Comparative Study of Mono- and Bimetallic (Ni–Co–Fe) Catalysts Supported on LaCeO₃ for Ammonia Decomposition

Abstract

Ammonia decomposition over non-precious metal thermos-catalysts offers a viable and cost-effective pathway for sustainable hydrogen production. In this study, LaCeO₃ perovskite was synthesized using a citric acid complexation method and employed as a support for mono- and bimetallic catalysts prepared by incipient wetness impregnation, maintaining a total metal loading of 10 wt%. Structural and surface properties were systematically investigated using BET, XRD, H₂-TPR, SEM, TEM, and CO₂-TPD. Among the monometallic catalysts (Ni, Co, and Fe), 10%Ni/LaCeO₃ exhibited the highest activity, which is attributed to its enhanced reducibility and optimal surface basicity, facilitating NH₃ activation. Bimetallic systems (Ni-Co, Ni-Fe, and Co-Fe) with equal metal loadings (5 wt% each) showed better activity compared to their monometallic counterparts following the order: 5%Ni–5%Co/LaCeO₃ > 5%Ni–5%Fe/LaCeO₃ > 5%Co–5%Fe/LaCeO₃. The improved performance of the Ni-Co system is due to structural interactions between Ni and Co, which promote hydrogen desorption and accelerate N–H bond cleavage, while suppressing nitrogen recombination as the rate-limiting step. Further systematic optimization of the Ni/Co ratio showed that 8%Ni–2%Co/LaCeO₃ had the highest catalytic activity with consistent performance over 50 h. This optimal composition provides a balanced distribution of active metallic sites and moderate-to-strong basic sites, enhancing NH₃ adsorption and intermediate transformation. These findings show that LaCeO₃-supported Ni-Co catalysts are promising candidates for efficient hydrogen production from ammonia without using noble metals.

1. Introduction

Hydrogen, as an ideal alternative to fossil energy, has the advantages of high energy density, zero carbon emissions, and easy access. However, the large-scale application of hydrogen is hindered by challenges related to its storage and transportation [1,2,3,4]. Ammonia as an ideal hydrogen carrier has the following advantages: low-cost production, a high energy density, ease of liquefaction for storage and transportation, and carbon-free participation in the whole process [5,6,7]. Nowadays, hydrogen production by ammonia decomposition has been regarded as a promising solution for hydrogen storage and transportation; thus, developing efficient and durable ammonia decomposition catalysts is the key to realizing this application [6]. In the last few decades, plenty of metal catalysts, such as Ruthenium (Ru), Nickel (Ni), Cobalt (Co), Iron (Fe), Copper (Cu), etc., have been found to have high ammonia decomposition activity [8,9,10,11,12]. Ru-based catalysts show the highest activity at low temperatures among them, but the scarce resources and exorbitant market costs of Ru hinder its large-scale industrial application. Much effort has therefore been devoted to developing non-precious metal catalysts for ammonia decomposition, and among them, Fe, Co, and Ni have attracted particular attention.

According to literature, the weak binding energy between metal (M) and nitrogen (N) in non-noble metal catalysts limits their effectiveness in ammonia decomposition. In a bimetallic system, the M-N binding energy can be tuned; hence, bimetallic catalysts have turned out to be potential materials for ammonia decomposition in comparison with their monometallic counterparts [13,14]. This is majorly attributed to the synergistic effect of the metal-metal combination in bimetallic catalysts [15]. Fujitani’s group examined the role of support materials in enhancing nickel’s catalytic activity for ammonia decomposition. Their findings revealed that the electronic structure of nickel, and consequently the Ni-N bond strength, is significantly influenced by the choice of support [16].

Over the past decades, researchers have examined various materials as support for H₂ production from NH₃ such as MgO, SiO₂, Al₂O₃, TiO₂, ZrO₂, carbon (activated), porous (meso and micro) materials, multi-walled carbon nanotubes (MWCNTs), etc. [13,17,18,19,20,21]. Literature research has shown that the basicity of a catalyst is crucial for its effectiveness in decomposing ammonia. Additionally, studies have demonstrated that using La₂O₃ or CeO₂ with a catalyst can enhance its intrinsic activity [22,23,24]. These rare earth oxides had been widely considered as catalyst carriers or important catalyst components, especially for ammonia decomposition reactions [25,26,27,28].

On the other hand, perovskite-type oxides (ABO₃) are proven remarkable materials due to their exceptional properties, like structural integrity even when modified with diverse elements or compositions, high conductivity, and high thermal stability [29]. Based on the above considerations, we chose LaCeO₃ perovskite oxide as support for the ammonia decomposition reaction. In this work, we studied several bimetallic systems on LaCeO₃ and compared the activity with their monometallic counterparts for ammonia decomposition. A review of the available literature indicates that studies on ammonia decomposition related to this system remain limited or have not been widely reported.

2. Results and Discussion

2.1. Textural Properties

The calcined catalysts are used for surface area analysis. The N₂ adsorption-desorption isotherms and pore-size distribution patterns of all prepared catalysts are shown in Figure 1 and Figure 2. The support LaCeO₃ showed type-IV isotherm with H3 hysteresis, indicating the presence of aggregates of plate-like particles giving rise to slit-shaped pores. All calcined monometallic catalysts displayed type-IV isotherms with an H3 hysteresis loop and showed mesopores with an average pore size in the range of 20–70 nm except 10%Fe/LaCeO₃ [30]. The 10%Fe/LaCeO₃ catalyst exhibited a type IV isotherm with H2 hysteresis, designating the existence of ink-bottle-shaped pores.

The BET specific surface area, total pore volume, and average pore diameter values are listed in

Table 1. Textural properties of catalysts supported on LaCeO₃.

Catalyst	Surface Area	Pore Volume	Pore Width
	(m2/g)	(cc/g)	(nm)
LaCeO3	6.6	0.026	52.0
10%Ni/LaCeO3	14.6	0.065	39.7
10%Co/LaCeO3	13.4	0.054	133.8
10%Fe/LaCeO3	21.4	0.078	93.3
5%Ni-5% Co/LaCeO3	16.3	0.073	133.8
5%Co-5% Fe/LaCeO3	12.7	0.062	191.7
5%Ni-5% Fe/LaCeO3	12.4	0.051	133.8
7% Ni-3%Co/LaCeO3	14.0	0.062	133.8
8%Ni-2%Co/LaCeO3	18.2	0.072	133.8

. The LaCeO₃ support showed a lower surface area than all catalysts prepared in this work. The BET specific surface areas increased after the addition of active metal, and the increment is high in 10%Fe/LaCeO₃ catalyst. This increase in catalyst surface area indicates either introducing pores or increased surface roughness within the composite, rather than pore blockage. As a result, there is a reduction in pore size upon metal in addition to the support. It is observed from Figure 1 that the LaCeO₃ showed pores in the 4 to 30 nm range. With the addition of cobalt to LaCeO₃ i.e., in 10%Co/LaCeO₃ there is a slight change in distribution of pores in the range of 5 to 20 nm. With the addition of Ni, the pores are predominantly distributed in the range of 2–20 nm. Finally, with the addition of Fe to LaCeO₃ the pores are distributed in the range of 5–15 nm, and the pore volume also changes. The order of surface area of single metallic catalysts is 10%Co/LaCeO₃ < 10%Ni/LaCeO₃ < 10%Fe/LaCeO₃. It was observed that the average pore width is higher in the 10%Co/LaCeO₃ catalyst than in the other mono-metallic catalysts supported on LaCeO₃.

On the other hand, the bimetallic catalyst showed both high pore volume and pore width, which indicates the pore size of the bimetallic catalyst markedly shifts to a larger pore size. As a result, the bimetallic catalysts were slightly higher in surface area. The order of surface area of bimetallic catalysts is 5%Ni-5% Fe/LaCeO₃ ~ 5%Co-5%Fe/LaCeO₃ < 5%Ni-5% Co/LaCeO₃.

2.2. XRD-Analysis

The XRD results of prepared lanthanum cerate are shown in Figure 3A. The diffraction peaks exhibit characteristic reflections of perovskite crystal structure, consistent with the standard JCPDS file No.: 01-080-5546, confirming cubic symmetry. As reported by Butt et al. [31], LaCeO₃ nanoparticles prepared via the precipitation method exhibit reflections at 2θ values of 27.79°, 31.93°, 45.51°, 54.74°, which are indexed to the (1,1,1), (2,0,0), (2,2,0) and (3,1,1) planes, respectively. The XRD pattern shown in Figure 3A displays similar diffraction peaks at corresponding 2θ positions, indicating good agreement with the reported structure. Furthermore, according to Mangu et al. [32], the planes (1,1,1) and (2,0,0) are highly reactive and play a significant role in enhancing the oxygen vacancy formation. Notably, no diffraction peaks corresponding to lanthanum oxide phases are observed, suggesting the complete formation of the perovskite structure.

The XRD patterns of mono- and bimetallic catalysts are presented in Figure 3B,C. Phase identification of the collected diffraction data was performed using Match! software. All catalyst samples exhibit diffraction patterns similar to that of the LaCeO₃ perovskite support, indicating that the perovskite structure is preserved even after metal addition. No additional diffraction peaks corresponding to Ni, Co, or Fe oxide phases are observed in any of the catalysts. The absence of distinct metal oxide reflections cannot be solely attributed to their amorphous nature or low crystallinity. Instead, several factors may contribute to this observation. First, the active metal species are likely highly dispersed on the LaCeO₃ support surface. When metal oxide particles are present as very small crystallites, their diffraction peaks become significantly broadened and weak, often falling below the detection limit of conventional XRD analysis. Second, the strong diffraction peaks associated with the crystalline LaCeO₃ perovskite structure can mask the weak reflections originating from highly dispersed metal oxide species. This effect is particularly significant in the bimetallic catalysts, where the individual metal loading is lower due to the distribution of the total metal content between two active components.

Furthermore, strong interactions between the transition metal species and the LaCeO₃ support may facilitate the formation of highly dispersed surface species or partial incorporation of metal ions into the perovskite lattice, thereby preventing the formation of large segregated crystalline metal oxide domains. Similar observations have been reported for supported Ni-, Co-, and Fe-based catalysts, where the absence of detectable metal oxide peaks has been associated with high metal dispersion and strong metal–support interactions [22,25]. The lack of any observable bulk metal oxide phases in the present study suggests that the active metals are well dispersed throughout the support matrix, which is beneficial for maximizing the number of accessible catalytic active sites for ammonia decomposition.

Additionally, the overall diffraction peak intensities of the catalysts are slightly lower than those of the support, indicating a reduction in crystallinity after metal incorporation. This decrease may result from slight structural distortion of the perovskite lattice and the introduction of defects or oxygen vacancies arising from metal-support interactions. These observations collectively confirm that the perovskite framework remains intact following metal loading while promoting a highly dispersed distribution of active metal species.

2.3. TPR Studies

The reducibility of all prepared catalysts and support was analyzed by H₂-TPR. The TPR patterns of support material & single metal catalysts are presented in Figure 4A. There is one small peak at 500 °C observed for the support material LaCeO₃ which might be due to the reduction of surface ceria species [23,33]. Based on the literature, the Ni catalysts, either on La₂O₃ or on CeO₂ showed reduction peaks in the temperature range of 220–430 °C which correspond to weak, medium, and strong interactions between NiO and the support [34]. The TPR profile of 10%Ni/LaCeO₃ showed a peak between 290 and 500 °C, which corresponds to weakly and medium interacted NiO species. A small lower temperature peak at 240 °C observed in the TPR pattern of the 10%Ni/LaCeO₃ catalyst that corresponds to non-interacted NiO species with support. According to literature, the TPR pattern of the cobalt catalyst shows two-step reduction profiles. In the first step Co₃O₄ reduced to CoO (Co₃O₄ + H₂ → 3CoO + H₂O), and in the second step CoO reduced to metallic cobalt (CoO + H₂ → Co + H₂O) [26,35]. The TPR profiles of 10%Co/LaCeO₃ showed multiple peaks that are related to the reduction of non-interacted and interacted cobalt species with support.

The TPR profile of iron oxide typically shows two successive reduction steps: (i) a small peak within the 350–450 °C temperatures, representing the reduction of Fe³⁺ to Fe²⁺ ions; and (ii) a broad reduction peak between 750 and 800 °C temperatures, indicating the subsequent reduction Fe²⁺ to Fe⁰ [36].

The TPR profile of 10%Fe/LaCeO₃ exhibits a low-intensity reduction peak around 400 °C corresponds to the reduction of Fe³⁺ to Fe²⁺. The reduction of Fe²⁺ to Fe⁰ is happening above 850 °C thus not detected in Figure 4A. These results clearly indicate a strong metal-support interaction in 10%Fe/LaCeO₃, which consequently leads to low metal reducibility in this catalyst.

Table 2. Hydrogen consumption of H₂-TPR profiles & TPD Quantification.

Catalyst	H2 Consumption	Total Basicity
	(µ mol g−1)	(µ mol g−1)
LaCeO3	241.7	285.74
10%Ni/LaCeO3	1530.2	581.72
10%Co/LaCeO3	537.1	494.01
10%Fe/LaCeO3	442.8	183.40
5%Ni-5% Co/LaCeO3	727.8	485.47
5%Co-5% Fe/LaCeO3	613.8	423.11
5%Ni-5% Fe/LaCeO3	769.0	315.18
7% Ni-3%Co/LaCeO3	972.3	474.88
8% Ni-2%Co/LaCeO3	1053.2	448.94

summarizes the hydrogen consumption values obtained from the TPR experiments for all catalysts. The results clearly indicate that the 10%Ni/LaCeO₃ catalyst exhibits the highest reducibility, while the 10%Fe/LaCeO₃ catalyst shows the lowest. The theoretical hydrogen consumption for the 10%Ni/LaCeO₃ catalyst was calculated to be approximately 1704 µmol g⁻¹, compared to the experimentally measured value of about 1530 µmol g⁻¹. Based on these values, the degree of reduction was determined to be 89.8%.

The TPR patterns of bimetallic catalysts are presented in Figure 4B. There is clear change in TPR profiles of bimetallic catalysts compared to single-metal catalysts. The Ni-Co catalyst showed a reduction peak in the range of 300–400 °C which corresponds to overlapped peaks for NiO and Co₃O₄ reduction. A small reduction peak also observed at 250 °C corresponds to the reduction peak of NiO not interacted with the support. The TPR profiles of 5%Ni-5%Fe/LaCeO₃ catalyst showed two main reduction peaks, a peak ranging from 207 to 310 °C and another peak between 317 and 560 °C. All these TPR peaks correspond to the reduction of NiO and Fe₃O₄ species. The TPR profiles of 5%Co-5%Fe/LaCeO₃ catalyst showed three reduction peaks, a peak ranging from 185 to 303 °C and another peak between 360 and 585 °C. In addition, there is a small peak observed in the TPR profile of 5%Co-5%Fe/LaCeO₃ catalyst at a temperature range from 303 to 357 °C. All these TPR peaks correspond to the reduction of Co₃O₄ and Fe₃O₄ species. From Figure 4B and Table 2, it can be concluded that the metal reducibility increased in the bimetallic catalyst compared to the single metallic catalyst, especially in comparison to 10%Fe/LaCeO₃ & 10% Co/LaCeO₃. On other hand, the 10%Ni/LaCeO₃ catalyst showed the highest reducibility among all catalysts. The TPR results clearly indicate that Ni-based catalysts supported on LaCeO₃ exhibit enhanced metal reducibility. Consequently, a higher density of accessible active sites is expected for the Ni-based catalysts.

2.4. CO₂-TPD

CO₂-TPD analysis was performed to evaluate the total basicity of the catalysts prepared in this study. Figure 5A presents the CO₂ desorption profiles of the LaCeO₃ support and all monometallic catalysts, while Figure 5B shows the profiles for the bimetallic catalysts. The CO₂-TPD pattern of the LaCeO₃ support (Figure 5A) displays three types of basic sites: (i) weak basic sites (<150 °C), (ii) moderate basic sites (200–400 °C), and (iii) strong basic sites (650–850 °C). Upon incorporation of single active metals (Ni, Co, or Fe) into the LaCeO₃ support, noticeable changes in the distribution of basic sites occur, particularly in the Ni/LaCeO₃ and Fe/LaCeO₃ catalysts. In contrast, only a slight change in basicity is observed for the 10%Co/LaCeO₃ catalyst relative to the support. Among all monometallic catalysts, 10%Ni/LaCeO₃ exhibits the highest concentration of moderate basic sites.

Table 2 summarizes the basicity values obtained from the CO₂-TPD experiments. The basicity of the LaCeO₃ support is lower than that of all the catalysts, except for 10%Fe/LaCeO₃, which shows the lowest basicity overall, likely due to strong Fe-support interactions. The 10%Ni/LaCeO₃ catalyst demonstrates the highest total basicity, dominated by medium-strength basic sites. The 10%Co/LaCeO₃ catalyst shows the second-highest basicity, with strong basic sites contributing most significantly. In contrast, the 10%Fe/LaCeO₃ catalyst exhibits the lowest basicity, attributed to strong metal-support interactions. The overall observation is that incorporating active metals significantly enhances the basicity of LaCeO₃, with Ni imparting the highest increase, while Fe results in the lowest basicity due to strong metal-support interactions.

The basicity profiles of the bimetallic catalysts, shown in Figure 5B, follow trends similar to those of the monometallic samples. The 5%Ni–5%Co/LaCeO₃ catalyst demonstrates the highest basicity among the bimetallic formulations, with a significant contribution from medium-strength basic sites. In contrast, the 5%Co–5%Fe/LaCeO₃ and 5%Ni–5%Fe/LaCeO₃ catalysts exhibit a predominance of strong basic sites. Among the bimetallic catalysts, 5%Ni–5%Fe/LaCeO₃ shows the lowest total basicity. Overall, bimetallic catalysts maintain similar basic site distributions to monometallic catalysts but exhibit enhanced basicity depending on the metal combination.

2.5. SEM Studies

The morphology of the supported nickel, cobalt and iron catalysts are studied by scanning electron microscopy characterization. The SEM images of all prepared catalysts are displayed in Figure 6. According to literature [32], the LaCeO₃ perovskite exhibits the morphology of nanoaggregates with a microstructure that has a large amount of honeycomb-like macropores. The addition of metal to LaCeO₃ morphology changed. The 10%Ni/LaCeO₃ catalyst showed aggregates of plates with narrow edges having slit-shaped pores with high surface roughness.

The 10%Co/LaCeO₃ catalyst contains aggregations of small plates in different shapes. The 10%Fe/LaCeO₃ catalyst contains bulk aggregates having space inside. From Figure 6 it seems 10%Fe/LaCeO₃ catalyst possesses pores with narrow width outside and wider pore inside. The SEM results are in good agreement with the BET data. The morphology of a bimetallic catalyst slightly changes in comparison to that of a monometallic catalyst. The 5%Ni-5%Co/LaCeO₃ catalyst showed aggregates of small plates with increased porosity. The bimetallic catalyst having Fe showed similar morphology with less porosity.

The EDS spectra for the selected area of the 10%Ni/LaCeO₃ catalyst are presented in Figure S1A. The quantitative results showed the presence of 39%, 37%, and 11% atomic weight percentages of La, Ce, and Ni, respectively. The scanning image and the EDS spectra analysis of the 10%Co/LaCeO₃ sample are displayed in Figure S1B. The quantification results showed that 33%, 33% and 12% atomic percentages of La, Ce, and Co, respectively, are present in the 10%Co/LaCeO₃ sample. The scanning image and the EDS spectra analysis of the 10%Fe/LaCeO₃ sample are displayed in Figure S1C. The quantification results showed that 41%, 35%, 9% and 13% atomic percentages of La, Ce, Fe, and O, respectively, are present in the 10%Fe/LaCeO₃ sample. Overall, the EDS analyses confirm the successful incorporation and uniform distribution of Ni, Co, and Fe on the LaCeO₃ support, with elemental compositions close to the intended loadings. The presence of La and Ce as major components in all samples, along with the corresponding transition metals, verifies the effective formation of the metal-supported LaCeO₃ catalysts without significant elemental loss or segregation during preparation.

The scanning image and the EDS spectra analysis of the 5%Ni-5%Co/LaCeO₃ sample are displayed in Figure S2A. The quantification results showed that 36%, 35%, 4%, 4% and 18% atomic percentages of La, Ce, Ni, Co, and O, respectively, are present in the 5%Ni-5%Co/LaCeO₃ sample. The Scanning image and the EDS spectra analysis of 5%Ni-5%Fe/LaCeO₃ sample are displayed in Figure S2B. The quantification results showed that 36%, 36%, 5%, 5% and 16% atomic percentages of La, Ce, Ni, Fe, and O, respectively, are present in 5%Ni-5%Fe/LaCeO₃ sample. The Scanning image and the EDS spectra analysis of 5%Co-5%Fe/LaCeO₃ sample are displayed in Figure S2C. The quantification results showed that 36%, 36%, 4%, 4% and 14% atomic percentages of La, Ce, Co, Fe, and O, respectively, are present in the 5%Co-5%Fe/LaCeO₃ sample.

These EDS results clearly demonstrate the successful co-incorporation of Ni-Co, Ni-Fe, and Co-Fe bimetallic species onto the LaCeO₃ support, with elemental compositions closely matching the nominal loadings. The comparable La and Ce contents across all samples and the uniform presence of both metal components indicate effective dispersion of the bimetallic phases without preferential segregation, confirming the formation of well-defined bimetallic LaCeO₃-based catalysts.

2.6. Catalyst Performance Studies

All prepared catalysts were evaluated for their catalytic performance in ammonia decomposition to hydrogen. The reaction was carried out using pure ammonia (100 vol%) at ambient pressure with a space velocity of 6000 mL g_cat⁻¹ h⁻¹, over a temperature of 300–600 °C. Figure 7A presents the catalytic activities of the monometallic catalysts, namely 10%Ni/LaCeO₃, 10%Co/LaCeO₃ and 10%Fe/LaCeO₃. As expected, ammonia conversion increased with rising temperature, which is consistent with the endothermic nature of the ammonia decomposition reaction [35].

Among the Ni, Co, and Fe catalysts on LaCeO₃, the 10%Ni/LaCeO₃ catalyst showed high activity. At 600 °C; the 10%Ni/LaCeO₃ catalyst achieved an NH₃ conversion of 95%, whereas 10%Co/LaCeO₃ and 10%Fe/LaCeO₃ catalysts showed lower conversions of 90% and 88% respectively. To better elucidate the differences in catalytic performance, the activities of the catalysts at 550 °C are compared in Figure S3A. A clear variation in activity among the monometallic catalysts is observed, with 10%Ni/LaCeO₃ showing the highest activity and 10%Fe/LaCeO₃ the lowest.

Perovskite oxides have strong electron-donating capability due to the presence of abundant basic sites, which facilitates electron transfer to the supported metal species [37]. According to Yin et al., [38] the conductivity of the support is very important for high catalytic efficiency, as it promotes electron transfer from the support to metal. This process facilitates the recombinative desorption of surface N atoms. Furthermore, the incorporation of active metal increases the basicity of monometallic catalyst, thereby facilitating nitrogen recombination.

Zhang et al. [39] reported that La and Ce promoters significantly influence the physicochemical properties of Ni catalysts, leading to enhanced activity in ammonia decomposition. Similarly, Liu et al. [40] demonstrated that La₂O₃ increases weak and medium-strength basic sites and improves the reducibility of metal species, particularly Ni.

A similar trend is observed in LaCeO₃-supported catalysts, where the Ni-based catalyst exhibited a high degree of reduction (90%), resulting in a greater number of active sites. In contrast, the 10%Fe/LaCeO₃ catalyst showed the lowest activity, attributed to strong metal-support interaction that limits metal reducibility (degree of reduction 26%). Although the 10%Co/LaCeO₃ catalyst exhibited significant basicity, particularly strong basic sites, its activity remained lower than that of the Ni catalyst due to its comparatively lower metal reducibility (31.5%). These results indicate that Ni is more favorable than Co and Fe for ammonia decomposition over LaCeO₃ support.

Figure 7B shows the catalytic activities of the bimetallic catalysts 5%Ni-5%Co/LaCeO₃, 5%Ni-5%Fe/LaCeO₃, 5%Co-5%Fe/LaCeO₃. The bimetallic catalysts are more active than the mono metallic catalysts except 5%Co-5%Fe/LaCeO₃. Catalysts 5%Ni-5%Co/LaCeO₃ and 5%Ni-5%Fe/LaCeO₃ showed 100% conversion at 600 °C.

The bimetallic catalysts showed higher surface area values than the pure support and monometallic catalysts. Notably, the metal reducibility increased, especially in Fe-based catalysts relative to their monometallic catalysts. On the other hand, the Ni-Co catalyst showed lower metal reducibility and reduced basicity compared to 10%Ni/LaCeO₃. The results suggest that the metal-support interactions decreased in Ni-Fe and Co-Fe systems whereas in Ni-Co system the interactions slightly increased relative to their mono metallic catalysts.

Despite its relatively lower basicity and reducibility, the Ni-Co catalyst exhibited superior catalytic activity among all catalysts studied. Haihua et al. [14] studied Ni, Co, and Ni-Co catalysts on CeO₂ support for the ammonia decomposition reaction. According to them, the synergy between Co and Ni can decrease the bond energy between N and the active phase. Hence, the nitrogen re-combinative desorption is fast, which is the rate-limiting step of the NH₃ decomposition reaction. The reduced bond energy can be used to explain the much-enhanced catalytic activity of the Ni-Co bimetallic catalyst. A similar synergistic effect is expected in this study; the synergy between Ni-Co catalysts leads to higher activity. To further clarify activity differences, catalytic performance at 550 °C was compared (Figure S3B). The sequential order activity by bimetallic catalysts is as follows: 5%Ni-5%Co/LaCeO₃ > 5%Ni-5%Fe/LaCeO₃ > 5%Co-5%Fe/LaCeO₃.

2.7. Optimization of Metal Loading in Bimetallic Catalysts

To finalize the study and optimize the active metal loading in Ni-Co bimetallic catalysts, samples with different Ni/Co weight percentages, namely 7%Ni–3%Co/LaCeO₃ and 8%Ni–2%Co/LaCeO₃ in addition to the 5%Ni–5%Co/LaCeO₃ catalyst, were synthesized and evaluated. As summarized in Table 1, the specific surface area increased with increasing Ni content, following the order: 5%Ni–5%Co < 7%Ni–3%Co < 8%Ni–2%Co.

The XRD patterns of the bimetallic catalysts with varying metal loadings are presented in Figure S4. No significant changes in the diffraction patterns were observed upon variation of the Ni/Co ratio, indicating that changes in metal loading did not substantially affect the crystal structure or crystallite size of the catalysts.

The H₂-TPR profiles of the calcined bimetallic catalysts used for loading optimization are shown in Figure 8A. As discussed previously, the Ni-Co catalysts exhibit a main reduction peak in the temperature range of 300–400 °C, corresponding to the overlapping reduction of NiO and Co₃O₄ species interacting with the support. In addition, a minor reduction peak observed around 250 °C is attributed to the reduction of NiO species weakly interacting with the support.

Comparison of the TPR profiles of 5%Ni–5%Co/LaCeO₃ and 7%Ni–3%Co/LaCeO₃ reveals no significant difference in their reduction behavior. However, for the 8%Ni–2%Co/LaCeO₃ catalyst, the reduction peak in the range of 170–275 °C becomes more intense and dominant, corresponding to an increased fraction of non-interacting or weakly interacting NiO species. Furthermore, as shown in Table 2, the overall reducibility of the catalysts increases with increasing Ni content from 5 to 8 wt%, indicating enhanced reduction characteristics at higher Ni loadings.

The basicity of the catalysts, evaluated by CO₂-TPD, is shown in Figure 8B. As previously discussed, Ni-Co bimetallic catalysts exhibit a significant contribution from medium-to-strong basic sites. A comparison among 5%Ni–5%Co/LaCeO₃, 7%Ni–3%Co/LaCeO₃, and 8%Ni–2%Co/LaCeO₃ reveals a similar distribution of basic sites; however, the total basicity is highest for the 5%Ni–5%Co/LaCeO₃ catalyst. In particular, the concentration of medium-strength basic sites is more pronounced in this sample. As summarized in Table 2, the total basicity decreases with increasing Ni content from 5 to 8 wt%, indicating that a reduction in Co loading leads to a corresponding decrease in surface basicity.

Overall, the combined H₂-TPR and CO₂-TPD results demonstrate that metal reducibility is enhanced by higher Ni loading, whereas surface basicity is favored by higher Co content on the LaCeO₃ support.

Transmission electron microscopy (TEM) was employed to investigate the morphology of the calcined 8%Ni–2%Co/LaCeO₃ catalyst. A representative TEM image is shown in Figure 9A, and the corresponding energy-dispersive X-ray (EDX) spectra are presented in Figure 9B. The calcined catalyst exhibits a nano-platelet–like morphology with stacked structures forming relatively large agglomerates, which can provide extended interfacial contact between the active metal species and the LaCeO₃ support.

The EDX analysis confirms the presence of O, La, Ce, Ni, and Co in the calcined 8%Ni–2%Co/LaCeO₃ catalyst, with atomic percentages of 71.77, 9.04, 10.64, 6.13, and 2.43%, respectively, in good agreement with the nominal composition.

The HRTEM image of the 8%Ni–2%Co/LaCeO₃ catalyst is presented in Figure 10a. Clear and well-defined lattice fringes are observed, with interplanar spacings of 0.148 nm and 0.124 nm, which are attributed to the NiO (220) and CoO (220) crystallographic planes, respectively [41]. According to the literature [41,42,43], these planes correspond to face-centered cubic (FCC) metal oxide phases at the interface. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) combined with EDS elemental mapping was carried out to evaluate the spatial distribution of the active metal species. The HAADF-STEM images acquired at a 10 nm scale, together with the corresponding elemental mapping (Figure 10b), reveal a uniform dispersion of Ni and Co species over the LaCeO₃ support. Such homogeneous dispersion and suitable metal-support interactions are known to enhance N-H bond activation by increasing the density of accessible bimetallic active sites, while simultaneously suppressing metal sintering at reaction temperatures. Consequently, these structural characteristics favor higher ammonia conversion and contribute to the excellent catalytic stability observed during prolonged ammonia decomposition tests.

Figure 11A presents the catalytic performance of all Ni-based catalysts prepared in this study for ammonia decomposition. The introduction of optimally balanced Ni-Co compositions over LaCeO₃ support significantly enhances catalytic activity. The ammonia conversion increased as the Ni loading was raised from 5 to 8 wt%. Among the investigated samples, 8%Ni–2%Co/LaCeO₃ exhibited the highest catalytic activity. Based on ammonia conversion, catalytic performance follows the order: 8%Ni-2%Co/LaCeO₃ > 7%Ni-3%Co/LaCeO₃ > 5%Ni-5%Co/LaCeO₃ > 10%Ni/LaCeO₃ > 10%Co/LaCeO₃.

The variation in the H₂ formation rate at 550 °C is clearly illustrated in Figure 11B. These results indicate that the incorporation of a small amount of Co is beneficial for enhancing the activity of Ni-based catalysts supported on LaCeO₃. He et al. [14] studied the Ni-Co bimetallic system on CeO₂ support and reported that doping of Co into the Ni-based catalysts significantly reduced the reaction order and followed the Temkin-Pyzhev mechanism. A similar phenomenon is expected in this study.

According to Fu et al., [44], the addition of cobalt to the Ni system reduces the energy barrier for the associative decomposition of M-N to form N₂ which is the step dictating ammonia decomposition. Cobalt reveals strong metal-support interactions, which can promote its dispersion on the LaCeO₃ surface and facilitate electronic interaction between Co and Ni species [45]. This interaction may enable electron transfer from the support to Ni via Co, thereby enhancing the intrinsic activity of Ni active sites. In contrast, higher Co loadings may partially block Ni-Co interfacial contact or hinder Ni accessibility, leading to a gradual decline in catalytic performance. These observations suggest that low, optimized Co content is sufficient to maximize the promotional effect.

The time-on-stream stability test was conducted using the 8%Ni-2%Co/LaCeO₃ catalyst at 550 °C and 600 °C for a total duration of 50 h, and the results are presented in Figure 12. Initially, the catalyst was evaluated at 550 °C for 15 h. Subsequently, the reaction temperature increased to 600 °C and was maintained for 20 h. Finally, the temperature decreased back to 550 °C, and the catalyst was tested for an additional 15 h. Remarkably, the 8%Ni-2%Co/LaCeO₃ catalyst exhibited excellent stability in terms of NH₃ conversion throughout the entire 50 h test period at both 550 °C and 600 °C. A time-on-stream study at a higher GHSV (12,000 h⁻¹) was conducted using the 8%Ni–2%Co/LaCeO₃ catalyst at 550 °C for 25 h. The results are presented in Figure S6. The conversion decreased to 85%. However, the catalyst exhibited stable performance even at this high space velocity throughout the 25 h test.

Previous studies have reported that synergistic interactions between Ni and Co species can influence the adsorption–desorption behavior of hydrogen and thereby enhance ammonia decomposition activity [14,46]. However, direct evidence for such effects, including hydrogen desorption measurements or operando spectroscopic investigations, was not obtained in the present study. Therefore, the enhanced catalytic performance of the Ni–Co bimetallic catalysts observed herein should be interpreted as being potentially associated with synergistic metal–metal interactions reported in the literature.

In addition, surface basic sites enhance electron donation to adsorbed ammonia molecules, thereby promoting N-H bond activation. While Co addition over LaCeO₃ increases surface basicity and strengthens metal-support interactions, resulting in decreased metal reducibility. Ni species supported on LaCeO₃ exhibit higher reducibility but comparatively lower basicity. Therefore, appropriate Co doping achieves a balance between metal reducibility and surface basicity, leading to improved catalytic activity of Ni-based catalysts for ammonia decomposition.

The reducible nature of the LaCeO₃ support (TPR results) may also contribute to catalytic performance through possible hydrogen spillover phenomena. Hydrogen atoms generated on Ni and Co active sites during ammonia decomposition can potentially migrate to the support surface via the metal–support interface, where they interact with oxygen vacancies and reducible surface species. Such processes may assist in the removal of adsorbed hydrogen from metallic active sites and contribute to enhanced catalytic activity [47].

Arrhenius plots of ln (k) versus 1/T for M/LaCeO₃ catalysts (M = Ni, Co, Fe, 5%Ni–5%Co, 7%Ni–3%Co, and 8%Ni–2%Co) are presented in Figure S5, and the corresponding apparent activation energies (Eₐ) are summarized in the inset. The monometallic catalysts 10%Ni/LaCeO₃, 10%Co/LaCeO₃, and 10%Fe/LaCeO₃ exhibit apparent activation energies of 68.66, 72.05, and 82.25 kJ mol⁻¹, respectively, indicating comparatively slower ammonia decomposition kinetics.

In contrast, the Ni-Co bimetallic catalysts show substantially lower activation energies, with values of 55.52, 52.52, and 51.92 kJ mol⁻¹ for 5%Ni-5%Co/LaCeO₃, 7%Ni-3%Co/LaCeO₃, and 8%Ni-2%Co/LaCeO₃, respectively. The progressive decrease in activation energy with increasing Ni content and optimized Co loading clearly demonstrates the promotional effect of Co on Ni-based catalysts.

These kinetic results are consistent with the catalytic activity trends and H₂ formation rates discussed earlier, where 8%Ni-2%Co/LaCeO₃ exhibited the highest ammonia conversion and hydrogen production. The reduced activation energy can be attributed to the synergistic interaction between Ni and Co. Among all investigated catalysts, 8%Ni-2%Co/LaCeO₃ exhibits the lowest activation energy, further confirming that an optimal, low Co loading maximizes the synergistic effect and leads to superior intrinsic catalytic activity for hydrogen production from ammonia.

Table 3. Comparison of catalytic activities with reported literature values at 550 °C.

Catalyst	GHSV (ml/g/h)	Conversion (%)	H2 Production (mmol/g/min)	Ref.
Ni1Co9/CeZrYO	6000	91.6	6.13	[48]
Co/CeZrYO	6000	67.8	4.54	[48]
Ni/CeZrYO	6000	66.7	4.47	[48]
4Ni/Ce0.8Zr0.2O2-SA	4500	95.7	3.99	[49]
CeNiO3	6000	99.0	6.5	[50]
30%Ni/CeO2	6000	75.0	4.8	[50]
Ni5Co5/SiO2	6000	94.7	6.34	[13]
La-Ce-Co	6000	92.0	6.2	[26]
La0.5Ce0.5NiO3	6000	99.0	6.7	[26]
Ni7.5Co2.5/CeO2	30,000	40	13.4	[14]
Ru/La0.33Ce0.67 a	6000	100	6.7	[51]
Ru/Ce1 a	6000	98.1	6.6	[51]
5Co-BaCeO	6000	80.0	5.40	[25]
Mg-Co-Fe	6000	99.0	6.63	[11]
10%Ni/La2O3	6000	62.7	4.20	[37]
40%Ni/Al2O3	6000	70.0	4.7	[52]
40%Ni/LaAlO3	6000	78.0	5.2	[52]
10%Ni/MgO-WI	12,000	25.0	3.35	[53]
10%Ni/MgO-CPT	12,000	97.3	17.1	[53]
10%Ni/LaCeO3	6000	76.0	5.09	This study
10%Co/LaCeO3	6000	68.8	4.61	This study
10%Fe/LaCeO3	6000	62.9	4.21	This study
5%Ni-5%Co/LaCeO3	6000	82.6	5.53	This study
5%Ni-5%Fe/LaCeO3	6000	80.1	4.72	This study
5%Co-5%Fe/LaCeO3	6000	70.5	5.41	This study
7%Ni-3%Co/LaCeO3	6000	85.3	5.71	This study
8%Ni-2%Co/LaCeO3	6000	90.3	6.05	This study

^a at 450 °C.

compares the NH₃ decomposition performance of the catalysts prepared in this study with those reported in the literature under comparable reaction conditions at 550 °C. The monometallic catalysts investigated in this work (10%Ni/LaCeO₃, 10%Co/LaCeO₃, and 10%Fe/LaCeO₃) exhibit ammonia conversions of 76.0%, 68.8%, and 62.9%, respectively, which are comparable to Ni- and Co-based catalysts supported on conventional oxides such as CeZrYO and Al₂O₃ reported in previous studies. Notably, the bimetallic catalysts synthesized in this study demonstrate improved performance relative to their monometallic counterparts. In particular, 5%Ni-5%Co/LaCeO₃ achieves an ammonia conversion of 82.6% with an H₂ production rate of 5.53 mmol g⁻¹ min⁻¹ at a GHSV of 6000 mL g⁻¹ h⁻¹, which is comparable to or higher than several reported non-noble metal catalysts, including Co-BaCeO and Ni/CeZrYO systems. Although noble-metal-based catalysts such as Ru/CeO₂ and Ru/La_0.33Ce_0.67O_x exhibit superior performance, the Ni–Co/LaCeO₃ catalysts developed in this study show competitive activity among noble-metal-free systems, highlighting the effectiveness of LaCeO₃-supported bimetallic formulations for ammonia decomposition.

3. Experimental

3.1. Chemicals

All chemicals were obtained from Fluka and Aldrich and used as received without further purification. The chemicals Fe(NO₃)₃·9H₂O (Fluka), Ni(NO₃)₂·6H₂O (Fluka), Co(NO₃)₂·6H₂O (Fluka), La(NO₃)₃·6H₂O (Aldrich), Ce(NO₃)₃·6H₂O (Aldrich), and citric acid (Aldrich) were used in catalyst preparation.

3.2. Synthesis of LaCeO₃ Perovskite Oxides

The support material (ABO₃-like) was prepared by using a combustion preparation technique with citric acid (CIA) as an organic fuel and enhancing gelation additive (metal precursor: CIA = 2). To prepare LaCeO₃ support, (0.0181 mol) of La(NO₃)₃·6H₂O and (0.0181 mol) of Ce(NO₃)₃·6H₂O salts were dissolved in de-ionized water (50 mL). The metal-to-CIA molar ratio of 1 was made by dissolving 0.0367 mol of CIA in 50 mL of de-ionized H₂O. Afterwards, the CIA solution was slowly added, while stirring, to the solution of metal precursors. To make a viscous gel, heat the mixture of metal precursor and CIA on a hot plate until it thickens. Additionally, the thick gel was immersed in a water bath and heated to 90 °C for a full day. After 24 h, the catalyst was dried in a standard oven set at 110 °C for 12 h. The resulting solid ground into a powder and subjected to a 6-h thermal treatment in a traditional muffle furnace at 650 °C. The synthesized oxide support is labelled LaCeO₃.

3.3. Preparation of LaCeO_3—Supported Catalysts

Using the standard wet-impregnation technique in a rotary evaporator, all catalysts were prepared with calcined LaCeO₃ oxide support material. In all catalysts, the total active metal loading was maintained as a nominal 10 wt%. After preparation, the catalyst was dried at 110 °C for 12 h. Next, the catalyst was calcined at 550 °C for 5 h in a muffle furnace. Nickel, cobalt and iron are impregnated on LaCeO₃ to prepare single active metallic catalysts and labelled as 10%Ni/LaCeO₃, 10%Co/LaCeO₃ and 10%Fe/LaCeO₃. The Ni-Co, Ni-Fe, and Co-Fe were impregnated on LaCeO₃ to prepare bimetallic active catalysts maintained equal weight percentage (5 wt%) and labelled as 5%Ni-5%Co/LaCeO₃, 5%Ni-5%Fe/LaCeO₃, 5%Co-5%Fe/LaCeO₃. The other catalysts were prepared with different metal loadings labeled as 7%Ni-3%Co/LaCeO₃ and 8%Ni-2%Co/LaCeO₃.

3.4. Catalyst Characterization

The crystallographic properties of the calcined catalysts were examined by X-ray powder diffraction (XRD) using a Shimadzu XRD-6100 diffractometer (Shimadzu Corporation, Kyoto, Japan). Diffraction patterns were collected employing Cu K_α radiation (λ = 1.54056 Å) over a 2θ range of 20–80° at a scanning rate of 5° min⁻¹. The diffractometer was operated at an accelerating voltage of 30 kV and a current of 30 mA. The diffraction data obtained were analyzed using Match! software 1.1 (Crystal Impact, Bonn, Germany), and phase identification was carried out by comparison with standard patterns from the Powder Diffraction File (PDF) database.

The textural properties of the calcined catalysts, including specific surface area and pore size distribution, were evaluated using a NOVA 2200e surface area and pore size analyzer (Quantachrome Instruments, Boynton Beach, FL, USA). Nitrogen adsorption-desorption measurements were carried out at −196 °C. Prior to analysis, the samples were degassed under vacuum at 200 °C for 2 h to remove physically adsorbed species.

The hydrogen temperature-programmed reduction (H₂-TPR) experiment was performed using a Micromeritics Autochem II 2950 (Micromeritics Instrument Corp., Norcross, GA, USA) apparatus equipped with a thermal conductivity detector (TCD) under atmospheric pressure. Initially, 100 mg of the calcined sample was loaded into a U-shaped quartz reactor. Prior to the reduction step, the catalyst was pretreated by heating it linearly from room temperature to 120 °C under a He gas flow of 15 mL min⁻¹ and maintaining this temperature for 60 min. This pretreatment is performed to remove any adsorbed water from the catalyst surface. Following the pretreatment, the catalyst cooled to 40 °C in the He atmosphere. Finally, the pretreated catalyst was subjected to a linear temperature ramp from 40 °C to 850 °C at a heating rate of 10 °C min⁻¹ in a flowing mixture of 10% hydrogen (H₂) and 90% argon (Ar) with a total flow rate of 40 mL min⁻¹.

A CO₂-TPD analysis was performed to evaluate the basic properties of the prepared catalysts. These analyses were carried out using the same instrument employed for H₂-TPR. Initially, 100 mg of calcined catalyst was reduced in a 10% H₂ and 90% Ar gas mixture (50 mL min⁻¹) at 550 °C for 2 h. The sample was then cooled to 50 °C under a helium flow (50 mL min⁻¹). Subsequently, a gas mixture containing 10% CO₂ in He (50 mL/min) was introduced into the sample bed for 1 h. After CO₂ adsorption, the catalyst was purged with He for 30 min. Finally, the sample was heated from 50 to 850 °C, and the amount of desorbed CO₂ was measured using a thermal conductivity detector (TCD).

The morphological properties of all prepared catalysts were determined using Scanning Electron Microscopy (SEM) with a JEOL JSM-6390LV instrument (JEOL Ltd., Akishima, Tokyo, Japan) operating at an accelerating voltage of 20 keV. This system was contained with an energy dispersive X-ray spectrometry (EDS) system for elemental analysis.

3.5. Evaluation of Catalytic Performance

A fixed-bed microreactor with an outer diameter of 9 mm was used for catalyst activity. 0.2 g of catalysts was used to load the reactor for ammonia decomposition. Before the test, the catalyst was reduced under a mixture of gases (H₂/N₂ = 1:1 volume ratio) at 550 °C for 5 h. After reduction, the system was cooled down to 300 °C and flushed with N₂ gas. A space velocity of 6000 h⁻¹ to feed pure 100% NH₃ into the reactor. The experiments were conducted at a temperature range of 300–600 °C with 50 °C step intervals. The results were gathered once the NH₃ decomposition reaction was stabilized at each temperature. The Varian 450 GC was used to test the gases at the outlet of the reactor with the help of a TCD. The following calculation was used to figure out the conversion % (X) of NH₃.

$$Ammonia conversion\left(X\%\right)={\displaystyle \frac{n_{{NH}_3}^i-n_{{NH}_3}^o}{n_{{NH}_3}^i}}⨯100$$

where nⁱ_NH3 = NH₃ number of moles at reactor inflow and n^o_NH3 = moles of NH₃ at reactor outflow

$${\mathrm{H}}_2 \mathrm{formation} \mathrm{rate} (\mathrm{mmol}/\mathrm{g}/\min )={\displaystyle \frac{\frac{V_{NH3}}{22.4}⁕Conversion(X)⁕\raisebox{1ex}{$3$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}{m_{cat}}}$$

The rate of H₂ formation is calculated using the above equation, where V_NH3 represents ammonia volumetric flow rate.

4. Conclusions

LaCeO₃ perovskite support was successfully synthesized using the citric acid complexation method and subsequently employed for the preparation of mono- and bimetallic catalysts via the impregnation technique. Monometallic M/LaCeO₃ catalysts (M = Ni, Co, and Fe) with a fixed metal loading of 10 wt% were evaluated for ammonia decomposition. Among these catalysts, 10%Ni/LaCeO₃ exhibited superior catalytic activity compared to 10%Co/LaCeO₃ and 10%Fe/LaCeO₃. This enhanced performance is attributed to the higher reducibility of Ni species combined with a moderate surface basicity, which together facilitate efficient N-H bond activation. In contrast, 10%Fe/LaCeO₃ showed the lowest activity, which can be ascribed to strong metal-support interactions leading to decreased metal reducibility and lower surface basicity.

Bimetallic catalysts containing equal weight percentages of the active metals were also synthesized and tested. All bimetallic formulations demonstrated improved catalytic performance relative to their monometallic counterparts. 5%Ni–5%Co/LaCeO₃ outperformed both 5%Ni–5%Fe/LaCeO₃ and 5%Co–5%Fe/LaCeO₃, highlighting the beneficial synergistic interaction between Ni and Co. This synergy is proposed to weaken the binding strength of surface nitrogen species on active sites, thereby facilitating recombinative desorption of N₂ during ammonia decomposition.

To enhance catalytic performance further, the Ni/Co ratio was systematically optimized. Among the investigated compositions, 8%Ni–2%Co/LaCeO₃ exhibited the highest catalytic activity. The superior performance of this catalyst is attributed to an optimal balance between high metal reducibility, enhanced surface basicity, and suitable Ni-Co synergistic interactions, which collectively promote hydrogen formation and improve intrinsic reaction kinetics.