Asymmetric substrate supported Ni catalysts for robust photothermal catalytic dry reforming of methane

Abstract

Photothermal catalytic dry reforming of methane with CO₂ has emerged as a promising yet nascent strategy for mitigating greenhouse gas emissions and enabling clean energy conversion. However, achieving optimal performance requires advances in both catalyst design and mechanistic understanding. Herein, we adopted a double-emulsion-guided micelle assembly strategy to synthesize asymmetric supports (AMONs and AMOMs), featuring unidirectional open/closed pore channels. This distinctive architecture enabled the formation of an asymmetric catalyst configuration through ethylene glycol-assisted selective confinement of Ni nanoparticles at the open-pore termini. Compared to conventional symmetric catalysts, the optimized 5% Ni AMONs EG and 5% Ni AMOMs EG exhibited higher specific surface areas and improved metal dispersion, resulting in an abundance of active sites. Moreover, the asymmetric design strengthened the built-in electric fields, directing more photogenerated hot carriers and localized thermal energy toward reactant activation. Consequently, 5% Ni AMOMs EG achieved a remarkable H₂ production rate of 2314.2 mmol g⁻¹ h⁻¹ and sustained H₂ yields over 50 hours, outperforming symmetric counterparts and even some reported noble metal-based catalysts. This work offers a smart photothermal catalyst candidate and elucidates its structure–performance relationship, advancing photothermal catalytic technology for solar fuel production.

---

Jinqiang Zhang

Dr Jinqiang Zhang is currently a Lecturer of Chemistry and ARC DECRA Fellow at the School of Molecular Sciences, The University of Western Australia. He is also an adjunct Lecturer at the School of Chemical Engineering, The University of Adelaide. He received his PhD from Edith Cowan University and completed postdoctoral research at The University of Adelaide. His research focuses on nanostructured catalysts for solar-driven chemical transformations, including green energy production, C1 conversion and environmental remediation. Dr Zhang is particularly interested in photothermal and photocatalytic mechanisms, aiming to enhance solar-to-chemical energy conversion through advanced material design. He leads a DECRA project and a Discovery Project and is committed to advancing sustainable energy solutions through collaborative, interdisciplinary research.

---

Introduction

Natural gas, with its high energy-to-mass ratio, serves as an efficient fuel for electricity generation, heating, and transportation. Its abundant global reserves ensure a relatively stable supply for the foreseeable future.¹ However, methane (CH₄), the primary component of natural gas, is also a potent greenhouse gas, approximately 25 times more effective at trapping heat than carbon dioxide (CO₂). Methane leaks during extraction, transportation, and storage, along with excessive emission of CO₂, have led to severe environmental issues, including global warming,^2,3 significantly accelerating climate change. In the context of global efforts to achieve net-zero carbon emissions, the efficient conversion of CH₄ and CO₂ into valuable fuels has become a critical approach for mitigating greenhouse gas emissions and promoting resource recycling.⁴

Dry reforming of methane (DRM) with CO₂ is a pivotal reaction that converts the two major greenhouse gases into high-value syngas (H₂/CO). However, the reaction (CH₄ + CO₂ → 2H₂ + 2CO, ΔH = +247 kJ mol⁻¹) is highly endothermic, typically requiring temperatures above 700 °C. Such harsh conditions pose significant challenges for conventional thermocatalytic processes, including high energy input, catalyst sintering, and deactivation caused by carbon deposition.^5,6 Alternatively, photothermal catalytic technology offers a promising alternative by synergistically combining photoexcited hot carriers and localized thermal energy within a single system. This approach can lower the energy barrier, enabling DRM under milder conditions.^7,8 Additionally, the involvement of hot carriers may alter the reaction pathway, improving catalytic efficiency and long-term stability. Nevertheless, continued catalyst innovation remains crucial for advancing this emerging photothermal catalytic technology and enabling the practical, clean, and sustainable production of solar-derived syngas.

Symmetric structures, such as spherical or cubic nanoparticles, are commonly used as supports for metal sites in thermocatalytic DRM (e.g., Ni/SiO₂ particles⁹ and Ni@SiO₂ core–shell catalyst¹⁰) owing to their uniform surface energy distribution and electronic structures, which facilitate scalable synthesis. However, these structures present significant drawbacks when applied to solar-driven catalysis. In particular, weak metal–support interactions in symmetric geometries always lead to metal particle agglomeration, reducing dispersion and limiting active site availability. Furthermore, symmetric structures struggle to direct localized photothermal fields, limiting the coupling efficiency between light absorption and heat conduction. The uniformity of their electronic structures also yields weak internal electric fields, slowing charge dynamics and diminishing redox capabilities.^11,12 These limitations have driven researchers to explore novel catalyst designs that break the constraints of symmetry.

By contrast, asymmetric-structured catalysts demonstrate remarkable advantages in solar driven catalysis due to their unique geometric and electronic properties. Unlike conventional symmetric materials, asymmetric designs disrupt spatial symmetry to enhance catalytic performance. For instance, bowl-shaped materials and Janus heterojunctions (e.g., Fe₃O₄@mC&mSiO₂ (ref. 13)) provide larger specific surface areas and abundant surface defects, improving the dispersion of active components while exposing more catalytic sites. Besides, asymmetric interfaces induce built-in electric fields, enhancing light absorption and facilitating efficient hot carrier separation and transport. Furthermore, polarization effects in asymmetric structures regulate the adsorption configurations of reaction intermediates, suppressing the production of byproducts, thus increasing the quality of targeted products.⁵ Despite their potential in enhancing solar energy harvesting and catalytic kinetics, asymmetric-structured catalysts remain underexplored in photothermal catalysis, particularly in the DRM process.

In this work, we employed a double-emulsion-guided micelle assembly strategy to synthesize morphologically asymmetric supports with distinct open and closed mesoporous termini (AMONs). Using an ethylene glycol-assisted impregnation method, we effectively dispersed and spatially confined metallic components within the open-ended pore channels (5% Ni AMONs EG). Furthermore, the catalyst architecture was systematically optimized, evolving from a simple two-dimensional asymmetric nanosheet configuration to a more complex three-dimensional structure, while preserving the fundamental asymmetric pore design (AMOMs). This structural engineering enabled a synergistic integration of enhanced active site accessibility and nanoconfinement effects, improving mass transfer. Additionally, the asymmetric configuration amplified built-in electric fields, promoting the participation of more photogenerated hot carriers and localised thermal energy in the DRM process. As a result, the unique structures achieved H₂ production rates of 2335.4 and 2314.2 mmol g⁻¹ h⁻¹ for 5% Ni AMONs EG and 5% Ni AMOMs EG, respectively, with stable performance over 50 hours in photothermal catalytic DRM. These findings open a new avenue for photothermal catalyst development, advancing the practical production of solar fuels.

Results and discussion

Catalyst synthesis and characterization

A dual-emulsion-guided micelle assembly strategy was adopted to synthesize two dimensional (2D) asymmetric monolayer mesoporous nanosheets (AMONs) and 3D asymmetric mesoporous microspheres (AMOMs) (Fig. 1).¹⁴ This approach utilized a silica precursor, specifically 1,2-bis(triethoxysilyl)ethane (BTSE), and a surfactant, cetyltrimethylammonium bromide (CTAB) within a water/oil/water (W/O/W) dual-emulsion system. During the synthesis, CTAB served multiple roles: forming micellar templates, stabilizing charges to prevent aggregation, and facilitating the formation of a dynamic cross-linked network. Simultaneously, 1,3,5-trimethylbenzene (TMB) acted as a micelle-swelling agent to regulate pore size and as an emulsion interface modulator to control the dimensionality of the assembly (2D vs. 3D). The ordered arrangement of micelles at the oil–water interface was driven by interfacial self-assembly within the dual-emulsion droplets, guiding the formation of mesoporous structures.¹⁵ Therefore, the synthesis mechanism involved micelle self-assembly and directional alignment at interfaces, followed by silica precursor hydrolysis/condensation and surfactant removal, ultimately yielding mesoporous materials.

Fig. 1 Schematic illustration of catalyst synthesis via a dual-emulsion-directed micelle assembly strategy for preparing 2D AMONs and 3D AMOMs with Ni loading.

Within these asymmetric porous nanosheets, numerous quasi-spherical semi-open mesopores (∼5 nm in diameter) were uniformly arranged on a single plane, imparting surface anisotropy to the nanosheets (several micrometers in size).¹⁴ Triangular gaps between adjacent mesopores facilitated interconnections between nanosheets, resulting in ultrahigh specific surface areas (Fig. 1). AMONs featured a monolayer nanosheet structure with nanoscale thickness, exhibiting semi-open channels on the surface and an asymmetric morphology, one side with open pores and the other with closed or partially closed structures. By contrast, AMOMs adopted a collapsed microsphere morphology with bowl-like shapes and micrometer-scale diameters, similarly decorated with ordered semi-open channels and asymmetric surface features.

Nickel-loaded catalysts, 5% Ni AMOMs EG and 5% Ni AMONs EG, were prepared by introducing metal active sites into the mesopores on the open side via an ethylene glycol (EG) assisted impregnation method.¹⁶ Specifically, nickel salts (nickel nitrate hexahydrate) were dissolved in an EG–water mixture to form a homogeneous solution. The low surface tension and excellent wettability of EG enabled deep penetration into the mesopores, where nickel ions adsorbed onto the inner surfaces via capillary forces and diffusion. During subsequent thermal treatment, EG facilitated uniform nickel ion distribution while acting as a reducing agent, converting nickel ions into well-dispersed metallic nanoparticles anchored within the mesopores. Fourier transform infrared spectroscopy (FTIR) analysis confirmed the absence of new chemical bond signatures after EG integration (Fig. S1†), demonstrating that the catalyst architecture maintained its chemical integrity. For comparison, catalysts prepared using aqueous impregnation without EG were designated as 5% Ni AMOMs without EG and 5% Ni AMONs without EG. Additionally, a reference catalyst with a conventional architecture was synthesized by depositing 5 wt% Ni nanoparticles onto a SiO₂ support (denoted as 5% Ni SiO₂), with the same metal loading as the experimental group but differing in structural configuration.

Electron microscope images of the as-synthesized samples were captured to observe their morphologies. Transmission electron microscope (TEM) images revealed that AMONs consisted of monolayer micelle-assembled 2D nanosheets with a thickness of less than 20 nm (Fig. 2a–c). Its unique asymmetric structure arose from the distinct morphologies on both sides (Fig. 2b): one side featured open mesoporous channels, while the other was closed. The open side contained abundant mesopores with sizes ranging from 5 to 10 nm. After loading Ni metal by EG impregnation, Ni nanoparticles in the 5% Ni AMONs EG catalyst were uniformly dispersed within the mesoporous channels on the open side, showing small particle sizes with minimal agglomeration. Statistical analysis shows that these nanoparticles ranged from 3.1 to 8.6 nm (Fig. S2†), with an average size of 5.6 nm. The larger particles likely corresponded to Ni nanoparticles loaded on the closed side, where the absence of mesopore confinement prevented size control.¹⁷

Fig. 2 Morphology observation. (a)–(c) TEM and HRTEM images of 5% Ni AMONs EG. (d) TEM image of 5% Ni AMONs without EG. (e)–(h) HAADF-STEM and EDX elemental mapping images of 5% Ni AMONs EG. (i) SEM, (j) TEM and (k) HAADF-STEM images of 5% Ni AMOMs EG. (l) TEM image of 5% Ni AMOMs without EG. (m)–(p) HAADF-STEM and EDX elemental mapping images of 5% Ni AMOMs EG.

Fig. 2d presents the TEM images of 5% Ni AMONs without EG, prepared via aqueous impregnation. Without EG, metal dispersion within the mesopores was ineffective, resulting in large aggregated particles on the catalyst surface. This suggested that the asymmetric carrier properties were not fully utilized, causing the catalyst to behave more like a symmetric exposed-type material. Particle size analysis revealed a broader distribution, ranging from 3.3 to 37.1 nm (Fig. S3†), with an average size of 10.6 nm, further supporting this conclusion. Fig. 2e–h show high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and elemental mapping images of 5% Ni AMONs EG, confirming the uniform distribution of Ni nanoparticles and the Si/O composition of the support.

In contrast, scanning electron microscope (SEM) image revealed that AMOMs exhibited a distinct bowl-like morphology (Fig. 2i), and TEM and HAADF-STEM images further indicated that their structural units consist of mesopores with one open and one closed side, resembling the architecture of AMONs (Fig. 2j and k). However, unlike the flat nanosheet structure of AMONs, AMOMs were asymmetric, bowl-shaped structures formed by the collapse of spherical nanosheet assemblies, with diameters of around 250 nm. This unique 3D architecture retained the open/closed asymmetry while enhancing active site dispersion. Size analysis of Ni nanoparticles on 5% Ni AMOMs EG showed a wider distribution (2.6–8.5 nm, Fig. S4†) and a smaller average size (5.1 nm) compared to 5% Ni AMONs EG, indicating better metal dispersion and a higher active site density. To verify the reproducibility of the synthesis, multiple batches of 5% Ni AMONs EG and 5% Ni AMOMs EG were prepared. TEM images consistently demonstrated the asymmetric structure and effective Ni dispersion, confirming the robustness of the synthesis method (Fig. S5†). Additionally, Fig. 2l presents the TEM image of 5% Ni AMOMs without EG, synthesized via aqueous impregnation. The particle size ranged from 5.1 to 26.1 nm, with an average of 12.1 nm (Fig. S6†), demonstrating that without EG, metal particles failed to penetrate the mesopores effectively and instead aggregated on the surface. HAADF-STEM and elemental mapping images of 5% Ni AMOMs EG (Fig. 2m–p) also visually confirmed superior Ni dispersion compared to 5% Ni AMONs EG.

Physicochemical properties of the catalyst

The crystalline structures of the prepared materials were studied using X-ray diffraction (XRD) patterns (Fig. 3a), which revealed that all the samples exhibited a broad, weak diffraction peak at approximately 15°–30°, indicative of an amorphous SiO₂ structure. Diffraction peaks at 2θ angles of 44.5°, 51.9°, and 76.5° respectively corresponded to the (111), (200), and (220) crystal planes of metallic Ni.¹⁸ Notably, apparent Ni diffraction peaks appeared in the XRD patterns of 5% Ni AMOMs without EG, 5% Ni AMONs without EG, and 5% Ni SiO₂ (aqueous impregnation), while unconspicuous metallic Ni peaks were observed for 5% Ni AMOMs EG and 5% Ni AMONs EG. This indicated that EG-assisted impregnation effectively confined Ni particles within the mesoporous channels, leveraging the asymmetric support structure with an open and a closed side. The improved metal dispersion achieved through EG impregnation resulted in smaller Ni particle sizes compared to aqueous impregnation (Fig. S2, S4 and S5†). Additionally, the reducing environment provided by EG promoted superior metal dispersion and ensured effective entrapment of Ni within the mesopores.

Fig. 3 Catalyst characterization. (a) XRD patterns. (b) N₂ adsorption/desorption isotherms and (c) corresponding pore size distribution curves. XPS spectra of (d) O 1 s and (e) Ni 2p. (f) H₂-TPR profiles of the as-prepared samples.

We then used nitrogen adsorption–desorption isotherms of 5% Ni AMOMs EG, 5% Ni AMONs EG, and 5% Ni SiO₂ to determine their textural properties, where all exhibited reversible type IV hysteresis loops (Fig. 3b). Notably, the asymmetric materials 5% Ni AMOMs EG and 5% Ni AMONs EG displayed H5-type hysteresis loops, characterized by narrow adsorption branches and broad desorption branches, indicating complex mesoporous structures with partially blocked channels and asymmetric open/closed configurations.¹⁹ The observed desorption hysteresis confirmed pore mouth restrictions, further supporting the structural asymmetry. In contrast, 5% Ni SiO₂ exhibited an H3-type hysteresis loop, typical of spherical nanoparticle aggregates (Fig. S7†).²⁰ BET analysis showed that 5% Ni SiO₂ had the smallest surface area (129.2 m² g⁻¹), whereas 5% Ni AMOMs EG and 5% Ni AMONs EG demonstrated significantly larger BET surface areas of 550.7 m² g⁻¹ and 551.1 m² g⁻¹, respectively. Pore size distributions calculated using the Barrett–Joyner–Halenda (BJH) method revealed that both 5% Ni AMOMs EG and 5% Ni AMONs EG possessed mesopores primarily ranging from 2 to 30 nm, with average diameters of 8.5 nm (5% Ni AMOMs EG) and 9.5 nm (5% Ni AMONs EG) (Fig. 3c). The smaller mesopores (2–5 nm) likely corresponded to triangular interstices between the regularly arranged semi-open mesopores in the asymmetric supports, while the larger pores (10–30 nm) may represent inter-sheet spacings within the nanosheet assemblies.²¹ Notably, 5% Ni SiO₂, as an exposed-site catalyst, suffered from limited metal dispersion and severe particle agglomeration due to its low surface area and lack of structural confinement. In contrast, both 5% Ni AMOMs EG and 5% Ni AMONs EG, as confinement-type catalysts, exhibited comparable surface areas. However, 5% Ni AMOMs EG demonstrated superior metal dispersion and more uniform particle sizes compared to 5% Ni AMONs EG, attributed to its smaller average pore diameter that strictly restricted Ni nanoparticle growth through enhanced spatial confinement effects.

The surface characteristics and chemical state of the metal species in the synthesized samples were unveiled by X-ray photoelectron spectroscopy (XPS) results. Fig. 3d displays the O 1s XPS spectrum, where the peak at 532.9 eV is primarily attributed to the bridging oxygen bonds (O–Si–O) in the support.²² The peak near 531.4 eV corresponds to lattice oxygen in NiO, likely resulting from the partial oxidation of the catalyst upon exposure to air. Additionally, a weak shoulder at 534.5 eV is assigned to adsorbed molecular water and hydroxyl groups (–OH).²³Fig. 3e presents the Ni 2p XPS spectrum. The peaks at 873.2 eV and 855.0 eV are assigned to Ni²⁺ 2p_1/2 and Ni²⁺ 2p_3/2, respectively, with satellite peaks observed near 879.0 eV and 860.0 eV, indicating the oxidation of the catalyst when exposed in air.²⁴ For 5% Ni AMOMs without EG and 5% Ni AMONs without EG samples, the peaks at 869.5 eV and 852.0 eV were ascribed to Ni⁰ 2p_1/2 and Ni⁰ 2p_3/2, respectively.²⁵ Notably, these Ni⁰ peaks in the 5% Ni AMOMs EG and 5% Ni AMONs EG samples shifted to higher binding energies at 870.8 eV and 852.9 eV, respectively. This positive binding energy shift strongly correlates with the size effect of Ni⁰ particles in the EG-impregnated samples.²⁶ TEM characterization revealed that EG-assisted synthesis confined the growth of Ni⁰ nanoparticles, yielding smaller and uniformly distributed particles, whereas water-impregnated samples exhibited larger aggregated particles with a non-uniform size distribution. Smaller Ni⁰ particles possessed a higher proportion of surface atoms, where enhanced electron cloud polarization and reduced shielding efficiency of core electron layers collectively elevated the apparent binding energy of Ni⁰. These findings underscored the critical role of synthesis strategies in tailoring the electronic structure of supported nickel catalysts.

H₂ temperature-programmed reduction (H₂-TPR) profiles of the materials were also collected and are shown in Fig. 3f. All four catalysts exhibited distinct dual-peak behavior, indicating the presence of two distinct forms of NiO species with varying metal–support interactions. For 5% Ni AMOMs EG and 5% Ni AMONs EG, a low-temperature reduction peak (300–400 °C) corresponded to weakly interacting NiO, while a high-temperature peak between 400 and 500 °C represented strongly bound NiO species. The relative peak areas reflected the proportions of reducible NiO in each state. Notably, the high-temperature peak of 5% Ni AMOMs EG shifted to a higher reduction temperature compared to 5% Ni AMONs EG, indicating stronger metal–support interactions in the former. This enhancement was attributed to the optimized configuration of 5% Ni AMOMs EG, which promoted superior metal dispersion, tighter spatial confinement, and smaller pore dimensions.

In contrast, the non-EG-impregnated catalysts (5% Ni AMOMs without EG and 5% Ni AMONs without EG) displayed a dominant low-temperature peak with a larger area and a minor high-temperature peak. This observation confirmed that, in the absence of EG, most Ni species formed large, surface-exposed aggregates, with only a small fraction penetrating the pores to establish stronger interactions with the support. These findings aligned with TEM and particle size analyses, further validating that non-EG catalysts failed to exploit the structural advantages of the asymmetric support, resulting in inferior catalytic performance compared to their EG-impregnated counterparts. Combined with TEM observations and metal particle size analysis, these results demonstrated that EG-assisted impregnation effectively confined a significant fraction of metallic species within the mesopores, thereby enhancing metal–support interactions.

Catalyst performance evaluation

Thermocatalytic DRM performance of the samples was initially evaluated in a fixed-bed reactor at 300 °C, 500 °C, and 700 °C, respectively (Fig. 4a). Both the symmetric-structured catalyst (5% Ni SiO₂) and the 2D asymmetric catalyst (5% Ni AMONs EG) exhibited negligible activity at low-to-medium temperatures (300 °C and 500 °C), reflecting the strong endothermic nature of the DRM process, which requires higher temperatures and more effective catalysts to overcome the energy barrier. In contrast, the 3D asymmetric catalyst (5% Ni AMOMs EG) demonstrated measurable activity at 500 °C, indicating that the optimized asymmetric architecture enhanced metal dispersion and offered more accessible active sites to effectively lower the reaction energy barrier. All three catalysts showed catalytic activity at 700 °C, with 5% Ni SiO₂ achieving a CO₂ conversion of 12.3%; in contrast 5% Ni AMONs EG exhibited a significantly higher CO₂ conversion (53.9%), with H₂ and CO production rates of 1086.8 and 1702.0 mmol g⁻¹ h⁻¹, respectively. More importantly, the optimized 3D asymmetric catalyst, 5% Ni AMOMs EG, achieved a CO₂ conversion of 50.4% at 700 °C, with H₂ and CO production rates of 956.6 and 1593.2 mmol g⁻¹ h⁻¹, respectively (Fig. S8†), showing comparable performance to its 2D counterpart. These results highlighted the superior catalytic activity of asymmetric-structured catalysts in DRM at sufficiently high temperatures, outperforming conventional symmetric catalysts. The enhanced performance stemmed from the structural complexity of the asymmetric supports, high dispersion of active sites, and effective spatial confinement of metal particles, which mitigated agglomeration and preserved active site accessibility.²⁷ Notably, the significant disparity between H₂ and CO production rates observed for 5% Ni AMONs EG at 700 °C was attributed to the pronounced reverse water–gas shift (RWGS) side reaction (CO₂ + H₂ → CO + H₂O), where H₂ preferentially reduced CO₂ to CO.

Fig. 4 Performance evaluation of the synthesized samples. (a) Thermocatalytic DRM performance at different temperatures. (b) and (c) Photothermal catalytic DRM performance at varying temperatures. (d) Stability tests of photothermal catalytic DRM at 700 °C over 10 cycles. (e) Long term stability test of photothermal catalytic DRM with 5% Ni AMOMs EG at 500 °C over 12 h. (f) Photothermal catalytic DRM performance under light irradiation at different wavelengths. (g) Photothermal catalytic DRM performance with continuous light on/off. (h) Calculated activation energy of asymmetric catalysts in photothermal catalytic DRM.

Photothermal catalytic capability of the samples in DRM was then evaluated in a self-designed fixed-bed reactor using a CEL-PF300-T8 light source (irradiation intensity of 3.7 W cm⁻²). 5% Ni AMONs EG and 5% Ni AMOMs EG exhibited significantly higher activity under both standalone light irradiation and combined photothermal conditions compared to the conventional symmetric 5% Ni SiO₂ catalyst. Under light-only conditions, the CO₂ conversion rates for the asymmetric catalysts, 5% Ni AMONs EG and 5% Ni AMOMs EG, reached 6.9% and 12.6%, respectively, whereas 5% Ni SiO₂ showed negligible activity (Fig. 4b), strongly reflecting the significant role of asymmetric structure in solar driven catalysis. At 700 °C under photothermal conditions, the asymmetric catalysts maintained their superiority. 5% Ni AMONs EG achieved a CO₂ conversion of 72.9%, with H₂ and CO production rates of 2335.4 and 2662.9 mmol g⁻¹ h⁻¹, respectively. The optimized 5% Ni AMOMs EG demonstrated even higher CO₂ conversion (77.7%) under the same conditions, with H₂ and CO production rates of 2314.2 and 2945.3 mmol g⁻¹ h⁻¹, respectively. In stark contrast, the symmetric 5% Ni SiO₂ catalyst achieved only 16.6% CO₂ conversion, with H₂ and CO production rates of 109.8 mmol g⁻¹ h⁻¹ and 424.1 mmol g⁻¹ h⁻¹, respectively (Fig. S9†). The photothermal catalytic performance of asymmetric catalysts obviously surpassed that of reported symmetric catalysts, including some noble metal-based catalysts, particularly in terms of cost-effectiveness, operational durability, and photothermal efficiency under comparable or milder conditions (Table S1†).

Notably, the photothermal catalytic performance of 5% Ni AMOMs EG at 300 and 500 °C surpassed both its standalone thermocatalytic activity and that observed at equivalent catalyst temperatures (i.e., 390 and 551 °C) during light illumination (Fig. S10†). More importantly, the photothermal activity surpassed the simple additive effect of light and thermal contributions, underscoring the critical synergistic advantages of photothermal catalysis and how the asymmetric architecture amplified these benefits.²⁸ The light-to-fuel efficiency calculated for the two catalysts also clearly confirmed the advantages of 5% Ni AMOMs EG (Fig. S11†), whose light-to-fuel efficiency was about three times higher than that of 5% Ni AMONs EG, demonstrating its high light solar energy conversion efficiency.

To validate the role of EG impregnation in establishing asymmetric active sites in the catalyst, control samples prepared via aqueous impregnation (5% Ni AMONs without EG and 5% Ni AMOMs without EG) were tested under identical conditions (Fig. 4c). These aqueous-impregnated catalysts exhibited lower activity across all temperature ranges, with the most significant performance gap observed under photothermal conditions at 700 °C, where CO₂ conversion decreased by 31.8% (5% Ni AMOMs without EG) and 29.2% (5% Ni AMONs without EG), respectively. Additionally, photothermal stability for the two aqueous-impregnated catalysts at 700 °C was significantly poorer than that of the two catalysts using EG impregnation (Fig. 4d): after 10 consecutive chromatographic cycles, CO₂ conversion declined by 30.8% and 34.6%, respectively. This rapid deactivation aligned with behaviour typical of exposed-site catalysts and was attributed to severe Ni agglomeration, with average particle sizes increasing to 19.0 nm for 5% Ni AMONs without EG and 17.9 nm for 5% Ni AMOMs without EG (Fig. S12 and S13†).

For 5% Ni AMOMs EG, the observed deactivation was attributed to Ni nanoparticles located near the closed ends of the asymmetric channels. These particles, lacking spatial confinement and existing as larger aggregates on the surface, led to very slight increases in average particle size (i.e., 6.8 nm for 5% Ni AMONs EG and 5.4 nm for 5% Ni AMOMs EG). In contrast, Ni nanoparticles confined within the open-ended pores of the support maintained high dispersion due to EG-assisted impregnation (Fig. S14 and S15†), confirming that EG-assisted impregnation effectively confined metal particles within the mesopores of the asymmetric support, thereby enhancing catalyst stability under critical reaction conditions.

The photothermal stability tests were conducted at a lower temperature of 500 °C to reduce energy consumption towards practical application. 3D asymmetric catalyst of 5% Ni AMOMs EG demonstrated remarkable stability over a 12-hour photothermal test (Fig. 4e). An initial activation period was observed, with CO₂ conversion increasing from 33.6% to 39.0% within the first 7 hours, followed by stable performance with a negligible decline of 0.3% over the subsequent 5 hours. Besides, a longer-duration test of 5% Ni AMOMs EG under photothermal DRM conditions revealed a slight decline in performance after 8 hours, which then stabilized after 29 hours (Fig. S16†). Thermogravimetric (TG) analysis indicated minimal carbon deposition on the spent catalyst (Fig. S17†), further supporting that the performance loss between 8 and 29 hours was primarily due to sintering of Ni nanoparticles located near the closed ends of the asymmetric channels. Therefore, this structural design maximized Ni nanoparticle confinement within open-ended mesopores, ensuring superior dispersion and minimizing agglomeration-driven deactivation. The increased structural complexity and optimized asymmetry suppressed metal sintering, enabling sustained photothermal catalytic performance.

The photothermal catalytic DRM performance of 5% Ni AMOMs EG was further evaluated under different light wavelength ranges (Fig. 4f). Comparative tests under different band regions including (300–780 nm), (420–780 nm) and (490–780 nm) revealed that light below 420 nm contributed most significantly to activity enhancement. Specific wavelengths, λ = 380 nm and λ = 420 nm, also moderately improved catalytic efficiency. These results confirmed that high-energy photons from strong UV light absorption by the material excited electrons to higher energy states, generating hot electrons for direct participation in DRM.²⁹

Photothermal repeatability experiments (Fig. 4g) demonstrated rapid performance degradation upon light cessation, followed by immediate recovery upon re-illumination. This reversible behavior persisted over multiple on/off cycles, confirming the outstanding photoresponsiveness of the asymmetric catalyst.³⁰ The activation energies of the two asymmetric catalysts were then calculated using the Arrhenius equation. The photothermal activation energy of 5% Ni AMONs EG was determined to be 31.0 kJ mol⁻¹, while that of 5% Ni AMOMs EG was 28.8 kJ mol⁻¹ (Fig. 4h). The lower activation energy of the asymmetric catalysts highlighted their structural advantages in reducing reaction energy barriers, accelerating kinetics, and enhancing efficiency. The even lower activation energy of 5% Ni AMOMs EG further underscored the superior catalytic performance of its optimized 3D asymmetric architecture.

Mechanism elucidation

UV-vis absorption capability of different samples was investigated to study the structure–performance relationship (Fig. 5a). Notably, the absorption values of 5% Ni AMOMs EG and 5% Ni AMONs EG were lower than those of Ni SiO₂, indirectly reflecting the shielding of Ni nanoparticles by the confinement of the asymmetric nanostructure.³¹ Despite the limited number of photons absorbed, the utilization efficiency of solar energy in photothermal catalysis was still significantly enhanced, demonstrating the greater solar conversion ability of the unique structure. As such, in situ EPR tests were initially conducted under a CO₂ atmosphere with light irradiation (3.7 W cm⁻²) to detect CO₂⁻ signals (Fig. 5b). No CO₂⁻ signals were detected in the dark. After 20 minutes of illumination, signal peaks were observed for all four materials. However, the DMPO-CO₂ adduct peak intensities of 5% Ni AMOMs EG and 5% Ni AMONs EG were weaker than those of 5% Ni AMOMs without EG and 5% Ni SiO₂, which can be attributed to the limited accessibility of DMPO to CO₂⁻ species confined within the pores of the asymmetric catalysts. Among the asymmetric systems, 5% Ni AMOMs EG exhibited stronger peak intensity than 5% Ni AMONs EG, demonstrating that the optimized asymmetric structure of 5% Ni AMOMs EG generated more energetic hot electrons directly participating in the reaction. This further validated its superior catalytic performance in the photothermal DRM reaction due to the well-designed asymmetric configuration.

Fig. 5 (a) UV-vis spectra of the prepared samples. (b) In situ EPR spectra using DMPO as a trapping agent under light and dark conditions. (c) FEM simulations for assessing the internal electrical field on 5% Ni SiO₂ and 5% Ni AMOMs EG. (d) Surface temperature of the catalyst during photothermal catalytic DRM (light intensity: 3.7 W cm⁻²), captured using an infrared thermal camera. (e) Catalyst bed temperature measured using a thermocouple. (f) FEM simulations for assessing photo-thermal conversion capability for 5% Ni SiO₂ and 5% Ni AMOMs EG.

Finite element method (FEM) simulations were employed to examine the local electric field distribution around Ni nanoparticles situated on the surface of a SiO₂ support and those confined within the bowl-like nanostructured SiO₂. Based on TEM observations, simulation models representing 5% Ni SiO₂ and 5% Ni AMOMs EG were constructed (Fig. S18†). The results (Fig. 5c) revealed that the electric field enhancement factor near a 15 nm Ni particle in 5% Ni SiO₂ was approximately 2.5, whereas a significantly higher value of 250 was observed near a 5 nm Ni particle in 5% Ni AMOMs EG. This two-order-of-magnitude increase was attributed to both the unique bowl-shaped SiO₂ architecture, which modifies the local dielectric environment, and the nanoscale size effect of the Ni particles. Together, these factors synergistically strengthened the built-in electric field of the 5% Ni AMOMs EG catalyst.

Additionally, the photothermal conversion efficiency of the catalysts was in situ investigated using infrared thermography. As shown in Fig. 5d, the surface temperatures of the three catalysts after 5 minutes of illumination revealed that 5% Ni AMONs EG exhibited the highest surface temperature (505.9 °C), followed by 5% Ni AMOMs EG (493.6 °C), both significantly surpassing that of the conventional symmetric nanocatalyst 5% Ni SiO₂ (445.7 °C). This demonstrated that the induced electrical field within asymmetric structures generated more hot carriers for heat dissipation. Additionally, the bowl-shaped asymmetric SiO₂ shells in 5% Ni AMONs EG and 5% Ni AMOMs EG restricted heat conduction and radiation through thermal insulation and infrared shielding effects, respectively, thereby amplifying the photothermal effect.³² In contrast, 5% Ni SiO₂, lacking such confinement architecture, suffered from substantial heat loss via conduction, radiation and convection, resulting in inferior photothermal conversion efficiency.

Indeed, simultaneously recorded thermocouple data from the catalysts' lower surfaces (Fig. 5e) showed that 5% Ni SiO₂ achieved the fastest heating rate, reaching 131.9 °C within 5 minutes, compared to 110.6 °C for 5% Ni AMOMs EG and 127.7 °C for 5% Ni AMONs EG. Notably, while 5% Ni SiO₂ displayed the lowest upper-surface temperature, it exhibited the highest lower-surface temperature. This paradox arose from its loose structure facilitating downward heat transfer, leading to a minimal temperature differential between surfaces and inefficient thermal management. In contrast, both 5% Ni AMONs EG and 5% Ni AMOMs EG effectively concentrated thermal energy near active sites through their confinement structures, significantly reducing downward conduction. Particularly, 5% Ni AMOMs EG's highly intricate three-dimensional architecture addressed the limitations of 5% Ni AMONs EG's two-dimensional confinement, enabling superior heat entrapment and minimized thermal dissipation. In addition, FEM simulations (Fig. 5f) further confirmed the superior photothermal conversion capability of 5% Ni AMOMs EG compared to 5% Ni SiO₂. The results revealed that the heat power density on the Ni nanoparticles of 5% Ni AMOMs EG was nearly four orders of magnitude higher than that of 5% Ni SiO₂. This substantial enhancement indicates significantly stronger absorption of incident light by 5% Ni AMOMs EG, leading to a marked increase in photothermal conversion efficiency.

To investigate the catalytic mechanism in DRM, the in situ diffuse reflectance infrared fourier transform spectroscopy (DRIFTS) test was employed to analyze intermediate species under three catalytic conditions: photocatalytic (PC), thermocatalytic (TC, 500 °C), and photo-thermal catalytic (PTC, 500 °C) conditions. For the 5% Ni AMOMs EG catalyst, background spectra of the pre-reaction gas mixture and 10 time-resolved spectra were collected. As shown in Fig. 6a, b and S19,† peaks at 3013 cm⁻¹ (C–H stretching) and 1301 cm⁻¹ (C–H deformation) were observed under all conditions, corresponding to adsorbed CH₄ on the Ni surface.^33,34 The peak at 2830 cm⁻¹ was tentatively assigned to the symmetric deformation vibration of suggesting CH₄ dissociation into and H* species. In the 1360–1470 cm⁻¹ region, gradually intensified peaks were detected with increasing reaction time, potentially indicating further dissociation of intermediates.^35,36 Notably, no significant signals were observed under PC conditions, whereas TC exhibited rapid signal enhancement. Under PTC, the CH_x signals stabilized at a high intensity, implying that photothermal coupling effect promotes stable dissociation. All conditions displayed a persistent peak at 1090 cm⁻¹, attributed to CHO* species,³⁷ suggesting that C–H bond oxidation occurred prior to complete dehydrogenation to carbon, thereby suppressing carbon deposition. Antiresonance bands at 1661 cm⁻¹ (assigned to C=O of HCOO⁻) and 1588 cm⁻¹ (assigned to ν_as(O–C–O) of HCOO⁻) indicated dynamic consumption of CO₂⁻ derived intermediates. The HCOO⁻ signal intensified with reaction progress, implying interactions between CH₄-derived H species and HCOO⁻ for CO generation.³⁸ The strongest signals under PTC conditions suggested enhanced CO₂ adsorption and activation upon light irradiation.

Fig. 6 Mechanism investigation. (a) In situ DRIFTS of 5% Ni AMOMs EG in thermocatalysis (500 °C) and (b) photothermal catalysis (500 °C). (c) In situ EPR for the detection of DMPO-CH₃ signals over both asymmetric catalysts. (d) Proposed reaction pathway for 5% Ni AMOMs EG in photothermal catalytic DRM.

In the photothermal catalytic DRM, the generation of is a crucial step in CH₄ activation. The DRIFTS results also confirmed that was a stable and important intermediate during CH₄ dissociation. Therefore, in situ EPR tests were further conducted on 5% Ni AMONs EG and 5% Ni AMOMs EG catalysts, and DMPO was used as a spin-trapping agent to detect DMPO-CH₃ signals under illumination. As shown in Fig. 6c, the EPR spectrum under dark conditions displayed a flat baseline, indicating no generation. Upon illumination, a characteristic six-line hyperfine splitting pattern (12 peaks) was observed, matching the fingerprint of DMPO-trapped methyl radicals (DMPO-CH₃). This confirmed the generation of consistent with the in situ DRIFTS results, and the DMPO-CH₃ signal intensity increased as the reaction time progressed (Fig. S20†). Comparative analysis of the EPR spectra revealed that under identical reaction durations, the DMPO-CH₃ signal intensity for 5% Ni AMOMs EG significantly exceeded that of 5% Ni AMONs EG, indicating higher production on 5% Ni AMOMs EG. This demonstrated that under equivalent energy input, 5% Ni AMOMs EG exhibited superior CH₄ dissociation efficiency compared to 5% Ni AMONs EG, correlating with its enhanced catalytic performance.

Therefore, the structure–performance relationship and reaction pathway of the asymmetric catalyst can be inferred based on the above characterization studies. Compared to symmetric catalysts, the unique asymmetry induced an enhanced electrical field for more hot carriers and photothermal contributions. The increased energy carriers easily realized the dissociation of CH₄ molecules to CHO* and the activation of CO₂ molecules to CO₂⁻/HCOO⁻, which were finally converted into syngas, rather than coking (Fig. 6d). Additionally, the abundant porous structures offered significant confinement to promote active site accessibility, reduce heat loss and importantly avoid catalyst deactivation, thereby collectively leading to high-performance and satisfactory stability in photothermal catalytic DRM.

Conclusions

We developed asymmetric photothermal catalysts (5% Ni AMONs EG and AMOMs EG) through micelle-guided synthesis, creating unidirectional pore channels that enable selective Ni confinement. This asymmetric architecture enhanced metal dispersion (specific surface area >550 m² g⁻¹) and amplified built-in electric fields, directing more energetic hot carriers and photothermal energy toward reactant activation. The optimized 5% Ni AMOMs EG achieved an outstanding H₂ production rate of 2314.2 mmol g⁻¹ h⁻¹, outperforming conventional symmetric catalysts and even some reported noble metal-based systems. By elucidating the structure–performance interplay between asymmetric configurations and photothermal catalysis, this work advances catalyst design principles for efficient solar-driven conversion of greenhouse gases into clean fuels.

Data availability

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

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

The support from the scientific and technological innovation project of carbon emission peak and carbon neutrality of Jiangsu Province (BE2022024) and the National Natural Science Foundation of China (51676096) is acknowledged. This work was also partially supported by the Australian Research Council (DP240102707). J. Z. acknowledges the financial support from the Australian Research Council via an ARC DECRA Fellowship (DE250100753).

References

1. X. Si, R. Lu, Z. Zhao, X. Yang, F. Wang, H. Jiang, X. Luo, A. Wang, Z. Feng, J. Xu and F. Lu, Nat. Commun., 2022, 13, 258.
2. W. Leal Filho, R. A. Abeldaño Zuñiga, J. Sierra, M. A. P. Dinis, L. Corazza, G. J. Nagy and Y. A. Aina, Sci. Rep., 2024, 14, 14147.
3. Y.-J. Shen, Y.-H. Hsu, Y.-C. Chang, J.-J. Ma, K.-S. Peng, Y.-R. Lu, S.-H. Hsu and S.-F. Hung, ACS Appl. Mater. Interfaces, 2025, 17, 9378–9390.
4. Y. Sun, J. Xie, Z. Fu, H. Zhang, Y. Yao, Y. Zhou, X. Wang, S. Wang, X. Gao, Z. Tang, S. Li, X. Wang, K. Nie, Z. Yang and Y.-M. Yan, ACS Nano, 2023, 17, 13974–13984.
5. Z. Rao, Y. Cao, Z. Huang, Z. Yin, W. Wan, M. Ma, Y. Wu, J. Wang, G. Yang, Y. Cui, Z. Gong and Y. Zhou, ACS Catal., 2021, 11, 4730–4738.
6. R. Nandhini, D. Berslin, B. Sivaprakash, N. Rajamohan and D.-V. N. Vo, Environ. Chem. Lett., 2022, 20, 1645–1669.
7. M. H. Zhanjun He, T. Lin and L. Zhong, Acta Phys. – Chim. Sin., 2023, 39, 2212060.
8. G. Ji, S. Wu, X. Song, L. Meng, Y. Jia and J. Tian, Int. J. Hydrogen Energy, 2024, 57, 696–708.
9. S. M. Kim, A. Armutlulu, W.-C. Liao, D. Hosseini, D. Stoian, Z. Chen, P. M. Abdala, C. Copéret and C. Müller, Catal. Sci. Technol., 2021, 11, 7563–7577.
10. X. Zhu, Y. Wang, L. Lv, W. Tang, T. Zhang, Z. Zou and S. Tang, Ind. Eng. Chem. Res., 2024, 63, 15038–15050.
11. J.-Y. Wu, T.-H. Lai, M.-J. Fang, J.-Y. Chen, M.-Y. Kuo, Y.-H. Chiu, P.-Y. Hsieh, C.-W. Tsao, H.-E. Chang, Y.-P. Chang, C.-Y. Wang, C.-Y. Chen, M. Sone, W.-W. Wu, T.-F. M. Chang and Y.-J. Hsu, ACS Appl. Nano Mater., 2022, 5, 8404–8416.
12. J. Chen, Z. Ye, F. Yang and Y. Yin, Small Sci., 2021, 1, 2000055.
13. T. Zhao, X. Zhu, C.-T. Hung, P. Wang, A. Elzatahry, A. A. Al-Khalaf, W. N. Hozzein, F. Zhang, X. Li and D. Zhao, J. Am. Chem. Soc., 2018, 140, 10009–10015.
14. H. Chen, C.-T. Hung, W. Zhang, L. Xu, P. Zhang, W. Li, Z. Zhao and D. Zhao, J. Am. Chem. Soc., 2023, 145, 27708–27717.
15. H. Jiang, L. Hong, Y. Li and T. Ngai, Angew. Chem., Int. Ed., 2018, 57, 11662–11666.
16. S.-B. Qiu, Q. Zhang, W. Lv, T.-J. Wang, Q. Zhang and L.-L. Ma, RSC Adv., 2017, 7, 24551.
17. Y. Li, Z. Li, H. Cong, C. Yang, G. Cheng and W. Luo, ACS Sustainable Chem. Eng., 2022, 10, 3682–3689.
18. D. Wang, Q. Li, C. Han, Z. Xing and X. Yang, ACS Cent. Sci., 2018, 4, 112–119.
19. S. Naghdi, A. Cherevan, A. Giesriegl, R. Guillet-Nicolas, S. Biswas, T. Gupta, J. Wang, T. Haunold, B. C. Bayer, G. Rupprechter, M. C. Toroker, F. Kleitz and D. Eder, Nat. Commun., 2022, 13, 282.
20. L. Zhang, K. Qian, X. Wang, F. Zhang, X. Shi, Y. Jiang, S. Liu, M. Jaroniec and J. Liu, Adv. Sci., 2016, 3, 1500363.
21. X.-Y. Yang, L.-H. Chen, Y. Li, J. C. Rooke, C. Sanchez and B.-L. Su, Chem. Soc. Rev., 2017, 46, 481–558.
22. F. Zhao, Y. Feng, Y. Wang, X. Zhang, X. Liang, Z. Li, F. Zhang, T. Wang, J. Gong and W. Feng, Nat. Commun., 2020, 11, 1443.
23. Y. Xing, B. Zhu, L. Hong, C. Xia, B. Wang, Y. Wu, H. Cai, S. Rauf, J. Huang, M. I. Asghar, Y. Yang and W.-F. Lin, ACS Appl. Energy Mater., 2022, 5, 15373–15384.
24. L. Wu, G. Wang, S. Shi, X. Liu, J. Liu, J. Zhao and G. Wang, Adv. Sci., 2023, 10, 2304218.
25. D. Zhang, H. Zhao, B. Huang, B. Li, H. Li, Y. Han, Z. Wang, X. Wu, Y. Pan, Y. Sun, X. Sun, J. Lai and L. Wang, ACS Cent. Sci., 2019, 5, 1991–1997.
26. H. Wang, X.-K. Gu, X. Zheng, H. Pan, J. Zhu, S. Chen, L. Cao, W.-X. Li and J. Lu, Sci. Adv., 2019, 5, eaat6413.
27. M. Wu, X. Yang, X. Cui, N. Chen, L. Du, M. Cherif, F.-K. Chiang, Y. Wen, A. Hassanpour, F. Vidal, S. Omanovic, Y. Yang, S. Sun and G. Zhang, Nano-Micro Lett., 2023, 15, 232.
28. L. Hong, H. Zhang, L. Hu, R. Xiao and S. Chu, Sci. Adv., 2024, 10, eadn9441.
29. X. Han, P. Cheng, S. Han, Z. Wang, J. Guan, W. Han, R. Shi, S. Chen, Y. Zheng, J. Xu and X.-H. Bu, Chem. Sci., 2024, 15, 3530–3538.
30. Q. Pan, M. Abdellah, Y. Cao, W. Lin, Y. Liu, J. Meng, Q. Zhou, Q. Zhao, X. Yan, Z. Li, H. Cui, H. Cao, W. Fang, D. A. Tanner, M. Abdel-Hafiez, Y. Zhou, T. Pullerits, S. E. Canton, H. Xu and K. Zheng, Nat. Commun., 2022, 13, 845.
31. Y. Zou, K. Xiao, Q. Qin, J.-W. Shi, T. Heil, Y. Markushyna, L. Jiang, M. Antonietti and A. Savateev, ACS Nano, 2021, 15, 6551–6561.
32. M. Cai, Z. Wu, Z. Li, L. Wang, W. Sun, A. A. Tountas, C. Li, S. Wang, K. Feng, A.-B. Xu, S. Tang, A. Tavasoli, M. Peng, W. Liu, A. S. Helmy, L. He, G. A. Ozin and X. Zhang, Nat. Energy, 2021, 6, 807–814.
33. W. Jiang, J. Low, K. Mao, D. Duan, S. Chen, W. Liu, C.-W. Pao, J. Ma, S. Sang, C. Shu, X. Zhan, Z. Qi, H. Zhang, Z. Liu, X. Wu, R. Long, L. Song and Y. Xiong, J. Am. Chem. Soc., 2021, 143, 269–278.
34. Y. Li, D. Li, H. Liu, Y. Lei, R. Zhao, D. He, Z. Zheng, H. Luo and A. Liu, Catal. Sci. Technol., 2024, 14, 2722–2729.
35. Z.-Y. Zhang and T. Xie, J. Colloid Interface Sci., 2025, 677, 863–872.
36. J. Chen, M. Jiang, W. Xu, J. Chen, Z. Hong and H. Jia, Appl. Catal., B, 2019, 259, 118013.
37. J. Zhang, K. Xie, Y. Jiang, M. Li, X. Tan, Y. Yang, X. Zhao, L. Wang, Y. Wang, X. Wang, Y. Zhu, H. Chen, M. Wu, H. Sun and S. Wang, ACS Catal., 2023, 13, 10855–10865.
38. Z.-Y. Zhang, T. Li, Z.-Y. Tang, D. He, J.-J. Tian, J.-Y. Chen and T. Xie, Chem. Eng. Sci., 2023, 274, 118710.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta01976a

---