Activation of CO₂ and NH₃ at interface between Ni and Mg–Al mixed oxide for CH₄ synthesis

Abstract

To reduce greenhouse gas emissions and achieve a sustainable society, synthesizing methane (CH₄) from carbon dioxide (CO₂) and green hydrogen (H₂) is a promising approach. However, the limited availability of renewable energy and the challenges of H₂ transport have created a demand for CH₄ synthesis using ammonia (NH₃) instead of H₂. To realize this process, the development of highly active catalysts for the one-step synthesis of CH₄ from CO₂ and NH₃ is essential, yet effective non-noble metal catalysts remain undeveloped. Here, we report the superior activity of supported Ni catalysts based on Mg–Al mixed oxide systems. Operando DRIFTS measurements revealed that CH₄ formed at the Ni–support interface through hydrogenation of isocyanate species (*NCO) derived from reaction of CO₂ and NH₃. Promoting this hydrogenation requires both high NH₃ decomposition activity and abundant Ni–support interface sites. The Mg–Al mixed oxide support, with its moderate basicity and favorable morphology for Ni dispersion, provided these features, resulting in high catalytic activity. The catalyst developed in this study demonstrates efficient CO₂ conversion into value-added products using NH₃ as a hydrogen carrier and offers a promising strategy toward a more sustainable energy system.

---

1. Introduction

In line with the goals of the Paris Agreement, countries worldwide have implemented policy measures to reduce carbon emissions. To achieve these goals, carbon capture and utilization has emerged as a key strategy for mitigating the impact of carbon emissions. This approach has the potential to reduce reliance on fossil fuels by converting CO₂ to value-added products through catalytic processes.¹ Among the various CO₂ utilization pathways, CO₂ methanation (CO₂ + H₂ methanation; eqn (1)) has gained attention because CH₄ serves as a valuable fuel for domestic, industrial, and power generation applications. Another advantage is that the synthesized methane can be directly injected into existing natural gas infrastructure without the need for infrastructure modification.^2,3

Securing large quantities of green H₂ is essential for carbon-neutral CH₄ synthesis. However, large-scale green H₂ production plants can be established in only a few locations due to limitations in renewable energy resources.⁴ Consequently, most countries and regions must import green H₂ from areas with abundant resources. However, H₂ is difficult to handle because it must be liquified at extremely low temperatures (−253 °C), leading to high transportation costs.⁵ NH₃ is considered the most promising H₂ carrier because it can be liquified under mild conditions (20 °C, 0.8 MPa)⁶ and has a high volumetric density (121 kg_H2 m⁻³),⁷ thereby reducing transportation costs. Thus, an increasing number of nations and regions are turning to NH₃ as a means to overcome the challenges of hydrogen storage and transportation.⁸

When synthesizing CH₄ using NH₃ as a starting material, two distinct steps are typically involved: NH₃ decomposition (eqn (2)) and CO₂ + H₂ methanation. However, NH₃ decomposition is an endothermic reaction, whereas CO₂ + H₂ methanation is an exothermic reaction, meaning that the two reactions have different optimal thermal environments. As a result, this continuous process requires two reactors with separate thermal control systems, leading to significant energy loss and increased equipment costs. From this perspective, the one-step synthesis of CH₄ from CO₂ and NH₃ (CO₂ + NH₃ methanation; eqn (3)) is considered a more efficient approach when NH₃ is used as the feedstock. Therefore, developing highly active catalysts for CO₂ + NH₃ methanation could help reduce carbon emissions in regions lacking renewable energy resources.

CO₂ + 4H₂ → CH₄ + 2H₂OΔH_298K = −165 kJ mol_CH4⁻¹ (1)

NH₃ → 1/2N₂ + 3/2H₂ΔH_298K = +46 kJ mol_NH3⁻¹ (2)

CO₂ + 8/3NH₃ → CH₄ + 4/3N₂ + 2H₂OΔH_298K = −43 kJ mol_CH4⁻¹ (3)

Several studies have reported that oxide-supported ruthenium (Ru) and nickel (Ni) catalysts exhibit activity for CO₂ + NH₃ methanation.^9,10 Based on these findings, our group is focused on using supported Ni catalysts to design cost-effective CO₂ + NH₃ methanation processes. In our previous work, we investigated the effects of catalyst supports and the reaction mechanism to establish design guidelines for highly active supported Ni catalysts.¹¹Operando DRIFTS measurements revealed that isocyanate species (*NCO) form during the reaction and act as intermediates. Furthermore, we found that the basicity of the support influences the reactivity of the *NCO. Stronger basicity results in lower *NCO reactivity because this species is acidic and therefore stabilized on strongly basic sites. Another key factor we found is NH₃ decomposition activity because a high partial pressure of H₂ is needed for the hydrogenation of *NCO. It is known that strong basicity promotes NH₃ decomposition by facilitating the associative desorption of adsorbed N atoms, which is the rate-determining step of the reaction.^12,13 High electron-donating ability also contributes by supplying electrons to the antibonding orbital of the metal–N bond of adsorbed N atoms. Thus, alkaline earth metals and alkali metals, which have strong basicity, are often used as promoters or as components of mixed oxide supports.^14,15 However, such strongly basic materials stabilize the *NCO intermediates, thereby lowering CO₂ + NH₃ methanation activity. Thus, to enhance the CO₂ + NH₃ methanation activity, it is necessary to develop catalysts with moderate basicity and high NH₃ decomposition activity.

Mg–Al mixed oxides are widely used as supports in various catalytic reactions due to their high specific surface area, versatile preparation methods, and tunable basicity, which depends on the Mg/Al ratio.^16–19 Coleman et al. used Ni/Mg–Al mixed oxide catalysts for steam reforming of ethanol and found that the catalysts exhibited high H₂ and CO selectivity.²⁰ In this reaction, strong acidic or basic properties can easily cause side reactions, such as the production of ethylene or acetylene. The coexistence of acidic (Al₂O₃) and basic (MgO) materials in the Mg–Al mixed oxide catalysts provides moderate basicity, which suppresses side reactions and enhances reaction selectivity. Yang et al. used a Ni/Mg–Al mixed oxide catalyst derived from Ni–Mg–Al hydrotalcite (a layered double hydroxide represented by the chemical formula: [M_1−x²⁺M_x³⁺(OH)₂](Aⁿ⁻)_x/n mH₂O, where M represents metals and A is an anion) for CO₂ + H₂ methanation.²¹ This catalyst showed excellent activity by activating CO₂ at Ni–support interface sites with moderate basicity.

The Ni/Mg–Al mixed oxide catalyst is also effective for NH₃ decomposition. We have found that Ni/Mg–Al catalysts obtained from Ni–Mg–Al hydrotalcite precursors exhibited excellent NH₃ decomposition activity.²² The basicity provided by the MgO and the high Ni dispersion resulting from the formation of a Mg(Ni, Al)O solid solution contributed to the high activity. Su et al. showed that by optimizing the Ni, Mg, and Al ratio in the hydrotalcite structure, a Ni/Mg–Al catalyst with NH₃ decomposition activity comparable to that of a supported Ru catalyst could be acheived.¹⁹ Qiu et al. reported that MgAl₂O₄ supports synthesized via the hydrothermal method possess a large specific surface area, which enhances Ni dispersion and leads to higher NH₃ decomposition activity than supports prepared by sintering MgO and Al₂O₃ mixtures or using the citric acid method.¹⁷ These findings suggest that high Ni dispersion enables Mg–Al mixed oxide-supported catalysts to exhibit superior NH₃ decomposition activity without the need for strongly basic additives. In other words, this support has strong potential as an effective catalytic material for CO₂ + NH₃ methanation.

In this study, we prepared Mg–Al mixed oxide supports with varying Mg/Al ratios for supported Ni catalysts and evaluated their CO₂ + NH₃ methanation activity. The Ni/Mg–Al mixed oxide catalysts showed higher CO₂ + NH₃ methanation activity than Ni/Al₂O₃ and Ni/MgO. Operando DRIFTS measurements and various characterizations revealed that this high activity was due to a superior NH₃ decomposition activity and an abundance of Ni–support interfaces where hydrogenation of the intermediate occurred. In particular, the catalyst with a support of Mg/Al = 0.5/0.5 showed significantly higher activity than previously reported supported Ni catalysts due to its moderate basicity and morphology conducive to high Ni dispersion.

2. Experimental

2.1. Support preparation

Mg–Al mixed oxide supports with different Mg²⁺ and Al³⁺ contents were synthesized by preparing Mg–Al mixed hydroxides as precursors and subsequently calcining them. The Mg–Al mixed hydroxides were prepared using a precipitation method, as detailed in the SI (S1.1.) The resulting Mg–Al mixed hydroxides were labeled Mg_xAl_1−x(OH)_n, where x represents the molar ratio of Mg to the total moles of Mg and Al (Mg/(Mg + Al)) in the samples. In this study, six values of x were used: 1.00, 0.90, 0.75, 0.50, 0.25, and 0 (Table S1). To synthesize Mg–Al mixed oxide supports, each hydroxide was calcined in air at 700 °C for 5 h. The resulting Mg–Al mixed oxides were labeled Mg_xAl_1−xO_n.

2.2. Catalyst preparation

Supported Ni catalysts were prepared by an impregnation method. Ni(C₅H₇O₂)₂ (Sigma-Aldrich) was used as the nickel precursor. A nickel precursor solution was prepared by dissolving Ni(C₅H₇O₂)₂ in tetrahydrofuran. The synthesized Mg–Al mixed oxides were then added to the nickel solution, and the mixture was stirred overnight at room temperature. After the stirring, the solvent was removed by evaporation under vacuum (180 hPa) at a constant temperature of 35 °C. The resulting slurry was dried overnight at 80 °C, then crushed and calcined at 500 °C for 5 h under flowing Ar (80 mL min⁻¹). The Ni loading was fixed at 20 wt% for all catalysts. The resulting supported Ni catalysts were labeled Ni/Mg_xAl_1−xO_n. The compositions of the prepared catalysts were analyzed using inductively coupled plasma mass spectrometry (ICP-MS; OptiMass9600, GBC).

2.3. Catalytic activity test

The obtained catalyst powders were pelletized and sieved to 250–500 μm. The CO₂ + NH₃ methanation activity of each catalyst was evaluated using a conventional flow system with a tubular reactor under atmospheric pressure. After loading 150 mg of catalyst, in situ pre-reduction was performed with pure H₂ (50 mL min⁻¹) for 1 h at 600 °C. Following reduction, the catalyst was cooled to 300 °C under flowing Ar (50 mL min⁻¹), after which a mixture of CO₂ (5 mL min⁻¹) and NH₃ (13 mL min⁻¹) was passed over the catalyst (space velocity [SV] = 7200 mL h⁻¹ g⁻¹). When CO₂ and NH₃ (and H₂O) are mixed below 150 °C, ammonium salts such as ammonium carbonate ((NH₄)₂CO₃), ammonium bicarbonate ((NH₄)HCO₃), and ammonium carbamate (NH₂COONH₄) may form and cause blockages in the gas lines.²³ To prevent this, a ribbon heater was used to maintain the gas line temperature at 150 °C. To remove unreacted NH₃, the effluent gas mixture was passed through a gas-washing bottle filled with pure water connected to the reactor outlet. A cold trap was attached to the outlet of the gas-washing bottle to remove water vapor. The catalyst temperature was held constant for 30 min, and the composition of the gas mixture exiting the cold trap was analyzed using an online gas chromatograph (Agilent 490 MicroGC, Agilent) equipped with a thermal conductivity detector and Molsieve 5A and PoraPLOT Q columns. Measurements were taken 50 °C intervals up to 700 °C. During the activity test, Ar (60 mL min⁻¹) was mixed into the gas stream after the reactor, and the total outlet gas flow rate was determined using Ar as the internal standard. CH₄ yield, CO yield, and NH₃ conversion were calculated as follows:

CH₄ yield (%) = CH_{4 outlet}/CO_{2 inlet} × 100

CO yield (%) = CO_outlet/CO_{2 inlet} × 100

NH₃ conversion (%) = NH_{3 consumed}/NH_{3 inlet} × 100

where CO_{2 inlet} and NH_{3 inlet} are the inlet flow rates of CO₂ (5 mL min⁻¹) and NH₃ (13 mL min⁻¹), respectively; CH_{4 outlet} and CO_outlet are the outlet flow rates of CH₄ and CO, respectively; and NH_{3 consumed} was determined from the stoichiometric ratio with the outlet N₂ flow rate as NH_{3 consumed} = N_{2 outlet} × 2.

For the stability test, CH_{4 outlet}, CO_outlet, and N_{2 outlet} were recorded over time on stream at an SV of 7200 mL h⁻¹ g⁻¹ at 500 °C.

Experimental results were compared with thermodynamic equilibrium values calculated using the HSC Chemistry 6.1 commercial steady-state simulation package (Outotec Research Oy, Pori).

Catalyst activity for NH₃ decomposition was measured using the same setup, catalyst amount, and pre-reduction conditions as in the CO₂ + NH₃ methanation experiment. In the NH₃ decomposition experiment, the total gas flow rate was adjusted by varying the He flow such that the NH₃ partial pressure matched that of the CO₂ + NH₃ methanation test (NH₃: 13 mL min⁻¹, He: 5 mL min⁻¹; SV = 7200 mL h⁻¹ g⁻¹).

2.4. Operando diffuse reflectance infrared Fourier transform spectroscopy (operando DRIFTS)

Operando DRIFTS was performed using an infrared spectrometer (FT/IR-4700 spectrometer; Jasco) equipped with an MCT detector. Approximately 15 mg of finely ground catalyst or Mg_xAl_1−xO_n support was placed in a ceramic cup, and the cup was loaded into the DRIFTS cell. Before measurement, the sample was reduced in situ under a 100% H₂ stream (20 mL min⁻¹) for 1 h at 600 °C. After the reduction, the sample chamber was cooled to 100 °C under He flow (20 mL min⁻¹). A background spectrum was collected in He at 100 °C with a resolution of 4 cm⁻¹ and an average of 64 scans, and this spectrum was later subtracted from that of the test sample. DRIFT spectra were then recorded under a CO₂ + NH₃ methanation atmosphere (CO₂/NH₃/He = 1/2.6/6.4 mL min⁻¹) while increasing the temperature of the sample chamber from 100 to 600 °C at a ramping rate of 10 °C min⁻¹. During the measurements, the effluent gas composition at the DRIFTS cell outlet was analyzed using a quadrupole mass spectrometer (BELMASS, MicrotracBEL, Japan).

2.5. Characterization of supports and catalysts

X-ray diffraction analysis was performed using a SmartLab X-ray diffractometer (Rigaku, Japan) equipped with a Cu Kα radiation source. Diffraction patterns were analyzed with the PDXL2 software (ver. 2.8.4.0; Rigaku) using reference data from the International Centre for Diffraction Data and the Crystallography Open database.²⁴

The specific surface areas of the supports and catalysts were determined by N₂ adsorption at −196 °C using a BELSORP-mini X instrument (MicrotracBEL, Japan) and the Brunauer–Emmett–Teller model. Before measurement, all supports and catalysts were degassed under vacuum for 3 h at 300 °C. The pore size distribution and pore volume of the supports were determined using the Barrett–Joyner–Halenda method from the desorption branch of the N₂ adsorption isotherm.

The microstructures of the prepared supports were observed using a field-emission scanning electron microscope (JEOL JSM-7500FA) operated at 2.00 kV. To prevent sample charging during observation, a 20 nm-thick layer of OsO₄ coating was applied to the sample using an Osmium Plasma Coater (OPC60A, Filgen).

H₂ temperature-programmed reduction measurements were performed using a BELCAT-II apparatus (MicrotracBEL) to investigate the reducibility of Ni species in each catalyst. For each run, 50 mg of calcined catalyst (before reduction) was loaded into the reactor and pretreated at 150 °C for 3 h. The sample was then reduced under 100% H₂ flow (50 mL min⁻¹) while heating from room temperature to 1000 °C at a rate of 10 °C min⁻¹. The H₂O and CH₄ profiles were monitored by a quadruple mass spectrometer at m/z = 18 and m/z = 16, respectively.

High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images of prepared catalysts were obtained using JEOL JEM-ARM 200F (cold) and JEOL JEM-2100F HK microscopes operated at 200 kV. Prior to observation, all catalysts were reduced under flowing H₂ (50 mL min⁻¹) for 1 h at 600 °C. Reduced samples were dispersed in ethanol at room temperature, and a drop of the suspension was placed onto a carbon-coated copper grid and vacuum-dried for 24 h at room temperature.

H₂-pulse chemisorption measurements were performed using a BELCAT-II apparatus to evaluate the Ni dispersion of each catalyst. In each run, 50 mg of calcined catalyst (before reduction) was loaded into the reactor and pre-reduced under 100% H₂ flow (50 mL min⁻¹) for 1 h at 600 °C. Then, a mixture of 4.97% H₂ in He was injected in 15 pulses. Ni dispersion (%) was calculated from the amount of adsorbed H₂ per gram of catalyst (V_m: cm³ g⁻¹) using the following equation:

Ni dispersion (%) = (V_m × (SF/22414) × MW)/c

where SF (=2) is the stoichiometric factor for H₂ chemisorption, MW (=58.69) is the atomic weight of Ni atom, and c is the mass of Ni in the catalyst, calculated using the actual Ni loading measured by ICP-MS (OptiMass9600, GBC).

3. Results and discussion

3.1. CO₂ + NH₃ methanation activity of Mg–Al mixed oxide-supported catalysts

The actual contents of Ni, Mg, and Al in each supported Ni catalyst were estimated by ICP-MS (Table S2). In Table S2, the actual Ni content is expressed as loading (wt%), calculated using the following equation:

Ni loading (wt%) = (Ni/(Mg_aAl_bO_((2a+3b)/2) + Ni)) × 100

where a and b were obtained from ICP-MS measurements. For all catalysts, the actual Mg/Al molar ratios and Ni loadings were nearly equal to the nominal values, indicating that all catalysts were successfully synthesized with the intended compositions.

Before evaluating the CO₂ + NH₃ methanation activity, the basicity and intrinsic NH₃ decomposition activity (i.e., NH₃ decomposition activity in the absence of CO₂) of the Mg–Al mixed oxide catalysts were investigated. To examine the basic properties of each catalyst, CO₂ temperature-programmed desorption measurements were conducted (Fig. 1). The CO₂ desorption peak temperatures of the Mg–Al mixed oxide catalysts lay between those of Ni/Mg_1.00Al₀O_n (MgO) and Ni/Mg₀Al_1.00O_n (Al₂O₃), indicating that the Mg–Al mixed oxide-supported catalysts exhibited moderate basicity due to the coexistence of Mg²⁺ and Al³⁺. Fig. S1 shows the intrinsic NH₃ decomposition activity of each supported catalyst. The Mg–Al mixed oxide-supported catalysts exhibited higher NH₃ decomposition activity than Ni/MgO and Ni/Al₂O₃. Among them, Ni/Mg_0.90Al_0.10O_n, Ni/Mg_0.75Al_0.25O_n, and Ni/Mg_0.50Al_0.50O_n, with higher Mg²⁺ content, showed greater activity, each exhibiting similar NH₃ conversion trends with reaction temperature. For NH₃ decomposition over oxide-supported Ni catalysts, the associative desorption of adsorbed N is considered the rate-determining step.^25,26 It has also been reported that electron donation associated with basic properties promotes the rate-determining step.^12,13 The CO₂ temperature-programmed desorption profiles (Fig. 1) show that the CO₂ desorption peak shifted to higher temperatures with increasing Mg²⁺ content, suggesting stronger basicity. Therefore, catalysts with higher Mg²⁺ content likely exhibited superior intrinsic NH₃ decomposition activity because the rate-determining step was promoted by the enhanced electron-donating ability stemming from their stronger basicity. Interestingly, Ni/Mg_0.90Al_0.10O_n, Ni/Mg_0.75Al_0.25O_n, and Ni/Mg_0.50Al_0.50O_n showed higher intrinsic NH₃ decomposition activity than Ni/MgO, despite having weaker overall basicity. This is presumably due to differences in the number of Ni sites (i.e., Ni dispersion), which is discussed in detail in section 3.3.

Fig. 1 CO₂ temperature-programed desorption profiles of each Mg–Al mixed oxide-supported Ni catalyst.

The activity of each Mg–Al mixed oxide-supported Ni catalyst for CO₂ + NH₃ methanation was evaluated. In this experiment, estimating CO₂ conversion based on the composition of the effluent gas was difficult. The effluent gas from the reactor was passed through a gas-washing bottle filled with pure water to remove unreacted NH₃; however, unreacted CO₂ was also captured by the water because dissolved NH₃ renders the solution alkaline, which enhances CO₂ absorption. For example, the CO₂ conversion calculated from the CO₂ concentration at the outlet of the gas-washing bottle for the Ni/Mg_0.50Al_0.50O_n catalyst was approximately 40% at 300 °C (Fig. S2), even though the reaction hardly proceeded because no CH₄ or CO production was detected (Fig. 2a and c). Therefore, in the following discussion, catalytic activity is evaluated in terms of CH₄ and CO yields.

Fig. 2 Evaluation of the performance of Ni/Mg_xAl_1−xO_n catalysts for CO₂ + NH₃ methanation. (a) Temperature dependence of CH₄ yields for Mg–Al mixed oxide-supported catalysts. (b) CH₄ yield versus Mg/Al molar ratio of the support. (c) Temperature dependence of CO yields. (d) Temperature dependence of CO₂ conversion. (e) Temperature dependence of NH₃ conversion. (f) Correlation between CH₄ yield and NH₃ conversion at 500 °C.

The relationship between CH₄ yield and reaction temperature is shown in Fig. 2a. Equilibrium values (denoted as equilibrium (1)) were calculated assuming five products: CH₄, CO, N₂, H₂, and H₂O. At temperatures below 350 °C, CH₄ formation was negligible for all catalysts. A marked increase of CH₄ yields was observed for most samples from 400 °C. Compared with Ni/MgO and Ni/Al₂O₃, the Mg–Al mixed oxide-supported catalysts showed higher CH₄ yields at all temperatures, with maximum values achieved at 500 °C. At 500 °C, the CH₄ yield displayed a convex upward trend with respect to the Mg/(Mg + Al) molar ratio, peaking at 0.50 (Fig. 2b). For Ni/Mg_0.50Al_0.50O_n, the CH₄ yield was approximately 50% at 500 °C.

The CO yield as a function of reaction temperature is shown in Fig. 2c. In Fig. 2c, equilibrium CO yields were calculated under two conditions (denoted equilibrium (1) and equilibrium (2)). Equilibrium (1) was calculated assuming five products, as in Fig. 2a: CH₄, CO, N₂, H₂, and H₂O. Equilibrium (2) was calculated assuming four products: CO, N₂, H₂, and H₂O. CO formation was observed over all catalysts from 350 °C, with CO yields increasing similarly over all catalysts up to 400 °C. At temperatures below 400 °C, CH₄ yields were very small for all the catalysts, suggesting that the CO yield at those temperatures follows equilibrium (2) rather than equilibrium (1). At 450 °C, differences in CO yield among the samples were observed. For the Mg–Al mixed oxide-supported catalysts, CO yields decreased slightly or remained constant as the temperature was increased from 400 to 450 °C. In contrast, CO yields continued to increase over Ni/MgO and Ni/Al₂O₃, which showed low CH₄ yields relative to temperature. These results suggest that the CO formation rate is related to the CH₄ formation rate. Above 500–550 °C, the CO yield of all catalysts followed equilibrium (1), showing a shift from equilibrium (2) to equilibrium (1) with increasing CH₄ yield.

Fig. S3a and b shows the effect of SV on CH₄ and CO yields over the Ni/Mg_0.50Al_0.50O_n catalyst. Both CH₄ and CO yields as a function of reaction temperature varied with SV. As SV decreased, CH₄ yield increased, while the CO yield decreased and approached equilibrium (1). These results suggest that CO acts as an intermediate for CH₄ formation, or that a reaction pathway favoring CO formation dominates when CH₄ formation is slow. At an SV of 3600 mL h⁻¹ g⁻¹, a CH₄ yield of approximately 57% was obtained at 500 °C (Fig. S3a). This is the highest value reported for any supported Ni catalyst to date (Table S3), demonstrating the superior CO₂ + NH₃ methanation activity of this catalyst.

Our previous study revealed that in CO₂ + NH₃ methanation, CO₂ is converted only to CH₄ or CO and not to CN compounds (e.g., HCN).¹¹ Therefore, the CO₂ conversion of each catalyst was calculated using the following equation:

CO₂ conversion (%) = (CH_{4 outlet} + CO_outlet)/CO_{2 inlet} × 100

Fig. 2d shows the CO₂ conversion as a function of reaction temperature. Equilibrium (1) was calculated assuming five products, as in Fig. 2a: CH₄, CO, N₂, H₂, and H₂O. The Ni/Mg_0.50Al_0.50O_n catalyst showed the highest CO₂ conversion relative to reaction temperature. This trend was consistent with that of CH₄ yield (Fig. 2a), indicating that the formation rate of CH₄ affects the conversion rate of CO₂.

NH₃ conversion during CO₂ + NH₃ methanation was examined for each catalyst (Fig. 2e). Equilibrium NH₃ conversion values (equilibrium (1)) were calculated under the same conditions as in Fig. 2a. The Mg–Al mixed oxide-supported catalysts exhibited higher NH₃ conversion than Ni/MgO and Ni/Al₂O₃. At 450–500 °C, NH₃ conversion varied slightly among the samples. At 500 °C, the relationship between CH₄ yield and NH₃ conversion was investigated (Fig. 2f), revealing a linear correlation. This indicates that NH₃ decomposition activity is a key factor for CO₂ + NH₃ methanation activity, with Ni/Mg_0.50Al_0.50O_n exhibiting superior NH₃ decomposition activity in an CO₂ + NH₃ methanation atmosphere. However, its intrinsic NH₃ decomposition activity in the absence of CO₂ was comparable to that of other Mg–Al mixed oxide catalysts (Ni/Mg_0.90Al_0.10O_n, Ni/Mg_0.75Al_0.25O_n). These results demonstrate that CO₂ + NH₃ methanation activity is not solely determined by intrinsic NH₃ decomposition, and that the decomposition mechanism is changed in the presence of CO₂.

3.2. Reaction mechanism of CO₂ + NH₃ methanation over Mg–Al mixed oxide-supported catalysts

In section 3.1, it was shown that NH₃ conversion is a key factor in CO₂ + NH₃ methanation. Therefore, investigating the role of NH₃ in the reaction is essential to clarify the origin of the superior CO₂ + NH₃ methanation activity of Mg–Al mixed oxide-supported catalysts. To this end, operando DRIFTS measurements were performed to examine the reaction mechanism. Fig. 3a shows the DRIFTS spectra obtained under CO₂ and NH₃ flow (CO₂: 1.0 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹) for the Ni/Mg_0.5Al_0.5O_n catalyst. During the measurements, the temperature of the sample chamber was increased from 100 to 600 °C at a ramping rate of 10 °C min⁻¹. The effluent gas composition at the outlet of the DRIFTS cell was analyzed by online MS (Fig. 3b).

Fig. 3 (a) DRIFTS spectra of the Ni/Mg_0.50Al_0.50O_n catalyst under CO₂ + NH₃ methanation conditions (CO₂: 1.0 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹). (b) Mass spectroscopy analysis of the effluent gas at the outlet of the DRIFTS cell.

In Fig. 3a, IR bands assigned to gas-phase CO₂ and NH₃ were observed at 2300–2400 cm⁻¹ and 1625, 3334 cm⁻¹, respectively.^27–29 At 100 °C, bands at 1440 and 1594 cm⁻¹ corresponded to the symmetric OCO stretching of bicarbonate and asymmetric OCO-stretching of formate (HCOO*), respectively. These CO₂ adsorption species readily form on oxide supports.^28,29 When the reaction temperature was increased to 200 °C, a new band appeared at 2186 cm⁻¹, assignable to isocyanate species (*NCO).^30,31 The formation of *NCO indicates that C=O bond cleavage in CO₂ and C–N bond formation occurred on the catalyst. Using CO₂ and NH₃ as raw materials, the Bazarov reaction synthesizes urea (NH₂CONH₂) without a catalyst under harsh conditions (250 bar, 270 °C) and involves both C=O bond cleavage and C–N bond formation.³² The reaction proceeds via the formation of ammonium carbamate (NH₂COONH₄) and its dehydration (eqn (4) and (5)). Barzagli et al. have reported that urea can be produced from ammonium carbamate under milder conditions (>14 bar, 150 °C) in a solid-state reaction using CuO as a catalyst.³² Urea also thermally decomposes over oxides such as Al₂O₃ to yield isocyanic acid (HNCO) (eqn (6)).³³ These findings support that *NCO species can form via ammonium carbamate, urea, or related intermediates (e.g., *H₂NCOOH, *NHCO)³⁴ under mild conditions (1 bar, 200 °C) on Mg–Al mixed oxide-supported catalysts (eqn (7)).

CO₂ + 2NH₃ → NH₂COONH₄ (4)

NH₂COONH₄ → NH₂CONH₂ + H₂O (5)

NH₂CONH₂ → HNCO + NH₃ (6)

CO₂ + NH₃ → HNCO + H₂O (7)

At 300 °C, a shoulder band appeared at 2244 cm⁻¹ adjacent to the *NCO band (2186 cm⁻¹), suggesting the formation of a new site for *NCO adsorption or a change in the configuration of *NCO. At 400 °C, the intensities of the *NCO and CO₂ adsorption bands slightly decreased. Upon further heating to 500 °C, both bands markedly diminished and a band assigned to gas-phase CH₄ appeared at 3015 cm⁻¹.

As shown in Fig. 3b, mass signals at m/z = 2 (H₂) and m/z = 28 (N₂, CO) began to increase from 400 °C and continued to rise with temperature. In this reaction, the NH₃/CO₂ ratio ([8/3]/1) in the feed gas was higher than the stoichiometric ratio for HNCO formation from CO₂ and NH₃ (1/1) (eqn (7)). Thus, excess NH₃ not involved in the formation of HNCO was likely adsorbed on Ni and decomposed from 400 °C. The mass signal at m/z = 15 (CH₄) increased from 500 °C, accompanied by further increases in m/z = 2 (H₂) and m/z = 28 (N₂, CO). These results suggest that a high partial pressure of H₂ (i.e., active NH₃ decomposition) is needed for CH₄ formation. In Fig. 3a, the intensities of the *NCO, , and HCOO* bands decreased at 500 °C, where the partial pressure of H₂ is high, indicating that either *NCO or CO₂ adsorption species may be hydrogenated to CH₄ at this temperature.

Next, to investigate the properties of *NCO in detail, we analyzed the sites where this species is formed. DRIFTS spectra were obtained under a CO₂ and NH₃ flow (CO₂: 1.0 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹) using the Mg_0.5Al_0.5O_n support (i.e., without Ni loading), as shown in Fig. 4. Similar to the Ni/Mg_0.5Al_0.5O_n catalyst, IR bands corresponding to *NCO, , and HCOO* were observed at temperatures above 200 °C. The wavenumber of the *NCO band of the Mg_0.5Al_0.5O_n support was comparable to that of the Ni/Mg_0.5Al_0.5O_n catalyst (Fig. 3a), suggesting that *NCO forms on the support surface. Interestingly, for the Mg_0.5Al_0.5O_n support (Fig. 4), the intensities of the and HCOO* bands decreased at higher temperatures (600 °C), while that of the *NCO band remained unchanged, indicating that *NCO is strongly adsorbed on the support. Fig. S4 shows a comparison of the H₂ (m/z = 2) mass signals for the Ni/Mg_0.5Al_0.5O_n catalyst and the Mg_0.5Al_0.5O_n support during the DRIFTS measurements. The much weaker m/z = 2 signal intensity for the Mg_0.5Al_0.5O_n support indicated that NH₃ decomposition hardly occurs without Ni and that the partial pressure of H₂ influences *NCO decomposition. Therefore, it is suggested that *NCO undergoes hydrogenation over the Ni/Mg_0.5Al_0.5O_n catalyst. In other words, *NCO may act as an intermediate in CH₄ formation (eqn (8)) during CO₂ + NH₃ methanation over Mg–Al mixed oxide-supported catalysts.

Fig. 4 DRIFTS spectra for the Mg_0.50Al_0.50O_n support under CO₂ + NH₃ methanation conditions (CO₂: 1.0 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹).

To confirm the proposed mechanism in which *NCO is hydrogenated to CH₄, the changes of the DRIFTS bands attributed to *NCO (an intermediate of CH₄) and gas-phase CH₄ were investigated under H₂ flow. To ensure the formation of *NCO on Ni/Mg_0.5Al_0.5O_n, CO₂ and NH₃ gases (CO₂: 1.0 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹) were passed over the catalyst at 300 °C for 20 min. The reactant gases were then stopped, and the sample chamber was purged with He (20 mL min⁻¹) while cooling to 100 °C. Once cooled, the temperature was increased from 100 to 500 °C under an H₂/He flow (H₂: 3.9 mL min⁻¹, He: 8.7 mL min⁻¹), and the transient response of the DRIFTS spectrum was recorded (Fig. 5a). The effluent gas composition was analyzed by online mass spectrometry, monitoring m/z = 15 for CH₄ formation (Fig. 5b). In Fig. 5a, IR bands of *NCO (2259, 2201 cm⁻¹), (1425 cm⁻¹), (1428 cm⁻¹), and HCOO* (1608 cm⁻¹) were observed at 100 °C, along with a CO adsorption band of Ni (CO*) at 1852 cm⁻¹. The CO* band was not observed under CO₂ + NH₃ methanation conditions (Fig. 3a), likely because the Ni surface was covered by NH₃, leaving no available CO adsorption sites. However, in this experiment (Fig. 5a), the He purge likely removed NH₃, allowing CO derived from CO₂ adsorption species to adsorb on Ni as CO*. When the temperature was increased over 200 °C, the *NCO band intensity markedly decreased and almost disappeared at 250 °C. The , HCOO*, and CO* bands also weakened with increasing temperature, particularly at temperatures above 250 °C, indicating that *NCO was more easily hydrogenated at high temperatures. In Fig. 5b, the m/z = 15 signal was increased at temperatures between 150 °C and 200 °C, where a significant decrease in the *NCO peak occurred, reaching a maximum at 250 °C, where the *NCO bands almost disappeared. These results confirm that *NCO acts as an intermediate in CH₄ formation over Ni/Mg_0.5Al_0.5O_n catalyst. As shown in Fig. 4, *NCO forms on the Mg_0.5Al_0.5O_n support. Therefore, it is proposed that H₂ generated from NH₃ decomposition spills over from the Ni surface to the Ni–support interface, where *NCO is hydrogenated to CH₄ (Scheme 1). At low temperatures, where NH₃ decomposition (and thus H₂ partial pressure) is limited, *NCO likely decomposes thermally to CO (eqn (9)) or CO was formed from the CO₂ adsorption species through the reverse water–gas shift reaction (eqn (10)). At higher temperatures, where NH₃ decomposition is faster (and H₂ partial pressure is high), *NCO hydrogenation to CH₄ (eqn (8)) predominates, and CO formation reactions (eqn (9) and (10)) are suppressed.

*NCO + 3H₂ → CH₄ + 0.5N₂ + H₂O (8)

*NCO → 0.5N₂ + CO (9)

CO₂ + H₂ → CO + H₂O (10)

From the DRIFTS results over the Mg–Al mixed oxide-supported catalysts, it was found that NH₃ does not simply decompose but also reacts with CO₂ to form *NCO, which serves as an intermediate in CH₄ formation. Our previous study showed that CH₄ formation proceeds via *NCO over Ni/Al₂O₃ catalyst, similar to the Mg–Al mixed oxide-supported catalysts.¹¹ Furthermore, *NCO formation was also observed over the Ni/MgO catalyst under CO₂ + NH₃ methanation conditions (Fig. S5). For this catalyst, the *NCO IR band decreased, and a gas-phase CH₄ band appeared at higher temperatures, similar to the behavior observed for Ni/Mg_0.5Al_0.5O_n (Fig. 3a). Therefore, it is suggested that CH₄ formation proceeds primarily via *NCO as an intermediate on all of the catalysts examined in this study.

Fig. 5 (a) Temperature-dependent changes in the DRIFTS peaks attributed to *NCO (100–500 °C) under H₂ flow (H₂: 3.9 mL min⁻¹, He: 8.7 mL min⁻¹) over the Ni/Mg_0.50Al_0.50O_n catalyst. Before measurement, the catalyst was exposed to CO₂ and NH₃ gas (CO₂: 1 mL min⁻¹, NH₃: 2.6 mL min⁻¹, He: 6.4 mL min⁻¹) at 300 °C and then purged with He gas. (b) Mass spectroscopy analysis (m/z = 15, CH₄) of the effluent gas at the outlet of the DRIFTS cell.

Scheme 1 Proposed reaction mechanism for CO₂ + NH₃ methanation.

The formation of *NCO occurs at low temperature (around 200 °C), where CH₄ formation is negligible (Fig. 3a and S5). Therefore, the formation rate of this intermediate likely has little influence on the overall CH₄ formation rate. As shown in Fig. 2f, NH₃ conversion during CO₂ + NH₃ methanation is a key factor determining the CH₄ formation rate. In this study, NH₃ conversion was estimated based on N₂ production, which arises from both NH₃ decomposition and *NCO hydrogenation (eqn (8)). This indicates that NH₃ conversion reflects the rates of these two processes. Thus, it is possible that both NH₃ decomposition and *NCO hydrogenation contribute to the overall CO₂ + NH₃ methanation activity of the catalysts.

3.3. Effect of the physicochemical properties of the catalysts on CO₂ + NH₃ methanation activity

We investigated which physicochemical properties of the supported Ni catalysts influence the rates of NH₃ decomposition and hydrogenation of *NCO to CH₄. First, the physicochemical properties of the supports were analyzed. Fig. S6 shows the X-ray diffraction patterns of Mg–Al mixed hydroxides (before calcination). Diffraction peak assignable to hydrotalcite were observed for Mg_xAl_1−x(OH)_n where x = 0.90 to 0.50. Formation of the hydrotalcite structure is favored when there is an excess of divalent cations (M²⁺) relative to trivalent cations (M³⁺). Therefore, the hydrotalcite structure is formed in Mg²⁺-rich compositions (i.e., Mg_0.90Al_0.10(OH)_n and Mg_0.75Al_0.25(OH)_n). The diffraction peaks of the hydrotalcite for Mg_0.75Al_0.25(OH)_n were sharper than those for Mg_0.90Al_0.10(OH)_n, indicating high crystallinity of the former. In addition to the hydrotalcite peaks, a diffraction peak assigned to Mg(OH)₂ was observed in Mg_0.90Al_0.10(OH)_n, suggesting that forming a highly crystalline hydrotalcite structure is difficult when the composition is too rich in Mg²⁺. Hydrotalcite peaks were also observed for Mg_0.50Al_0.50(OH)_n; however, additional peaks corresponding to MgAl₂(OH)₈ appeared, indicating that maintaining the hydrotalcite structure is difficult at high Al³⁺ content. Mg_0.25Al_0.75(OH)_n, which contained the highest Al³⁺ content among the prepared Mg–Al mixed samples, showed MgAl₂(OH)₈ as the main crystal phase.

Fig. 6a shows the diffraction patterns of Mg–Al mixed oxide supports. Diffraction peaks characteristic of MgO were observed in all samples. However, as the Al³⁺ content increased, additional peaks assignable to MgAl₂O₄ appeared. Fig. 6b presents an enlarged view of the MgO (220) diffraction peaks in Mg_0.90Al_0.10O_n, Mg_0.75Al_0.25O_n, and Mg_0.50Al_0.50O_n. With increasing Al³⁺ content, the peaks gradually shifted toward higher angles, suggesting incorporation of Al³⁺ ions, which have a smaller ionic radius than do Mg²⁺, into the MgO lattice (i.e., formation of Mg(Al)O solid solution). The peaks also broadened with increasing Al³⁺ content, indicating partial collapse of the MgO structure.

Fig. 6 (a) X-ray diffraction patterns of each Mg–Al mixed oxide support. (b) Enlarged views of the diffraction peaks for MgO (220) in MgO, Mg_0.90Al_0.10O_n, Mg_0.75Al_0.25O_n, and Mg_0.50Al_0.50O_n.

The morphology of the Mg–Al mixed oxides was examined using field-emission scanning electron microscopy (Fig. S7). Aggregates composed of flake-like primary particles with diameters ≥100 nm were observed in MgO (Fig. S7a and b). Similar aggregates were also observed in Mg_0.90Al_0.10O_n, although the flake-like primary particles were smaller (Fig. S7c and d). With increasing Al³⁺ content, the particles became rounder. For Mg_0.75Al_0.25O_n, Mg_0.50Al_0.50O_n, and Mg_0.25Al_0.75O_n, pores of approximately 100 nm were observed between the primary particles (Fig. S7e–j). In contrast, almost no pores were seen in Al₂O₃ (Fig. S7k and l), which exhibited a solid surface structure.

The surface structures of the Mg–Al mixed oxide supports were analyzed using nitrogen adsorption–desorption measurements. The N₂ adsorption–desorption isotherms are shown in Fig. S8. All supports exhibited typical type IV isotherms, indicating mesoporous structures. The Barrett–Joyner–Halenda plots derived from the N₂ adsorption isotherms are shown in Fig. 7. MgO showed a broad, pore-size distribution, including pores larger than 100 nm. Scanning electron microscopy images (Fig. S7a and b) revealed that these large pores corresponded to the spaces between the flake-like particles. Al₂O₃ primarily contained pores with diameters of 1–10 nm, which were too small to be detected in the scanning electron microscopy images. All of the Mg–Al mixed oxides had pores mainly with diameters of 10–100 nm, which were considered to arise from the gaps between primary particles smaller than 100 nm.

Fig. 7 Barrett–Joyner–Halenda plots of the Mg–Al mixed oxide supports.

The specific surface area (SSA) and pore volume (V_p) of the oxides are shown in Table 1. V_p values were obtained by analyzing the Barrett–Joyner–Halenda plots in the 1–100 nm range. Pore volumes divided into pore diameter ranges of 1–10 nm and 10–100 nm (V_{p 1–10} and V_{p 10–100}, respectively) are also shown. MgO exhibited the smallest SSA, probably due to it also having the smallest V_p. Al₂O₃ showed a high SSA (253 m² g⁻¹) due to its large number of small pores (i.e., a higher V_{p 1–10} than the other supports). Although the V_{p 1–10} of Mg_0.50Al_0.50O_n was smaller than that of Al₂O₃, it exhibited the largest SSA (259 m² g⁻¹) because its total V_p was the largest among the supports.

Table 1 Surface structures of the Mg–Al mixed oxide supports

Support	SSA^a (m^2 g^−1)	V p (cm^3 g^−1)	V p 1–10 (cm^3 g^−1)	V p 10–100 (cm^3 g^−1)
a Specific surface area. b Pore volumes estimated for the pore diameter ranges of 1–100 nm, 1–10 nm, and 10–100 nm, respectively, using the Barrett–Joyner–Halenda plots show in Fig. 7.
MgO	81.6	0.27	0.091	0.18
Mg0.90Al0.10On	151	0.83	0.075	0.75
Mg0.75Al0.25On	178	0.79	0.13	0.66
Mg0.50Al0.50On	259	1.05	0.17	0.88
Mg0.25Al0.75On	141	0.62	0.098	0.52
Al2O3	253	0.35	0.32	0.026

---

Next, the physicochemical properties of the supported Ni catalysts were analyzed. The reduction behavior of the Ni species on each Mg–Al mixed oxide support was analyzed by H₂ temperature-programmed reduction, where H₂O formation (m/z = 18) was used to estimate reduction behavior (Fig. S9). The maximum H₂O formation occurred at approximately 250 °C for all catalysts, showing no significant difference in Ni reducibility among the supports (Fig. S9a). In addition, CH₄ formation (m/z = 16) was observed at 500–600 °C for all catalysts (Fig. S9b), indicating that carbon species derived from the Ni precursor (Ni(C₅H₇O₂)₂) remained after calcination and were hydrogenated to CH₄. Based on these results, pre-reduction for the activity test was conducted at 600 °C, a temperature that is sufficient to reduce most Ni species and remove residual carbon species. Therefore, the influence of remaining carbon species on the CH₄ and CO yields in the activity test was considered negligible.

Fig. S10 shows the XRD patterns of each supported Ni catalyst after reduction at 600 °C. Diffraction peaks attributed to the support oxide and Ni metal were observed for each sample, confirming the reduction of Ni species. Broad Ni peaks were seen for catalysts using MgO and Mg–Al mixed oxide supports, indicating the formation of small Ni particles on these supports. In contrast, Al₂O₃ exhibited a sharp peak for Ni metal, suggesting that agglomeration of Ni nanoparticles occurred during reduction.

The crystallite sizes of Ni, estimated using the Scherrer formula based on the half–width of the diffraction peak of the Ni 200 plane (which does not overlap with the diffraction peaks of the support oxides), are summarized in Table 2. Although exact values were difficult to determine due to the broad Ni peaks, all Mg-containing catalysts exhibited small crystallite sizes of ≤5 nm, while Ni/Al₂O₃ exhibited a much larger size (>20 nm).

Table 2 Physicochemical properties of the Mg–Al mixed oxide-supported Ni catalysts

Catalyst	SSA (m^2 g^−1)	Ni crystallite size (nm)	Ni dispersion (%)
Ni/MgO	84.4	4.3	9.6
Ni/Mg0.90Al0.10On	125	5.4	11.4
Ni/Mg0.75Al0.25On	154	4.0	11.8
Ni/Mg0.50Al0.50On	197	3.9	12.6
Ni/Mg0.25Al0.75On	186	4.9	11.1
Ni/Al2O3	166	23.6	4.1

---

Fig. 8 shows HAADF-STEM images of Ni/MgO, Ni/Mg_0.50Al_0.50O_n, and Ni/Al₂O₃. For Ni/MgO, Ni particles of around 5 nm were mainly observed, with some approaching 10 nm. For Ni/Mg_0.50Al_0.50O_n, most of the Ni particles were ≤5 nm and uniformly dispersed on the support. In contrast, Ni/Al₂O₃ contained coarse particles measuring several tens to 100 nm alongside small particles with sizes below 10 nm. As shown in Table 1, the SSA and pore structures of the oxides differed, influencing Ni dispersion. The Al₂O₃ support, in which aggregation of Ni particles was observed, had a lower V_{p 10–100} than the other supports.

Fig. 8 High-angle annular dark-field scanning transmission electron microscopy images of Ni/MgO, Ni/Mg_0.50Al_0.50O_n, and Ni/Al₂O₃.

During impregnation, it is likely difficult for the Ni precursor to penetrate supports with small pores,³⁵ limiting the available dispersion sites and resulting in particle aggregation. In contrast, the Mg_0.50Al_0.50O_n support showed the highest V_{p 10–100} among the supports, facilitating Ni precursor penetration and uniform dispersion. Similarly, the other Mg–Al mixed oxide-supported catalysts also showed small (≤10 nm), uniformly dispersed Ni particles (Fig. S11), likely due to their larger V_{p 10–100} compared with the MgO and Al₂O₃ supports.

The Ni dispersion for each supported Ni catalyst was analyzed by H₂-PULSE chemisorption measurements (Table 2). Ni/Mg_0.50Al_0.50O_n exhibited the highest Ni dispersion among the catalysts, while Ni/Al₂O₃ exhibited the lowest. From the HAADF-STEM images (Fig. 8), Ni dispersion was found to correlate with Ni particle size. Therefore, among the Mg–Al mixed oxide-supported catalysts, aggregation of Ni particles was most effectively suppressed in Ni/Mg_0.50Al_0.50O_n. This is likely because the Mg_0.50Al_0.50O_n support had the largest V_10–100, providing abundant sites for Ni loading.

The correlation between CH₄ yield (at 500 °C) and Ni dispersion was investigated (Fig. 9). CH₄ yield increased almost linearly with Ni dispersion, indicating that Ni dispersion is a key physicochemical property in CO₂ + NH₃ methanation. Since NH₃ decomposition occurs on the Ni surface and CH₄ formation (i.e., hydrogenation of *NCO) occurs at the Ni–support interface (see Section 3.2.), Ni dispersion influences the number of sites for both processes. Although Ni/Mg_0.50Al_0.50O_n had similar basicity and almost the same intrinsic NH₃ decomposition activity as the other Mg–Al mixed oxide catalysts (Ni/Mg_0.90Al_0.10O_n and Ni/Mg_0.75Al_0.25O_n), its superior CO₂ + NH₃ methanation activity is attributed to its higher Ni dispersion, which provides more sites for hydrogenation of *NCO to CH₄.

Fig. 9 Correlation between CH₄ yield (at 500 °C) and Ni dispersion.

3.4. Stability test of Mg–Al mixed oxide-supported catalyst for CO₂ + NH₃ methanation

Finally, a stability test was performed over Ni/Mg_0.50Al_0.50O_n at 500 °C, the temperature that yielded the maximum CH₄ yield (Fig. 2a). CH₄ and CO yields and NH₃ conversion remained nearly constant for up to 100 h, indicating good catalyst stability (Fig. 10). Fig. S12 shows the XRD patterns of Ni/Mg_0.50Al_0.50O_n before and after the stability test. The Ni crystallite size changed only slightly (3.9 to 4.2 nm), indicating that Ni/Mg_0.50Al_0.50O_n does not lose active sites under prolonged exposure to the reaction atmosphere. Although the intensity of the MgAl₂O₄ diffraction peaks increased after the stability test, suggesting support sintering, this structural change is considered to have had little impact on catalyst activity.

Fig. 10 Stability test over Ni/Mg_0.50Al_0.50O_n at 500 °C with a space velocity of 7200 mL h⁻¹ g⁻¹.

4. Conclusion

Ni-loaded Mg–Al mixed oxides with varying Mg/Al ratios were prepared to investigate the relationship between their physicochemical properties and CO₂ + NH₃ methanation activity. Operando DRIFTS measurements revealed that *NCO species were formed from CO₂ and NH₃ on the support surface, and *NCO was hydrogenated to CH₄ at the Ni–support interface. These results indicate that Ni dispersion is a key physicochemical factor affecting catalytic activity, because a greater number of Ni–support interface sites enhance *NCO hydrogenation. The Mg–Al mixed oxide-supported catalysts showed moderate basicity and higher Ni dispersion than Ni/MgO and Ni/Al₂O₃ because of their favorable support structures. Among them, Mg_0.50Al_0.50O_n had the larger pore diameter and the largest pore volume, facilitating Ni precursor loading. Consequently, Ni/Mg_0.50Al_0.50O_n achieved the highest Ni dispersion and the best CO₂ + NH₃ methanation performance, yielding 57% CH₄ at 500 °C (SV: 3600 mL h⁻¹ g⁻¹), which is much higher than that of previously reported supported Ni catalysts (Table S3). The developed catalyst is expected to contribute to improving carbon capture and utilization efficiency using ammonia as an H₂ carrier, thereby helping to reduce carbon emissions in regions with limited renewable energy resources.

Conflicts of interest

There are no conflicts of interest to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI).

Supplementary information: detail preparation procedure of Mg–Al mixed hydroxide, ICP-MS measurement, activity test of NH₃ decomposition, CO₂ conversion estimated from unreacted CO₂, dependence of CH₄, CO yields and NH₃ conversion on space velocity, mass signals during DRIFTS measurements, DRIFTS spectrums, XRD patterns, SEM images, N₂ adsorption–desorption isotherms, temperature-programed reduction profiles, TEM images, XRD patterns of catalysts before and after stability test. See DOI: https://doi.org/10.1039/d5cy01327e.

Acknowledgements

This work was partly supported by JST FOREST (JPMJFR223N) and Akira Yoshino Research Grant of The Chemical Society of Japan.

References

1. R. Zhang, Z. Xie, Q. Ge and X. Zhu, J. CO2 Util., 2024, 89, 102973.
2. N. Rui, X. S. Zhang, F. Zhang, Z. Y. Liu, X. X. Cao, Z. H. Xie, R. Zou, S. D. Senanayake, Y. H. Yang, J. A. Rodriguez and C. J. Liu, Appl. Catal., B, 2021, 282, 119581.
3. G. Varvoutis, M. Lykaki, S. Stefa, V. Binas, G. E. Marnellos and M. Konsolakis, Appl. Catal., B, 2021, 297, 120401.
4. M. Sadiq, R. J. Alshehhi, R. R. Urs and A. T. Mayyas, Renewable Energy, 2023, 219, 119362.
5. N. Armaroli and V. Balzani, ChemSusChem, 2011, 4, 21–36.
6. R. Lan, J. T. S. Irvine and S. Tao, Int. J. Hydrogen Energy, 2012, 37, 1482–1494.
7. J. W. Makepeace, T. J. Wood, H. M. A. Hunter, M. O. Jones and W. I. F. David, Chem. Sci., 2015, 6, 3805–3815.
8. X. Hu, B. Guan, J. Chen, Z. Zhuang, C. Zheng, J. Zhou, T. Su, C. Zhu, S. Zhao, J. Guo, H. Dang, Y. Zhang, Y. Yuan, C. Yi, C. Xu, B. Xu, W. Zeng, Y. He, Z. Wei and Z. Huang, Fuel, 2025, 381, 133134.
9. H. Saima, R. Sunamoto, H. Miyaoka and T. Ichikawa, J. Chem. Eng. Jpn., 2023, 56(1), 2248176.
10. M. A. Uddin, Y. Honda, Y. Kato and K. Takagi, Catal. Today, 2017, 291, 24–28.
11. Y. Ueda, K. Nagaoka and K. Sato, JACS Au, 2025, 5, 6112–6126.
12. Y. Huang, H. Ren, H. Fang, D. Ouyang, C. Chen, Y. Luo, L. Lin, D. Wang and L. Jiang, Appl. Surf. Sci., 2024, 669, 160517.
13. T. Furusawa, H. Kuribara, K. Kimura, T. Sato and N. Itoh, Ind. Eng. Chem. Res., 2020, 59, 18460–18470.
14. J. Zhang, H. Xu and W. Li, Appl. Catal., A, 2005, 296, 257–267.
15. Y. Im, H. Muroyama, T. Matsui and K. Eguchi, Int. J. Hydrogen Energy, 2020, 45, 26979–26988.
16. J.-E. Min, Y.-J. Lee, H.-G. Park, C. Zhang and K.-W. Jun, Ind. Eng. Chem., 2015, 26, 375–383.
17. Y. Qiu, E. Fu, F. Gong and R. Xiao, Int. J. Hydrogen Energy, 2022, 47, 5044–5052.
18. A. Rizzetto, E. Sartoretti, M. Piumetti, R. Pirone and S. Bensaid, Chem. Eng. J., 2024, 501, 157585.
19. Q. Su, L. Gu, Y. Yao, J. Zhao, W. Ji, W. Ding and C.-T. Au, Appl. Catal., B, 2017, 201, 451–460.
20. L. J. I. Coleman, W. Epling, R. R. Hudgins and E. Croiset, Appl. Catal., A, 2009, 363, 52–63.
21. X. Yang, M. Huang, H. Huang, D. Li, Y. Zhan and L. Jiang, Int. J. Hydrogen Energy, 2022, 47, 22442–22453.
22. K. Sato, N. Abe, T. Kawagoe, S.-I. Miyahara, K. Honda and K. Nagaoka, Int. J. Hydrogen Energy, 2017, 42, 6610–6617.
23. S. Mohan and P. Dinesha, Chem. Pap., 2022, 76, 6551–6556.
24. S. Gražulis, D. Chateigner, R. T. Downs, A. F. Yokochi, M. Quirós, L. Lutterotti, E. Manakova, J. Butkus, P. Moeck and A. Le Bail, J. Appl. Crystallogr., 2009, 42, 726–729.
25. J. C. Ganley, F. S. Thomas, E. G. Seebauer and R. I. Masel, Catal. Lett., 2004, 96, 117–122.
26. A. M. Karim, V. Prasad, G. Mpourmpakis, W. W. Lonergan, A. I. Frenkel, J. G. Chen and D. G. Vlachos, J. Am. Chem. Soc., 2009, 131, 12230–12239.
27. NIST Chemistry WebBook, SRD 69, https://webbook.nist.gov/chemistry/form-ser/, (accessed 07–10, 2024).
28. S. Collins, M. Baltanas and A. Bonivardi, J. Catal., 2004, 226, 410–421.
29. A. Solis-Garcia, J. F. Louvier-Hernandez, A. Almendarez-Camarillo and J. C. Fierro-Gonzalez, Appl. Catal., B, 2017, 218, 611–620.
30. T. Bánsági, T. S. Zakar and F. Solymosi, Appl. Catal., B, 2006, 66, 147–150.
31. G. Piazzesi, D. Nicosia, M. Devadas, O. Kröcher, M. Elsener and A. Wokaun, Catal. Lett., 2007, 115, 33–39.
32. F. Barzagli, F. Mani and M. Peruzzini, Green Chem., 2011, 13, 1267–1274.
33. A. M. Bernhard, D. Peitz, M. Elsener, T. Schildhauer and O. Kröcher, Catal. Sci. Technol., 2013, 3, 942–951.
34. J. Ding, R. Ye, Y. Fu, Y. He, Y. Wu, Y. Zhang, Q. Zhong, H. H. Kung and M. Fan, Nat. Commun., 2023, 14, 4586.
35. Z.-W. Wu, J. Xiong, C.-W. Wang and Y.-H. Qin, Int. J. Hydrogen Energy, 2023, 48, 4728–4737.

---