Ce–La ratio-controlled structural transitions in Ni/Ce–La oxide catalysts for direct decomposition of methane

Abstract

Ce–La mixed oxides represent widely applicable catalyst supports owing to their controllable crystal structures, oxygen storage capacity, and electronic properties. In this study, Ni/Ce_1−xLa_x catalysts were systematically synthesized to investigate how Ce–La composition determines the lattice structure, redox behavior, and the resulting catalytic performance. Structural analyses by XRD and H₂-TPR revealed that moderate La incorporation (x ≤ 0.3) expands the CeO₂ fluorite lattice and enhances oxygen vacancy formation, thereby stabilizing highly dispersed Ni⁰ nanoparticles. XPS and soft XAS confirmed that this composition maximizes the Ce³⁺ fraction and optimizes the electronic interaction between Ni and the support. By contrast, excessive La content (x ≤ 0.5) induces a structural transition toward perovskite phases (LaNiO₃ and La₂NiO₄) and secondary La-based oxides (La₂O₂CO₃ and La(OH)₃), which embed Ni within the lattice and reduce its accessibility as active sites. These structural differences critically influence the quality of carbon products formed via methane decomposition. The optimally doped Ni/Ce_0.7La_0.3 catalyst exhibited remarkable performance, producing CO₂-free hydrogen together with highly crystalline CNTs featuring narrow diameters, high graphitization, and strong π* transitions.

---

1. Introduction

Supported metal catalysts have attracted considerable attention due to their unique catalytic properties in heterogeneous reactions. While extensive research studies have focused on the dispersion, oxidation state, and stability of the active metal species, equal emphasis has been placed on the characteristics of the support itself.^1,2 Such studies demonstrate that the types of support materials, the incorporation of multiple cations, and even the morphology of the support strongly influence catalytic activity, selectivity, and stability.³ These findings collectively indicate that the support does not play a passive role; rather it actively governs metal–support interactions (MSI), redox behavior, and the stabilization of active sites. Moreover, the support often serves as storage for lattice oxygen and a source of defect sites that participate in redox cycling and reactant activation, influencing intrinsic performance and long-term stability of the catalysts.^4,5 Accordingly, they can dramatically change the adsorption–desorption behavior of key intermediates, thereby determining the dominant reaction pathway. In other words, the support does not merely help the active metal; it actively tailors the whole catalyst system, participating in catalytic reactions and modulating both activity and selectivity together with the active metal species.

Amongst various lanthanide oxides, ceria (CeO₂) has been considered as one of the most promising and effective support materials in supported catalysts, owing to its ability to enhance metal dispersion, promote redox cycling, and strengthen metal–support interactions.⁶ Based on these attributes, CeO₂-supported catalysts have been widely applied in thermo-catalytic reactions such as steam reforming, dry reforming of methane, the water-gas shift reaction, and selective catalytic reduction.^6–8 To further improve oxygen ion conductivity and facilitate rapid oxygen transport, CeO₂ has frequently been combined with trivalent lanthanide dopants, such as gadolinium (Gd), samarium (Sm), and lanthanum (La).^9–11 Despite forming a trivalent oxide, lanthanum is known to exhibit high solubility (up to ca. 50 mol%) in the fluorite lattice of ceria. Ce–La mixed oxides can therefore retain a disordered fluorite phase or undergo structural transitions toward perovskite-like arrangements depending on the Ce/La ratio.¹² Such transformations are not merely compositional. They constitute structural engineering strategies that modify lattice constants, oxygen vacancy concentrations, and MSI at a fundamental level. Nevertheless, the mechanistic role of Ce–La structural transitions in governing physicochemical properties of active metals remains insufficiently understood.

Catalytic decomposition of methane (CDM) provides a unique pathway for the co-production of CO₂-free hydrogen and graphitic carbon nanomaterials such as carbon nanotubes (CNTs). Unlike non-catalytic thermal decomposition, which requires extremely high temperatures and typically yields low-quality carbon, the catalytic route enables the reaction to proceed under milder conditions with enhanced selectivity toward crystalline carbon structures.^13–15 A primary requirement for catalysts is to maintain stable activity under harsh operating conditions: elevated reaction temperature and continuous carbon deposition on the catalyst surface.¹⁶ These factors often lead to particle sintering, encapsulation by carbon layers, and ultimately rapid deactivation of conventional catalysts. To address these challenges, Ni has been the most widely employed active metal because of its high intrinsic activity for C–H bond cleavage, favorable selectivity toward CNT growth, and cost-effectiveness compared to noble metals.^13,15 However, the performance of Ni catalysts critically depends on the characteristics of the support to stabilize Ni nanoparticles and tune their electronic structure. The support not only influences metal dispersion and reducibility but also determines the extent to which carbon deposition can be mitigated during long-term operation.⁴ In this context, lanthanide oxides such as CeO₂ and La₂O₃ have attracted significant attention. When employed as supports or dopants, these oxides exhibit enhanced redox properties, facilitate the generation and migration of oxygen vacancies, and strengthen metal–support interactions.¹⁷ These features are particularly promising for improving the stability and performance of Ni-based catalysts, as demonstrated in this study. Thus, employing Ce–La-based mixed oxides as catalyst supports represents a promising approach for the efficient co-production of hydrogen and CNTs via the methane decomposition reaction. A series of Ce–La mixed oxides with varying Ce/La ratios were synthesized by a co-precipitation method, followed by the addition of Ni through the incipient wetness impregnation technique. The catalyst samples were characterized by a suite of X-ray and electron-microscopy techniques, and their catalytic performance was evaluated to establish an integrated, multidimensional correlation between catalytic methane decomposition (CMD) and the physicochemical and structural properties of the supports.

2. Experimental

2.1 Catalyst preparation

In this study, a Ce_1−xLa_x support material was synthesized using a co-precipitation method. First, cerium and lanthanum nitrate hexahydrates were dissolved in distilled water (2.0 L). The weight ratios of Ce to La were adjusted to 1 : 0, 0.9 : 0.1, 0.7 : 0.3, 0.5 : 0.5, 0.3 : 0.7, 0.1 : 0.9, and 0 : 1. A precipitating solution containing 15 wt% potassium carbonate was added dropwise until the pH reached 10.1. The resulting slurry was aged at room temperature for 4 h, and the precipitate was thoroughly washed several times to remove any impurities. After washing, it was dried at 110 °C for 24 h and calcined in air at 600 °C for 4 h.

Ni, using nickel nitrate hexahydrate as a precursor, was loaded onto the Ce_1−xLa_x support via wet impregnation. The loading amount of active metals was fixed at 10 wt%, and after impregnation, the samples were dried at 110 °C for 24 h and then calcined in air at 600 °C for 4 h. The catalysts are denoted as Ce_1−xLa_x, where x represents the weight fraction of Ce.

2.2 Catalyst characterization

The specific surface area, total pore volume, and N₂ adsorption–desorption isotherms of the powdered samples were determined via N₂ adsorption at −196 °C using the Brunauer–Emmett–Teller (BET) method with a BELSORP-max apparatus (BEL Japan, Inc., Japan). Powder XRD analysis was conducted using a D/MAX-2500 diffractometer (Rigaku, Japan) with Cu-Kα radiation over a 2θ range of 10–90° to investigate the crystal structures of the as-prepared and reduced catalysts. H₂-TPR analysis was performed using an Autochem II 2920 analyzer (Micromeritics, USA). The surface electronic states and elemental compositions of the reduced samples were characterized by XPS. Measurements were conducted using an ESCALAB Mark II spectrometer (Vacuum Generators, UK) equipped with Al-Kα radiation (hν = 1486.6 eV) at a constant pass energy of 50 eV. The binding energies were calibrated with reference to the C 1s peak at 285.0 eV. Quantitative analysis was performed via peak deconvolution and integration of the fitted peak areas to determine the relative elemental concentrations. Soft X-ray absorption spectroscopy (XAS) experiments were conducted at beamline 7.3.1, Advanced Light Source, Lawrence Berkeley National Laboratory (USA). Data acquisition was performed in the total fluorescence yield (TFY) mode, providing a probing depth of approximately 100 nm. All spectra were normalized to the beam flux, which was measured using an upstream gold mesh. The photon energies in the C K-edge and Ni L₃-edge spectra were calibrated using a standard sample with a resolution of 0.2 eV. CO₂-TPD analysis was conducted using an Autochem II 2920 analyzer (Micromeritics, USA) to evaluate the basicity of the prepared catalysts. Before the measurements, the catalyst (∼100 mg) was loaded into a quartz U-tube reactor and reduced under a 10 vol% H₂/Ar flow at 500 °C for 1 h. After cooling to 50 °C, the sample was exposed to a 10 vol% CO₂/He stream for 1 h to allow adsorption. The physically adsorbed CO₂ was subsequently removed by purging with pure He. The sample was then heated from 50 to 900 °C at a ramping rate of 5 °C min⁻¹. The number of basic sites (mmol g⁻¹) was quantified by integrating the desorption peaks.

CH₄-TPD analysis was performed using a BELCAT-II instrument (BEL Japan Inc., Japan) to investigate the CH₄ adsorption and desorption characteristics of the reduced catalysts. For the analysis, the catalyst (∼100 mg) was loaded into a quartz tubular reactor and reduced under a 10 vol% H₂/N₂ flow at 500 °C for 2 h. After cooling to 50 °C, the sample was exposed to a 5 vol% CH₄/N₂ gas mixture for 1 h to enable methane adsorption, followed by a 30 min N₂ purge to remove physically adsorbed species. The TPD measurements were conducted by heating the sample from 50 to 900 °C at a rate of 5 °C min⁻¹ under a continuous N₂ flow. During the desorption process, the evolution of CH₄ (m/z = 16), CO₂ (m/z = 44), and CO (m/z = 28) was monitored in real-time using a quadrupole mass spectrometer to evaluate CH₄ decomposition behavior.

The morphological characteristics of the carbon materials grown on the catalysts were examined by field-emission scanning electron microscopy (FE-SEM) using a Regulus 8220 microscope (HITACHI, Japan). Microstructural features and elemental distribution were examined using transmission electron microscopy (TEM) and energy-dispersive X-ray spectroscopy (EDX) mapping, performed on a F200 microscope (JEOL, Japan) operated at 200 kV. Raman spectroscopy was performed to evaluate the crystallinity of the carbon materials using an NRS-5100 spectrometer (JASCO Co., Japan) equipped with a 532 nm excitation source from an Nd-YAG laser. The spectra were recorded in the wavenumber range of 1000–4000 cm⁻¹. Thermogravimetric (TG) analysis was conducted using an SDT-Q600 thermal analyzer (TA Instruments, USA) under a controlled air atmosphere. The samples were heated from room temperature to 1000 °C at a constant heating rate of 10 °C min⁻¹, and the corresponding weight changes were continuously monitored to assess the thermal stability and compositional characteristics until a steady state was achieved.

2.3 Catalyst performance test

A thermogravimetric (TG) analyzer and tubular chemical vapor deposition (CVD) reactor were employed to conduct the catalytic reactions. The TG analyzer (Fig. 1(a)) was particularly useful for methane decomposition experiments, as it enabled quantitative analysis of the catalytic decomposition behavior, carbon yield, and temperature-dependent reaction trends using a sensitive balance. A catalyst sample (5 mg) was placed in a crucible mounted on an analytical balance and pretreated under a 10% H₂/N₂ flow at 500 °C for 1 h for reduction. Subsequently, the temperature was increased to 600 °C at a rate of 10 °C min⁻¹ and held for 1 h under specific reaction conditions. The reactant gas, a mixture of 50% CH₄ in N₂, was introduced into the catalytic reactor at a weight hourly space velocity (WHSV) of 360 L g_cat⁻¹ h⁻¹ based on CH₄ gas. After the reaction, the carbon yield was estimated by monitoring the weight change of the crucible containing the reaction products. The methane decomposition rate (R_CH4,TG) for each catalyst was evaluated using eqn (1), where m_i is the initial mass of the catalyst, Δm is the change in the mass of sample during the reaction and M_c is the molecular weight of carbon.

Fig. 1 Schematic of the experimental setup: (a) thermogravimetric analyzer and (b) tubular chemical vapor deposition reactor.

A tubular CVD reactor (Fig. 1(b)) was used for large-scale carbon production and analysis of the reaction gases, with the temperature of the Pt crucible positioned at the center of the reactor controlled by infrared radiation. The catalyst (50 mg) was placed in the crucible of the reactor, reduced under a 10% H₂/N₂ flow at 500 °C for 1 h, and then the reaction gas (50% CH₄/N₂) was introduced at 600 °C. This reaction gas was supplied to the catalytic reactor at a WHSV of 36 L g_cat⁻¹ h⁻¹ based on CH₄. The compositions of the reaction gases were analyzed in real-time using an AO2000 analyzer (ABB, UK) equipped with a non-dispersive infrared gas sensor (Uras 26) and thermal conductivity detector (Caldos 27), enabling continuous monitoring. Following the reaction, the crucible containing the products was weighed directly using an analytical balance, and the carbon yield (C_yield) was calculated using eqn (2). The methane decomposition rate (R_CH4,CVD)and hydrogen production rate (R_H2,CVD) in the tubular CVD reactor were given by eqn (3) and (4), respectively.

In eqn (3) and (4), F_CH4,in and F_CH4,out refer to the inlet and outlet flow rates of methane, respectively. F_CH4,out was determined as F_total,out × P_CH4,out, where F_total,out was calculated from F_N2,in/P_N2,out. Eqn (4) provides the calculation formula based on the methane decomposition reaction (CH₄ (g) → C (s) + 2H₂ (g)), where no considerable side reactions are detected.

R_H2,CVD = R_CH4,CVD × 2 (4)

3. Results and discussion

3.1 Structural and textural properties of Ni/Ce_1−x-La_x catalysts

Powder XRD provides the structural framework against which subsequent spectroscopic and catalytic results are interpreted. XRD analysis was performed to determine the crystal structure of the catalyst before and after reduction. The analysis identifies the crystallographic phases of the Ce–La mixed-oxide supports and Ni-containing species. It also monitors La-induced lattice expansion via systematic 2θ shifts of fluorite reflections, and reveals potential composition-dependent transitions toward perovskite-like phases as the Ce/La ratio varies.

Fig. 2(a) shows the XRD patterns of the Ni/Ce_1−xLa_x (x = 0–1.0) catalysts calcined at 600 °C. The Ni/Ce_1.0, Ni/Ce_0.9La_0.1, and Ni/Ce_0.7La_0.3 samples exhibit diffraction peaks corresponding to the CeO₂ phase (JCPDS 34-0394), along with peaks attributed to the NiO phase (JCPDS 47-1049, denoted by ). A shift of the CeO₂ diffraction peaks toward lower angles is observed with increasing La concentration up to 30%, which is attributed to the incorporation of La³⁺ into the CeO₂ lattice.¹² When the La content exceeds 50%, peak positions and relative intensities change markedly, indicating the emergence of mixed oxides such as LaNiO₃ () and La₂NiO₄ () and a structural transition of the catalyst from the fluorite to the perovskite phase.^17,18 In the perovskite structures, Ni³⁺ is incorporated into the crystal lattice. Accordingly, no distinct diffraction peaks corresponding to NiO are observed in the La-rich catalysts. Unlike the NiO species supported on fluorites, this structure is less likely to participate directly in the reaction.¹⁹ Characteristic peaks corresponding to La₂O₂CO₃ () were also detected for the Ni/Ce_0.1La_0.9 and Ni/La_1.0 catalysts. This phase likely formed from the reaction between La and carbonate precursors, indicating incomplete decomposition of the carbonate precursor.²⁰

Fig. 2 X-ray diffraction patterns of Ni/Ce_1−xLa_x catalysts in (a) calcined and (b) reduced states.

Fig. 2(b) shows the XRD patterns of the catalysts after reduction at 500 °C. For the CeO₂-dominant catalysts (Ce 0.5–1.0), the fluorite-type CeO₂ phase remains unchanged, which is consistent with those observed in the as-synthesized samples. The presence of metallic Ni peaks () indicates the successful reduction of Ni species.²¹ In contrast, for Ni/Ce_0.1La_0.9 and Ni/La_1.0, additional diffraction peaks appeared, which can be attributed to La₂NiO₄ () and La(OH)₃ (). Since these phases are absent prior to reduction, their appearance suggests structural decomposition and phase segregation during H₂ treatment. The appearance of these La-containing phases indicates structural instability, implying that uniform active species were not successfully formed.^17,19 Furthermore, incomplete reduction under the applied conditions likely led to the formation of secondary phases such as La₂O₃, La(OH)₃, and La₂NiO₄.²⁰ These changes reflect a structural transformation from a CeO₂-type fluorite structure toward complex La-containing oxide phases as the La content increases.

The structural transformation of the lattice observed with the XRD analyses is further supported by the following TEM-based analyses of the reduced catalyst samples. Fig. 3(a) shows that high Ce content catalysts exhibit well-defined, spherical Ni nanoparticles on the surface. However, as the La content increases above 0.5, the reduced Ni nanoparticles are no longer distinguishable in the TEM images, and the EDS maps confirm that Ni is uniformly distributed alongside Ce and La within the catalyst particles. This observation supports the hypothesis that, in La-rich environments, Ni is incorporated into perovskite-like structures (LaNiO₃ and La₂NiO₄) within the oxide matrix, which inhibits the segregation of reduced Ni on the surface.

Fig. 3 (a) TEM-EDS elemental mapping images of Ni, Ce, La, and O species and (b) atomistic structural models of LaNiO₃, La₂NiO₄, and Ni/CeO₂.

Fig. 3(b) shows atomic models illustrating the structural differences in Ni incorporation depending on the Ce/La ratio. In La-rich environments, the formation of perovskite-type structures such as LaNiO₃ and La₂NiO₄ is thermodynamically favored, with Ni atoms structurally embedded within the La–O lattice, as represented by the left and middle models in Fig. 3(b).^18,19 This configuration effectively stabilizes Ni within the oxide framework, restricting its reduction and subsequent segregation into metallic nanoparticles on the surface. In contrast, the right model in Fig. 3(b), representing the high Ce content condition, shows Ni species segregated on the CeO₂ surface as metallic nanoparticles, rather than being incorporated into the oxide lattice.⁶ This structural arrangement facilitates the reduction of Ni²⁺ to metallic Ni⁰, resulting in the formation of well-defined Ni nanoparticles observed in the TEM images, which is consistent with the clear metallic Ni diffraction peaks in the XRD patterns after reduction.

Thus, the combined XRD and TEM-EDS results strongly suggest that the Ce–La mixed oxide system effectively tunes the reducibility and distribution of Ni species by controlling the structural evolution of the catalyst, thereby modulating the metal–support interactions and catalytic performance during methane decomposition. This structural transition was accompanied by a notable increase in the particle size and decrease in the BET surface area, as summarized in Table 1. The perovskite-type oxides typically form dense, highly crystalline structures with low surface areas and limited porosity.^18,19 Moreover, the Ni species incorporated into these lattice frameworks are less likely to be exposed on the surface as active sites. These structural changes are likely to limit the catalytic activity by reducing both surface accessibility and Ni dispersion.²²

Table 1 Textural properties of the investigated catalysts and their corresponding reaction rates

Sample	CH4 conversion rate^a [mmol g^−1 min^−1]	Relative standard deviation^a [%]	BET surface area [m^2 g^−1]	Metal particle size^b [nm]	Metal dispersion^b [%]	Ni loading^c [wt%]	Ce/La ratio^c [wt%]
a Measured by TG analysis (Fig. 8(a)). b Estimated by CO chemisorption. c Measured by ICP analysis.
Ni/Ce1.0	2.59	1.81	39.5	35.6	2.84	10.4	1.0 : 0
Ni/Ce0.9La0.1	3.66	1.25	15.0	41.3	2.45	11.6	0.9 : 0.1
Ni/Ce0.7La0.3	11.97	0.39	14.1	45.6	2.22	11.0	0.69 : 0.31
Ni/Ce0.5La0.5	8.24	0.44	9.3	55.9	1.72	10.7	0.48 : 0.52
Ni/Ce0.3La0.7	3.82	0.83	5.5	92.3	1.09	11.8	0.31 : 0.70
Ni/Ce0.1La0.9	4.22	0.89	5.5	108.5	0.93	11.8	0.89 : 0.19
Ni/La1.0	1.99	1.06	7.0	73.0	1.39	11.2	0 : 1.0

---

3.2 Physicochemical properties influenced by the Ce–La effect

The reduction characteristics of the calcined catalysts were analyzed using H₂-TPR, as shown in Fig. 4. Prior to discussing the reduction behavior, ICP analysis confirmed that the Ni loading and Ce/La weight ratios of all catalysts were consistent with the nominal compositions, indicating that unintended compositional deviation did not influence the H₂-TPR profiles. All samples exhibit reduction peaks corresponding to the stepwise reduction of the Ni species and associated structural transitions. For Ni/Ce_1.0, a distinct reduction peak is observed at 220 °C, corresponding to the reduction of surface NiO species weakly interacting with the CeO₂ support, while a second peak at 302 °C is ascribed to the reduction of bulk NiO species more strongly interacting with the support. A broad peak above 700 °C is associated with the Ce⁴⁺ → Ce³⁺ reduction.^25,26

Fig. 4 Redox properties of Ni/Ce_1−xLa_x catalysts characterized by H₂ temperature-programmed reduction.

As the La content increases to x = 0.3 (Ni/Ce_0.7La_0.3), the main NiO reduction peak gradually shifts to higher temperatures, indicating that balanced La incorporation strengthens the metal–support interaction, as discussed in the previous section. Such a strong interaction facilitates the maintenance of highly dispersed and reducible Ni species during reduction.²⁷ Alternatively, for La-rich samples (x ≥ 0.5), the reduction peaks shift back to lower temperatures, and broad reduction features appear over a wide temperature range, especially in the Ni/Ce_0.1La_0.9 and Ni/La_1.0 catalysts. This profile is attributed to the formation of various La-based oxide phases, as confirmed by XRD. These phases tend to contain weakly bound or phase-separated Ni species.¹⁸ Specifically, Ni/La_1.0 shows three well-defined peaks at 344, 464, and 601 °C, which are assigned to the stepwise reduction of LaNiO₃ → La₂NiO₄ → La₂O₃ + Ni⁰, respectively.^18,20 The presence of La₂O₂CO₃, which typically decomposes in the temperature range of 400–600 °C, may also contribute to these reduction steps.²¹ Overall, the H₂-TPR profiles reveal that the La content influences the reduction behavior of NiO species. At lower La contents, the growth of the low-temperature reduction peak indicates facilitated reduction of surface NiO, consistent with improved oxygen mobility and vacancy formation. With further La incorporation, however, the reduction features progressively shift toward the medium- and high-temperature regions, reflecting stronger Ni–O–Ce/La interactions and partial stabilization of bulk NiO.^12,25,27 This trend explains that excessive La doping disrupts the fluorite lattice through excessive lattice expansion and structural distortion, which gradually undermines the stability of the metal–support interactions and deteriorates the overall structural integrity of the mixed oxide matrix.²⁸

XPS analysis was conducted to investigate the surface composition and oxidation states of the catalysts, offering insight into the electronic structure and metal–support interactions that govern their catalytic properties. Therefore, X-ray photoelectron spectroscopy (XPS) was carried out to determine the oxidation states of Ce in the Ni/Ce_1−xLa_x catalysts. The relative Ce³⁺/Ce⁴⁺ ratio, as derived from the Ce 3d spectra (Fig. 5(a)), serves as a key descriptor of oxygen vacancy formation and provides complementary evidence for the trends observed in the reduction behavior. The Ce³⁺ species are labeled as v₀, v′, u₀, and u′, whereas the Ce⁴⁺ species are labeled as v, v″, v‴, u, u″, and u‴.^29–31 The relative Ce³⁺ concentration, summarized in Table 2, increases with the La content up to x = 0.3, reaching a maximum of 43.7%. This indicates that moderate La doping enhances the formation of Ce³⁺ species, which are often associated with the presence of oxygen vacancies. These oxygen vacancies facilitate improved redox behavior and can promote the activation of CH₄ molecules.³² Furthermore, a slight shift in the binding energies of the Ce 3d peaks is observed upon the addition of La. This suggests an electronic structure modification around the Ce–O environment, possibly due to lattice distortion or electron redistribution induced by La³⁺ substitution.^31,33 The La 3d spectra were also examined to investigate the electronic structure of the La species within the Ce–La mixed oxide system. As shown in Fig. 5(b), all La-containing samples exhibit characteristic La 3d_5/2 and La 3d_3/2 spin–orbit doublets, corresponding to La³⁺ in an oxide environment.^32,34 As the La ratio decreases, the La 3d peaks of the Ce–La mixed oxides shift slightly to lower binding energies, indicating that the La³⁺ species are not present as phase-separated La₂O₃ domains but are instead incorporated into the CeO₂ lattice, forming a more homogeneous mixed oxide phase.^12,35

Fig. 5 X-ray photoelectron spectroscopy and Raman spectroscopy of Ni/Ce_1−xLa_x catalysts: (a) Ce 3d, (b) La 3d, and (c) O 1s XPS spectra, and (d) Raman spectra.

Table 2 Surface atomic ratios of Ce and O species, and binding energies of La species determined by XPS

Sample	Relative concentration [%]						Binding energy [eV]
	Ce 3d		O 1s				La 3d5/2
	Ce^3+	Ce^4+	O^2−	O^−/O^2−	CO3^2−/OH^−	OH^−
Ni/Ce1.0	29.5	70.5	67.6	25.0	—	7.4	—
Ni/Ce0.9La0.1	31.5	68.5	41.7	19.5	19.7	19.2	832.7
Ni/Ce0.7La0.3	43.7	56.3	29.4	32.5	20.8	17.3	832.4
Ni/Ce0.5La0.5	36.1	63.9	39.7	22.4	23.8	14.2	833.0
Ni/Ce0.3La0.7	32.1	67.9	17.1	10.0	55.2	17.7	834.1
Ni/Ce0.1La0.9	33.0	67.0	10.1	25.4	34.6	29.9	834.2
Ni/La1.0	—	—	20.8	18.0	37.5	23.7	833.8

---

The O 1s XPS spectra of the catalysts were analyzed to examine the surface oxygen chemical states, as shown in Fig. 5(c). The O 1s XPS spectra of the as-prepared catalysts were deconvoluted into four primary components corresponding to lattice oxygen (O²⁻, ∼528.5 eV), surface oxygen species (O⁻/O²⁻, ∼530.1 eV), carbonate or hydroxyl groups (CO₃²⁻/OH⁻, ∼531.2 eV), and surface hydroxyls (OH⁻, ∼532.5 eV).^36,37 As summarized in Table 2, the relative content of lattice oxygen (O²⁻) is highest in Ni/Ce_1.0 and progressively decreases as the La content increases, reaching as low as 10.1% for Ni/Ce_0.1La_0.9. This decrease explains a disruption of the CeO₂ fluorite lattice due to La incorporation, which may weaken the Ce–O bond and alter oxygen mobility.¹² In contrast, the content of surface oxygen species and hydroxyls (O⁻/O²⁻, CO₃²⁻/OH⁻, and OH⁻) increased significantly with La ratio. Among all the samples, Ni/Ce_0.7La_0.3 and Ni/Ce_0.5La_0.5 showed a higher ratio of surface oxygen species (O⁻/O²⁻) compared with those of the others, which is consistent with their enhanced CH₄ decomposition activity. These species are generally regarded as highly active and closely related to the presence of oxygen vacancies, which is also supported by the increased Ce³⁺ concentration observed in the Ce 3d spectra.^36,37 As the La content increases (x ≥ 0.5), the contributions of CO₃²⁻ and OH⁻ species also increase, consistent with the appearance of La-based crystalline phases in the XRD patterns of La₂O₂CO₃, La(OH)₃, and La₂NiO₄ (Fig. 2). The presence of such secondary phases suggests structural heterogeneity and phase segregation, which may lead to the accumulation of loosely bound carbonate or hydroxyl groups on the catalyst surface.^19,21,36 Consequently, excessive La incorporation disrupts the formation of a uniform mixed-oxide lattice and reduces the availability of catalytically active oxygen species. In such cases, even a high ratio of surface oxygen species does not necessarily enhance CH₄ conversion. Indeed, these results highlight that balanced La incorporation is crucial for maximizing active surface oxygen species while minimizing phase separation and the accumulation of inactive surface adsorbates.

While XPS provides valuable information on the surface redox state of cerium, its surface-sensitive nature limits insight into bulk defect structures and lattice disorder. Therefore, Raman spectroscopy was subsequently employed to probe vacancy-related lattice perturbations and structural modifications within the CeO₂-based framework. The Raman spectra of the reduced catalysts, as shown in Fig. 5(d), exhibit several characteristic bands that can be assigned to ceria- and lanthanum-related lattice vibrations. For Ce-containing samples, a pronounced Raman band centered at approximately 461 cm⁻¹ is observed, which corresponds to the F₂g symmetric stretching vibration of the Ce–O bond in the fluorite CeO₂ lattice. This mode is a signature of the well-ordered fluorite structure and is commonly used as a reference for evaluating structural disorder in ceria-based materials.²⁹ Moreover, samples dominated by the CeO₂ fluorite lattice exhibit a broad band near 561 cm⁻¹. This feature is generally attributed to defect-induced Raman modes associated with oxygen vacancies as well as local lattice distortions, which are generated by the partial reduction of Ce⁴⁺ to Ce³⁺ within the CeO₂ framework.²⁹ It should be recognized that the Ce³⁺/Ce⁴⁺ ratio obtained from XPS and the Raman defect-band intensity represent complementary descriptors of defect-related surface reduction and bulk lattice disorder, respectively, and that a strictly linear correlation between the two is not necessarily required. The relative contribution of this defect-related band was evaluated using the integrated area ratio with respect to the F₂g mode, providing an indirect descriptor of vacancy-related lattice disorder. Based on this analysis, the defect-band/F₂g area ratio increases sequentially in the order of Ni/Ce_1.0 (0.3739) < Ni/Ce_0.9La_0.1 (0.4718) < Ni/Ce_0.7La_0.3 (0.6476) < Ni/Ce_0.5La_0.5 (1.0842), indicating progressively enhanced defect-related lattice perturbation upon La incorporation into the CeO₂ lattice. Notably, although XPS Ce 3d analysis reveals the highest surface Ce³⁺ fraction for the Ni/Ce_0.7La_0.3 catalyst, the Raman-derived defect index reaches its maximum for Ni/Ce_0.5La_0.5. This discrepancy can be rationalized by considering that the Raman band at ∼561 cm⁻¹ reflects not only oxygen-vacancy formation but also lattice disorder and structural perturbation. At higher La contents, charge compensation induced by La³⁺ substitution leads to significant lattice distortion and partial disruption of the fluorite CeO₂ framework, thereby amplifying the defect-related Raman response even without a proportional increase in Ce³⁺ concentration.^5,20 Consequently, the elevated defect index observed for Ni/Ce_0.5La_0.5 is attributed to enhanced lattice disorder arising from La-induced structural modification rather than a simple increase in oxygen-vacancy concentration alone. For La-rich catalysts, additional Raman bands located at approximately 349 and 459 cm⁻¹ become prominent.⁴⁹ These bands are assigned to La–O lattice vibrations characteristic of the hexagonal La₂O₃ phase, indicating a gradual loss of fluorite-type CeO₂ dominance and the emergence of La-based oxide structures. In this compositional regime, the Raman response is primarily governed by structural transformation and phase heterogeneity rather than ceria-related oxygen-vacancy defects, rendering the conventional defect-band/F₂g analysis less applicable. Nevertheless, the emergence of Raman bands associated with La–O lattice vibrations indicates an increased contribution of La-based oxide structures, accompanied by a diminished relative contribution of ceria-related defect features.

To elucidate the electronic structure of Ni within the bulk phase of Ni/Ce_1−xLa_x catalysts, soft XAS was performed at the Ni L₃-edge in the total fluorescence yield mode (Fig. 6). Unlike XPS, which primarily probes the surface region, soft XAS enables the investigation of unoccupied electronic states and provides element-specific information about the local coordination and oxidation state of Ni in the bulk region. The main spectral feature corresponds to the Ni L₃-edge transition (2p_3/2 → 3d) at 852.6 eV.¹⁴ In this analysis, an additional peak was consistently observed at approximately ∼851 eV, which was attributed to the La M₄-edge, in agreement with previous studies on La-containing oxide systems.³⁸ This La M₄-edge feature gradually increased with increasing La content in the Ce–La supports, indicating the presence of La within the support lattice. However, the spectral overlap between the La M₄-edge and Ni L₃-edges complicates the direct interpretation of the Ni electronic state, particularly in samples with higher La content.³⁸ In order to quantify the oxidation state of the Ni species, linear combination fitting was employed using the reference spectra of Ni foil (Ni⁰), NiO (Ni²⁺), and Li(NiMnCo)O₂ (Ni³⁺). This approach enables the effective deconvolution of the Ni L₃-edge spectra despite interference from the La M₄-edge, allowing the relative concentration of Ni⁰, Ni²⁺, and Ni³⁺ species to be accurately determined in the series of catalysts.³⁸ The analysis reveals that the relative content of metallic Ni increased with the La content, reaching a maximum of 81% for the Ni/Ce_0.7La_0.3 catalyst. However, further increases in the La content led to higher contents of Ni²⁺ and Ni³⁺ species, which were attributed to the formation of perovskite-type structures that stabilize oxidized Ni within the lattice. These results align with the XRD and TEM findings, indicating that excessive La promotes the self-incorporation of Ni into the perovskite framework, thereby reducing metallic Ni on the catalyst surface.

Fig. 6 Ni L₃-edge soft X-ray absorption spectroscopy and linear combination fitting results for Ni/Ce_1−xLa_x catalysts.

CO₂-TPD is a well-established technique for determining the total basicity of solid catalysts (Fig. 7 and Table 3). All samples exhibit three distinct desorption regions, corresponding to weak (50–150 °C), medium (150–500 °C), and strong (>500 °C) basic sites.^39–41 Among the various types of basic sites, medium basic sites are especially noteworthy because of their association with surface oxygen species capable of facilitating CH₄ activation via H₂ abstraction and stabilizing active CH_x intermediates.⁴¹ Notably, catalysts with higher CH₄ decomposition activities are associated with pronounced contributions from medium basic sites.³⁹ This observation implies that adjusting the Ce/La ratio facilitates the formation of metal-oxygen pairs (Mⁿ⁺–O²⁻), which are characteristic of moderately basic sites. These sites serve as bifunctional centers capable of donating and accepting electrons during CH₄ activation. This enables C–H bond weakening and electron-density redistribution, collectively reducing the activation energy required for bond cleavage.^40,41 By comparison, samples with excessive La content (x ≥ 0.7) exhibited a marked increase in strong basic sites, which are typically associated with low-mobility O²⁻ species and La-rich domains (e.g., La₂O₃).³⁷ Excessively strong basic sites may increase activation energy barriers or impede the desorption of surface species, potentially resulting in surface poisoning or the accumulation of carbonaceous deposits, such as amorphous carbon.³⁹

Fig. 7 CO₂ temperature-programmed desorption profiles for the series of Ni/Ce_1−xLa_x catalysts.

Table 3 CO₂ uptake profiles and relative contributions of medium basic sites

Sample	Weak basic sites	Medium basic sites	Strong basic sites	Ratio of medium basic sites
	CO2 uptake [µmol g^−1]	CO2 uptake [µmol g^−1]	CO2 uptake [µmol g^−1]
Ni/Ce1.0	109.9	188.5	162.7	0.41
Ni/Ce0.9La0.1	34.7	526.1	395.4	0.55
Ni/Ce0.7La0.3	38.5	717.2	157.4	0.78
Ni/Ce0.5La0.5	36.9	221.7	68.6	0.67
Ni/Ce0.3La0.7	23.6	394.8	725.7	0.34
Ni/Ce0.1La0.9	14.0	315.7	1234.0	0.20
Ni/La1.0	38.6	162.9	1509.6	0.46

---

3.3 Catalytic performance of the Ni/Ce_1−x-La_x catalysts

Fig. 8(a) presents the carbon yield and CH₄ conversion rates of the prepared catalysts supported on Ce–La mixed oxides to investigate the effect of the Ce/La ratio under standard reaction conditions. The reaction was conducted at 600 °C for 1 h with a WHSV of 360 L g⁻¹ h⁻¹ using the TG analyzer. Among the catalysts, Ni/Ce_0.7La_0.3 exhibits the highest carbon yield of 811.4 wt%, which is nearly five times that of the Ni/Ce_1.0 catalyst. The activity order of the samples is as follows: Ni/Ce_0.7La_0.3 > Ni/Ce_0.5La_0.5 > Ni/Ce_0.9La_0.1 ≈ Ni/Ce_0.3La_0.7 ≈ Ni/Ce_0.1La_0.9 > Ni/Ce_1.0 ≈ Ni/La_1.0 (Fig. 8(a) and Table 1). The hydrogen production rates are not monotonic with Ce content, indicating that catalytic performance is governed by the Ce : La ratio rather than by the absolute Ce concentration alone. The superior performance of Ni/Ce_0.7La_0.3 is associated with its balanced reduction behavior and stable metal–support interaction. The balance between Ce and La is crucial because an excessive proportion of La in the support tends to induce a perovskite-like structure, which is less likely to participate directly in the catalytic reaction, as confirmed by XRD and H₂-TPR analyses. In addition, the CO₂-TPD results indicate that the dominance of medium-strength basic sites contributes to its high catalytic activity. Previous theoretical studies have shown that such medium-strength sites maintain more favorable C–H activation barriers (64–91 kJ mol⁻¹) compared to strong acid-base pairs, which increase the barrier to 94–126 kJ mol⁻¹.^40,41 Therefore, medium basic sites in the form of metal-oxygen pairs play a crucial role in the initial activation of CH₄ and are considered one of the key factors in lowering the barrier for C–H bond dissociation.

Fig. 8 Catalytic performance and kinetic analysis of Ni/Ce_1−xLa_x catalysts for methane decomposition: (a) carbon yield and methane conversion rate over Ni/Ce_1−xLa_x catalysts after methane decomposition at 600 °C; (b) carbon yield as a function of temperature for representative catalysts; (c) CH₄ decomposition rate profiles of each catalyst as a function of reaction time; (d) H₂ production rate profiles; (e) CO and CO₂ production profiles under identical reaction conditions; and (f) Arrhenius plots derived from kinetic data for representative catalysts with corresponding activation energy (E_a) values.

To further clarify the impact of reaction temperature, additional catalytic tests were conducted between 500 and 800 °C in a tubular CVD reactor (WHSV = 36 L g⁻¹ h⁻¹). As shown in Fig. 8(b), the highest activity is achieved at 600 °C. At temperatures above 700 °C, catalytic deactivation occurs, likely due to sintering of Ni active sites or coverage by amorphous carbon, both of which inhibit catalytic function. Based on these findings, 600 °C is selected as the optimal temperature for long-term testing.

Following the initial activity tests, long-term experiments were performed to evaluate the lifespan of the catalysts. Rapid carbon deposition on the catalyst surface is widely recognized as a major cause of catalyst deactivation during the catalytic decomposition of methane. Fig. 8(c and d) shows the long-term activity profiles of Ni/Ce_1−xLa_x catalysts at 600 °C, evaluated using a tubular CVD reactor. Most catalysts reach their peak activities within 45 min. The Ni/Ce_0.7La_0.3 and Ni/Ce_0.5La_0.5 catalysts exhibit remarkable catalytic stability, maintaining high H₂ production rates of 16.5 and 14.8 mmol g⁻¹ min⁻¹, respectively, for approximately 300 min without noticeable decline. Notably, these catalysts sustain H₂ production rates comparable to those observed during the initial 30 min of the reaction, demonstrating their excellent long-term stability under CH₄ decomposition conditions. Ni/Ce_0.9La_0.1 also demonstrates a high initial activity and an extended catalytic lifetime; however, its H₂ production rate gradually decreases after 140 min. In the case of Ni/Ce_0.3La_0.7, Ni/Ce_0.1La_0.9, and Ni/La_1.0 catalysts, despite achieving high initial H₂ production rates of 12.9, 12.1, and 11.3 mmol g⁻¹ min⁻¹, respectively, rapid activity decline is observed within 120 min of the reaction. This is associated with the weakened metal–support interactions and the reduced structural stability observed in the physicochemical characterization studies. In addition, the catalytic performance is also related to the Ni particle size. In general, smaller Ni particles exhibit higher initial activity due to their larger surface area and greater number of exposed active sites.^21,23 As shown in Table 1, when the La content in the catalyst exceeds 50%, the Ni particle size increases significantly. Consequently, exceeding the optimal particle size range reduces the exposure of active sites, which can lead to a decline in catalytic performance and durability.²⁴

The CO and CO₂ formation profiles are presented in Fig. 8(e). Although CO and CO₂ were simultaneously plotted for comparative understanding, their concentrations were consistently below approximately 0.02% throughout the reaction, indicating that the catalytic pathway was dominated by direct CH₄ → C + 2H₂ routes. Accordingly, the CH₄ decomposition rate profiles included in Fig. 8(c) validate the reliability of the performance comparison originally presented using hydrogen production in Fig. 8(d).

Fig. 8(f) presents the Arrhenius plots for the Ni/Ce_1.0, Ni/Ce_0.7La_0.3, and Ni/La_1.0 catalysts based on the CH₄ decomposition rates measured at different temperatures (Fig. 7(b)). Arrhenius plots provide further evidence for the role of the Ce/La ratio in modulating CH₄ activation. Ni/Ce_0.7La_0.3 exhibited the lowest apparent activation energy (62.9 kJ mol⁻¹ K⁻¹), indicating that partial La substitution enhances the ease of CH₄ dissociation, likely through synergistic modifications of surface basicity and oxygen mobility. Conversely, Ni/La_1.0 showed the highest activation energy (129.0 kJ mol⁻¹ K⁻¹), consistent with excessive lattice distortion and destabilized Ni–support interactions at high La contents. Ni/Ce_1.0 displayed an intermediate value of 114.1 kJ mol⁻¹ K⁻¹. Taken together, these results demonstrate that the catalytic performance trends directly reflect the structure–property relationships derived from Ce/La optimization, with partial La incorporation providing an optimal balance of lattice stability, oxygen vacancy formation, and Ni–support interactions.

It should be noted that the remarkable activity and durability of Ni/Ce_0.7La_0.3 can be attributed to its optimized physicochemical properties, including an increased Ce³⁺ fraction, balanced reducibility of Ni species without excessive incorporation into the perovskite-type lattice and the effect of the basic site. Their synergetic features promote the catalytic performance.

To more rigorously evaluate the long-term stability requirement, an extended methane decomposition experiment was conducted using the Ni/Ce_0.7La_0.3 catalyst beyond the initially evaluated 300 min (more than 600 min). As shown in Fig. S2, the catalyst maintains a high and stable H₂ production rate of approximately 16 mmol g⁻¹ min⁻¹ for nearly 5 h. After this period, a gradual decline in activity is observed, indicating the onset of catalyst deactivation under the fixed reaction conditions employed in this study. Although the reaction could not be sustained up to 10 h with preserved activity, the extended test clearly demonstrates the superior durability of Ni/Ce_0.7La_0.3 compared to other compositions investigated in this work. It should be emphasized that all experiments were conducted under identical reaction parameters to ensure a fair comparison, without adjusting gas composition, pressure, or space velocity. While deactivation becomes evident when the reaction time is extended past 5 h under these conditions, further optimization of operating parameters is expected to extend the catalyst lifetime. This aspect will be a primary issue for the practical application of the catalyst.

3.4 Characterization of carbon materials

To directly elucidate the structural nature of the carbon materials formed during methane decomposition, high-resolution transmission electron microscopy (HRTEM) analysis was performed prior to further morphological and spectroscopic characterization. As shown in Fig. 9(a), the carbon product exhibits a well-defined hollow tubular morphology composed of multiple concentric graphitic layers, which is a characteristic feature of multi-walled carbon nanotubes (MWCNTs). While the formation of concentric graphitic layers confirms the presence of CNT structures, the HRTEM analysis further reveals that La-rich compositions (La ≥ 0.7) tend to exhibit additional amorphous carbon domains adjacent to the tubular carbon. This suggests that excessive La incorporation promotes less ordered carbon deposition, leading to the coexistence of CNTs and amorphous carbon species. This direct structural evidence confirms that CNTs constitute the dominant carbon nanostructure formed under the present reaction conditions. Based on this confirmation, complementary analyses including SEM, diameter distribution, Raman spectroscopy, C K-edge XAS, and TG analysis were subsequently employed to further evaluate the morphology, graphitization degree, and thermal stability of the CNT products.

Fig. 9 (a) Transmission electron microscopy and (b) scanning electron microscopy images and (c) corresponding CNT diameter distribution graphs for catalysts after CH₄ decomposition.

The physicochemical properties of the carbon materials formed on the catalyst surface depend significantly on the catalyst characteristics. Fig. 9(b and c) shows the FE-SEM images of the carbon materials produced on different Ce_1−xLa_x-supported catalysts after CH₄ decomposition, along with the corresponding CNT diameter distributions. All catalysts led to the formation of CNTs, but the morphology and diameter distribution varied significantly depending on the catalyst composition. Ni/Ce_1.0 and Ni/Ce_0.9La_0.1 generated relatively thin and tangled CNTs with moderate uniformity. The Ni/Ce_0.7La_0.3 catalyst exhibited the narrowest diameter distribution and smallest average diameter (24.5 nm), indicating more controlled CNT nucleation and growth. Ni/Ce_0.5La_0.5 formed slightly thicker carbon nanomaterials than Ni/Ce_0.7La_0.3 but maintained a relatively well-aligned structure with an average diameter of 34.6 nm. However, catalysts from Ni/Ce_0.3La_0.7 to Ni/La_1.0 produced thicker and more irregular CNTs, with average diameters ranging from 41.5 to 48.3 nm, accompanied by disordered structures and particle encapsulation. Strong CH₄ adsorption can lead to the excessive accumulation of carbonaceous species on the surface, which may hinder proper CNT nucleation and promote the growth of amorphous carbon over filamentous carbon structures.^13,24,42 These morphological trends indicate that excessively strong CH₄ adsorption can lead to undesirable carbon accumulation, whereas a balanced adsorption-activation behavior is essential for the growth of high-quality CNTs.

TG analysis is useful for both qualitative and quantitative investigation of carbon products formed on catalysts. A mixture of different carbon species can result in a stepwise weight loss curve during TG analysis. Moreover, the temperature at which weight loss occurs can help identify the types of carbon materials present, such as amorphous carbon, single-walled CNTs, and multi-walled CNTs.^43,44 As shown in Fig. 10, all catalysts exhibited remarkable weight loss in the range of 500–700 °C, corresponding to the oxidation of carbon products formed during CH₄ decomposition. Most samples showed a single sharp oxidation step, which is indicative of more homogeneous and structured carbon, likely consisting predominantly of well-formed CNTs with minimal amorphous content. Interestingly, the Ni/Ce_0.1La_0.9 and Ni/La_1.0 catalysts displayed a two-step weight loss pattern, suggesting the presence of carbon species with different structural characteristics. The initial weight loss at lower temperatures was attributed to the oxidation of amorphous or poorly ordered carbon, which is less thermally stable.⁴⁴ The subsequent gradual weight loss at higher temperatures corresponds to the oxidation of more graphitized, crystalline CNT structures, which resist oxidation owing to their higher structural order.⁴³ This stepwise oxidation behavior implies that amorphous carbon and CNTs coexist on these catalysts.

Fig. 10 Thermogravimetric (TG-DTG) analysis of Ni/Ce_1−xLa_x catalysts under an air atmosphere.

Following thermal analysis, Raman spectroscopy and carbon K-edge soft XAS were performed, to evaluate the structural ordering and chemical purity of the CNTs formed on the catalyst surface. As shown in Fig. 11(a) and summarized in Table 4, the I_D/I_G ratios obtained from the Raman spectra ranged from 0.79 and 1.10. The La-based catalysts, such as Ni/Ce_0.3La_0.7, Ni/Ce_0.1La_0.9, and Ni/La_1.0, showed relatively low I_D/I_G ratios (0.93–1.00), indicating a high degree of graphitization.^45,46 In contrast, Ni/Ce_0.7La_0.3 exhibited a slightly higher I_D/I_G ratio of 0.98, despite producing the most uniform and well-ordered CNTs, according to SEM (Fig. 10(b)).

Fig. 11 (a) Raman and (b) C K-edge soft X-ray absorption spectra of carbon materials formed on Ni/Ce_1−xLa_x catalysts.

Table 4 I _D/I_G ratios and oxidation temperatures of carbon materials on Ni/Ce_1−xLa_x catalysts

Sample	I D/IG ratio of CNTs	Oxidation temperature [°C]
Ni/Ce1.0	0.92	612
Ni/Ce0.9La0.1	1.10	605
Ni/Ce0.7La0.3	0.98	655
Ni/Ce0.5La0.5	1.00	644
Ni/Ce0.3La0.7	0.79	637
Ni/Ce0.1La0.9	0.93	336 and 632
Ni/La1.0	0.93	350, 606, and 717

---

To provide a more detailed understanding of the electronic structure and bonding characteristics of the carbon materials generated during the reaction, carbon K-edge near-edge X-ray absorption fine structure spectroscopy was performed. The resulting spectra are displayed in Fig. 11(b). All samples exhibited clear resonances associated with π* and σ* transitions, offering valuable insights into the hybridization states within the carbon framework. The π* resonance at 285 eV, indicative of transitions to unoccupied π* orbitals of C=C double bonds, is characteristic of the sp²-hybridized carbon structures typical of CNTs.^47,48 Meanwhile, the σ* resonance was observed in the approximate range of 291–292 eV, corresponding to transitions to unoccupied σ* states of C–C single bonds, reflecting sp³-hybridization within the carbon matrix.^47,48 In samples with higher Ce content, the π* peak appeared relatively broad and of lower intensity, suggesting greater presence of defects and amorphous carbon species.^47,48 In contrast, increasing the La content in the catalyst resulted in a sharper and more intense π* resonance, indicating the development of well-ordered sp²-hybridized carbon networks and high structural alignment within the nanotube walls.⁴⁷ Moreover, with increasing La content, a distinct peak at 288 eV associated with C–H related resonances became evident, implying the presence of surface-bound functional groups likely formed during the CNT growth process.⁴⁷ The concurrent presence of a pronounced π* resonance and C–H feature at 288 eV indicates that the CNTs preserve their high crystallinity and graphitic purity while also exhibiting a certain level of surface functionalization.

Overall, the above analyses suggest that a low I_D/I_G ratio alone does not ensure the purity of the CNTs, as it can be affected by the presence of graphitized but hydrogenated carbon species.

4. Conclusion

In this study, a series of Ni/Ce_1−xLa_x catalysts was synthesized and evaluated for catalytic methane decomposition to simultaneously produce hydrogen and CNTs. XRD and XPS analyses confirm that the partial substitution of La³⁺ into the CeO₂ lattice induced lattice distortion and electronic coupling without phase segregation at appropriate La contents. In contrast, excessive La incorporation (x ≥ 0.5) yields La-based perovskite-type phases, such as LaNiO₃ and La₂NiO₄, as well as La₂O₂CO₃ and La(OH)₃, which significantly alter the catalyst structure. These secondary phases immobilize Ni within the lattice as Ni³⁺ species, thereby suppressing its reduction to active Ni⁰. The shift in reducibility and appearance of strong basic sites are associated with poor CH₄ activation and rapid deactivation. This structural transition from the fluorite to perovskite phases is a key factor contributing to the decreased activity and carbon yield of the La-based catalysts.

The performance of CH₄ decomposition highly depends on the Ce/La ratio of the support. Ni/Ce_0.7La_0.3 exhibited the highest carbon yield (811.4 wt%) and sustained hydrogen production at 600 °C over extended reaction times, reflecting its favorable reduction behavior and optimized Ni–support interactions. However, catalysts with excessive La content show weakened reducibility and poor long-term stability, consistent with the structural instability and strong basic sites inferred from physicochemical characterization. The apparent activation energies confirm this trend, with Ni/Ce_0.7La_0.3 showing the lowest barrier for CH₄ activation, whereas Ni/La_1.0 required substantially higher energy input.

Carbon characterization clarifies the structure–performance relationship. Ni/Ce_1.0, Ni/Ce_0.9La_0.1, Ni/Ce_0.7La_0.3, and Ni/Ce_0.5La_0.5 catalysts facilitated the growth of uniform CNTs with narrow diameter distributions (average 24.5–39.1 nm), whereas catalysts with excessive La content predominantly produced amorphous and disordered carbon structures. Although some La-based catalysts exhibited low I_D/I_G ratios, soft XAS analysis confirmed the presence of hydrogenated carbon species, suggesting that a low I_D/I_G ratio does not necessarily reflect high-purity CNTs.

These results emphasize the critical role of precisely controlling La incorporation to suppress unfavorable phase transitions while retaining the structural and electronic features necessary for efficient methane activation and CNT growth. More importantly, this work advances the understanding of structure-property-performance relationships in Ce–La mixed oxides and presents a rational design strategy for developing high-performance Ni-based catalysts, thereby enabling efficient and CO₂-free co-production of hydrogen and carbon nanomaterials. Such carbon nanostructures are also highly promising for applications in nanoscale devices and lithium-ion batteries, further broadening the impact of this work.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Acknowledgements

This work was conducted under framework of the research and development program of the Korea Institute of Energy Research (C5-2402). This research was also supported by the National Research Council of Science & Technology(NST) grant by the Korea government (MSIT) (No. GTL25101-000).

References

1. D. Leybo, U. J. Etim, M. Monai, S. R. Bare, Z. Zhong and C. Vogt, Metal-support interactions in metal oxide-supported atomic, cluster, and nanoparticle catalysis, Chem. Soc. Rev., 2024, 53(21), 10450–10490,  10.1039/d4cs00527a.
2. J. L. Snider, V. Streibel, M. A. Hubert, T. S. Choksi, E. Valle and D. C. Upham, et al., Revealing the Synergy between Oxide and Alloy Phases on the Performance of Bimetallic In-Pd Catalysts for CO₂ Hydrogenation to Methanol, ACS Catal., 2019, 9(4), 3399–3412,  DOI:10.1021/acscatal.8b04848.
3. M. Grabchenko, G. Pantaleo, F. Puleo, T. S. Kharlamova, V. I. Zaikovskii and O. Vodyankina, et al., Design of Ni-based catalysts supported over binary La-Ce oxides: Influence of La/Ce ratio on the catalytic performances in DRM, Catal. Today, 2021, 382, 71–81,  DOI:10.1016/j.cattod.2021.07.012.
4. K. Sutthiumporn and S. Kawi, Promotional effect of alkaline earth over Ni-La₂O₃ catalyst for CO₂ reforming of CH₄: Role of surface oxygen species on H₂ production and carbon suppression, Int. J. Hydrogen Energy, 2011, 36(22), 14435–14446,  DOI:10.1016/j.ijhydene.2011.08.022.
5. A. Gupta, U. V. Waghmare and M. S. Hegde, Correlation of oxygen storage capacity and structural distortion in transition-metal-, noble-metal-, and rare-earth-ion-substituted CeO₂ from first principles calculation, Chem. Mater., 2010, 22(18), 5184–5198,  DOI:10.1021/cm101145d.
6. M. J. Kim, J. R. Youn, H. J. Kim, M. W. Seo, D. Lee and K. S. Go, et al., Effect of surface properties controlled by Ce addition on CO₂ methanation over Ni/Ce/Al₂O₃ catalyst, Int. J. Hydrogen Energy, 2020, 45(46), 24595–24603,  DOI:10.1016/j.ijhydene.2020.06.144.
7. M. Greluk, M. Rotko and S. Turczyniak-Surdacka, Enhanced catalytic performance of La₂O₃ promoted Co/CeO₂ and Ni/CeO₂ catalysts for effective hydrogen production by ethanol steam reforming: La₂O₃ promoted Co(Ni)/CeO₂ catalysts in SRE, Renew. Energy, 2020, 155, 378–395,  DOI:10.1016/j.renene.2020.03.117.
8. I. Luisetto, S. Tuti, C. Romano, M. Boaro, E. Di Bartolomeo and J. K. Kesavan, Dry reforming of methane over Ni supported on doped CeO₂ : New insight on the role of dopants for CO₂ activation, J. CO2 Util., 2019, 30, 63–78,  DOI:10.1016/j.jcou.2019.01.006.
9. P. Vinchhi, A. Ray, K. Mallik and R. Pati, Gd-doped ceria with extraordinary oxygen-ion conductivity for low temperature solid oxide fuel cells, Sci. Rep., 2024, 14(1), 1–13,  DOI:10.1038/s41598-024-59030-6.
10. X. Hou, Q. Lu and X. Wang, Enhanced catalytic properties of La-doped CeO₂ nanopowders synthesized by hydrolyzing and oxidizing Ce₄₆La₅C₄₉ alloys, J. Sci. Adv. Mater. Devices, 2017, 2(1), 41–44,  DOI:10.1016/j.jsamd.2017.02.006.
11. Y. Xu, L. Gao, Q. Hou, P. Wu, Y. Zhou and Z. Ding, Enhanced Oxygen Storage Capacity of Porous CeO₂ by Rare Earth Doping, Molecules, 2023, 28(16), 6005,  DOI:10.3390/molecules28166005.
12. R. Zhou, M. Mohamedali, Y. Ren, Q. Lu and N. Mahinpey, La-Ce binary oxide catalysts for low-temperature dry reforming, Int. J. Hydrogen Energy, 2023, 48(89), 34766–34782,  DOI:10.1016/j.ijhydene.2023.05.281.
13. S. Takenaka, S. Kobayashi, H. Ogihara and K. Otsuka, Ni/SiO₂ catalyst effective for methane decomposition into hydrogen and carbon nanofiber, J. Catal., 2003, 217(1), 79–87,  DOI:10.1016/S0021-9517(02)00185-9.
14. J. R. Youn, M. J. Kim, K. C. Kim, M. Kim, T. Jung and K. S. Go, et al., Highly efficient Co-added Ni/CeO₂ catalyst for co-production of hydrogen and carbon nanotubes by methane decomposition, Fuel Process. Technol., 2024, 263, 108130,  DOI:10.1016/j.fuproc.2024.108130.
15. A. Rastegarpanah, F. Meshkani and M. Rezaei, CO_x-free hydrogen and carbon nanofibers production by thermocatalytic decomposition of methane over mesoporous MgO·Al₂O₃ nanopowder-supported nickel catalysts, Fuel Process. Technol., 2017, 167, 250–262,  DOI:10.1016/j.fuproc.2017.07.010.
16. S. Afzal, A. Prakash, P. Littlewood, H. Choudhury, Z. K. Ghouri and S. Mansour, et al., Catalyst Deactivation by Carbon Deposition: The Remarkable Case of Nickel Confined by Atomic Layer Deposition, ChemCatChem, 2021, 13(13), 2988–3000,  DOI:10.1002/cctc.202100109.
17. Y. Xu, J. Qiao, W. Sun, Z. Wang and K. Sun, Enhancement of CO₂ activation and coke-resistant ability on Ni/CeO₂ catalyst with La doping for dry reforming of methane, Int. J. Hydrogen Energy, 2024, 88, 1451–1462,  DOI:10.1016/j.ijhydene.2024.07.027.
18. M. F. Bekheet, P. D. K. Nezhad, N. Bonmassar, L. Schlicker, A. Gili and S. Praetz, et al., Steering the Methane Dry Reforming Reactivity of Ni/La₂O₃ Catalysts by Controlled in Situ Decomposition of Doped La₂NiO₄ Precursor Structures, ACS Catal., 2021, 11(1), 43–59,  DOI:10.1021/acscatal.0c04290.
19. A. Bhaskaran, S. A. Singh, P. D. Costa and S. Roy, Unveiling the catalytic behaviour of LaNiO₃ and La₂NiO₄ for dry reforming of methane, Int. J. Hydrogen Energy, 2024, 68, 623–634,  DOI:10.1016/j.ijhydene.2024.04.295.
20. R. Pereñíguez, V. M. González-DelaCruz, J. P. Holgado and A. Caballero, Synthesis and characterization of a LaNiO₃ perovskite as precursor for methane reforming reactions catalysts, Appl. Catal., B, 2010, 93(3–4), 346–353,  DOI:10.1016/j.apcatb.2009.09.040.
21. B. M. Faroldi, J. F. Múnera and L. M. Cornaglia, In situ characterization of phase transformation and reactivity of high surface area lanthanum-based Ru catalysts for the combined reforming of methane, Appl. Catal., B, 2014, 150–151, 126–137,  DOI:10.1016/j.apcatb.2013.12.005.
22. S. Takenaka, E. Kato, Y. Tomikubo and K. Otsuka, Structural change of Ni species during the methane decomposition and the subsequent gasification of deposited carbon with CO₂ over supported Ni catalysts, J. Catal., 2003, 219(1), 176–185,  DOI:10.1016/S0021-9517(03)00152-0.
23. P. G. Lustemberg, Z. Mao, A. Salcedo, B. Irigoyen, M. V. G. Pirovano and C. T. Campbell, Nature of the Active Sites on Ni/CeO₂ Catalysts for Methane Conversions, ACS Catal., 2021, 11(16), 10604–10613,  DOI:10.1021/acscatal.1c02154.
24. W. Liang, H. Yan, C. Chen, D. Lin, K. Tan and X. Feng, et al., Revealing the effect of nickel particle size on carbon formation type in the methane decomposition reaction, Catalysts, 2020, 10(8), 1–20,  DOI:10.3390/catal10080890.
25. S. Chen, L. Higgins, I. Giarnieri, P. Benito and A. M. Beale, A detailed characterization study of Ni/CeO₂ catalysts identifies Ni availability as the primary factor affecting CO₂ methanation performance, J. Catal., 2024, 439, 115778,  DOI:10.1016/j.jcat.2024.115778.
26. J. Lucas, N. S. P. Naveen, M. J. Janik, K. Alexopoulos, G. Noh and D. Aireddy, et al., Improved Selectivity and Stability in Methane Dry Reforming by Atomic Layer Deposition on Ni-CeO₂-ZrO₂/Al₂O₃ Catalysts, ACS Catal., 2024, 14(12), 9115–9133,  DOI:10.1021/acscatal.4c02019.
27. M. Greluk, M. Rotko and S. Turczyniak-Surdacka, Enhanced catalytic performance of La₂O₃ promoted Co/CeO₂ and Ni/CeO₂ catalysts for effective hydrogen production by ethanol steam reforming: La₂O₃ promoted Co(Ni)/CeO₂ catalysts in SRE, Renew. Energy, 2020, 155, 378–395,  DOI:10.1016/j.renene.2020.03.117.
28. M. Y. Dogan, H. Arbag, H. M. Tasdemir, N. Yasyerli and S. Yasyerli, Effect of ceria content in Ni–Ce–Al catalyst on catalytic performance and carbon/coke formation in dry reforming of CH₄, Int. J. Hydrogen Energy, 2023, 48(60), 23013–23030,  DOI:10.1016/j.ijhydene.2023.04.011.
29. B. Choudhury and A. Choudhury, Ce³⁺ and oxygen vacancy mediated tuning of structural and optical properties of CeO₂ nanoparticles, Mater. Chem. Phys., 2012, 131(3), 666–671,  DOI:10.1016/j.matchemphys.2011.10.032.
30. W. Ma, T. Mashimo, S. Tamura, M. Tokuda, S. Yoda and M. Tsushida, et al., Cerium oxide (CeO_2-x) nanoparticles with high Ce³⁺ proportion synthesized by pulsed plasma in liquid, Ceram. Int., 2020, 46(17), 26502–26510,  DOI:10.1016/j.ceramint.2020.07.093.
31. S. G. Peera and S. W. Kim, Rare Earth Ce/CeO₂ Electrocatalysts: Role of High Electronic Spin State of Ce and Ce³⁺/Ce⁴⁺ Redox Couple on Oxygen Reduction Reaction, Nanomaterials, 2025, 15(8), 600,  DOI:10.3390/nano15080600.
32. P. R. L. Keating, D. O. Scanlon and G. W. Watson, The nature of oxygen states on the surfaces of CeO₂ and La-doped CeO₂, Chem. Phys. Lett., 2014, 608, 239–243,  DOI:10.1016/j.cplett.2014.05.094.
33. N. Shehata, K. Meehan, M. Hudait and N. Jain, Control of oxygen vacancies and Ce⁺³ concentrations in doped ceria nanoparticles via the selection of lanthanide element, J. Nanopart. Res., 2012, 14(10), 1–10,  DOI:10.1007/s11051-012-1173-1.
34. J. Zhang, H. Wong, D. Yu, K. Kakushima and H. Iwai, X-ray photoelectron spectroscopy study of high-k CeO₂/La₂O₃ stacked dielectrics, AIP Adv., 2014, 4(11), 117117,  DOI:10.1063/1.4902017.
35. N. Shehata, K. Meehan, M. Hudait and N. Jain, Control of oxygen vacancies and Ce⁺³ concentrations in doped ceria nanoparticles via the selection of lanthanide element, J. Nanopart. Res., 2012, 14(10), 1–10,  DOI:10.1007/s11051-012-1173-1.
36. Y. Yu, Y. M. Gan, C. Huang, Z. H. Lu, X. Wang and R. Zhang, et al., Ni/La₂O₃ and Ni/MgO–La₂O₃ catalysts for the decomposition of NH₃ into hydrogen, Int. J. Hydrogen Energy, 2020, 45(33), 16528–16539,  DOI:10.1016/j.ijhydene.2020.04.127.
37. K. Sutthiumporn and S. Kawi, Promotional effect of alkaline earth over Ni-La₂O₃ catalyst for CO₂ reforming of CH₄: Role of surface oxygen species on H₂ production and carbon suppression, Int. J. Hydrogen Energy, 2011, 36(22), 14435–14446,  DOI:10.1016/j.ijhydene.2011.08.022.
38. M. Hepting, D. Li, C. J. Jia, H. Lu, E. Paris and Y. Tseng, et al., Electronic structure of the parent compound of superconducting infinite-layer nickelates, Nat. Mater., 2020, 19(4), 381–385,  DOI:10.1038/s41563-019-0585-z.
39. A. Figueiredo, M. A. Soria, L. M. Madeira and C. Rocha, Preparation and Characterization of Materials for Low- to Intermediate-Temperature CO₂ Adsorption, Processes, 2024, 12(11), 2403,  DOI:10.3390/pr12112403.
40. A. Misol, I. Giarnieri, F. Ospitali, A. Ballarini, J. Jiménez-Jiménez and E. Rodríguez-Castellón, et al., CO₂ hydrogenation over Ru hydrotalcite-derived catalysts, Catal. Today, 2024, 425, 3–11,  DOI:10.1016/j.cattod.2023.114362.
41. Y. R. Dias and O. W. Perez-Lopez, CO₂ methanation over Ni-Al LDH-derived catalyst with variable Ni/Al ratio, J. CO2 Util., 2023, 68, 102381,  DOI:10.1016/j.jcou.2022.102381.
42. L. Dong, Y. Du, J. Li, H. Wang, Y. Yang and S. Li, et al., The effect of CH₄ decomposition temperature on the property of deposited carbon over Ni/SiO₂ catalyst, Int. J. Hydrogen Energy, 2015, 40(31), 9670–9676,  DOI:10.1016/j.ijhydene.2015.06.005.
43. E. Mansfield, A. Kar, C. M. Wang and A. N. Chiaramonti, Statistical sampling of carbon nanotube populations by thermogravimetric analysis, Anal. Bioanal. Chem., 2013, 405(25), 8207–8213,  DOI:10.1007/s00216-013-7221-6.
44. A. Buzarovska, V. Stefov, M. Najdoski and G. Bogoeva-Gaceva, Thermal analysis of multi-walled carbon nanotubes material obtained by catalytic pyrolysis of polyethylene, Maced. J. Chem. Chem. Eng., 2015, 34(2), 373–379,  DOI:10.20450/mjcce.2015.620.
45. B. J. Kim, J. P. Kim and J. S. Park, Effects of Al interlayer coating and thermal treatment on electron emission characteristics of carbon nanotubes deposited by electrophoretic method, Nanoscale Res. Lett., 2014, 9(1), 1–6,  DOI:10.1186/1556-276X-9-236.
46. C. Zhou, R. Senegor, Z. Baron, Y. Chen, S. Raju and A. A. Vyas, et al., Synthesis and interface characterization of CNTs on graphene, Nanotechnology, 2016, 28(5) DOI:10.1088/1361-6528/28/5/054007.
47. N. H. Tran, M. A. Wilson, A. S. Milev, J. R. Bartlett, R. N. Lamb and D. Martin, et al., Photoemission and absorption spectroscopy of carbon nanotube interfacial interaction, Adv. Colloid Interface Sci., 2009, 145(1–2), 23–41,  DOI:10.1016/j.cis.2008.07.006.
48. V. Shmatko, D. Leontyeva, N. Nevzorova, N. Smirnova, M. Brzhezinskaya and G. Yalovega, Interaction between NiO_x and MWCNT in NiO_x/MWCNTs composite: XANES and XPS study, J. Electron Spectrosc. Relat. Phenom., 2017, 220, 76–80,  DOI:10.1016/j.elspec.2017.03.016.
49. J. Xiong, H. Yu, Y. Wei, C. Xie, K. Lai and Z. Zhao, et al., Metal Ions (Li, Mg, Zn, Ce) Doped into La₂O₃ Nanorod for Boosting Catalytic Oxidative Coupling of Methane, Catalysts, 2022, 12, 713,  DOI:10.3390/catal12070713.

---