, when the MSI is too strong, it may lower the reducibility and initiate the metal sintering. For the oxygen defects, redox property and lattice distortion are the main origins of oxygen vacancies, which facilitate the lattice and surface oxygen migration and enhance the surface oxygen concentration, resulting in an efficient elimination of carbon deposits and metal sulfides. As to the surface basicity, medium basicity might be favored in most cases since CO₂ only physically adsorbs onto the weak basic sites, whereas activation is reluctant at basic sites that are too strong. The reversible adsorption of CO₂ benefits the dissociation into CO and O radicals and the formation of active carbonates, which further converts the carbon intermediates or deposits into gaseous products. Despite the achievements in related fields, several issues and possible solutions are proposed as below.

First, there is a debate regarding the impacts of NiAl₂O₄. In particular, the Ni metal particles extracted from the spinel phase may interact strongly with the matrix which prevents the metal sintering; on the other hand, the low reducibility might be a concern for effective activation during the reaction; therefore, efforts should be put into the clarification of the effects of spinel phase on the concentration and agglomeration of active sites.

Second, the influences of in situ formed carbon nanotubes (CNTs) are still unclear. In certain scenarios where the Ni detachment occurs due to the weak MSI, the activity may not be adversely affected since the Ni sites located at the tip still activate the CH₄; however, in other cases where metals are strongly anchored at the support surface, CNTs might cover the Ni phase and intensify the metal growth from the neck, which would be encapsulated by carbon clusters. An in-depth study is recommended to monitor the reaction progress and elucidate the critical impacts of CNTs in different conditions.

Third, oxygen defects have been reported as being effective in improving sulfur resistance when metal oxides with multiple valence states and redox properties are doped in the catalysts; however, owing to the fact that some metal oxides with oxygen vacancies still suffer the deactivation caused by the S poisons, other determining factors need to be explored.

Fourth, CH₄ activation at the acidic sites and CO₂ dissociation at the basic sites need to be well balanced with the help of multi-functional metal oxides, which possess both acidic and basic groups. To address the issue, the co-existence of these two types of sites in mixed oxides or co-supports with Ni doping are promising solutions to simultaneously activate the CH₄ and remove the carbon intermediates.