Excessive carbon dioxide (CO₂) emissions from fossil fuel combustion are a primary factor in the disruption of the natural carbon cycle. Electrochemical conversion of CO₂ into high-value products (CO, CH₄, C₂H₄, HCOOH, C₂H₅OH, etc.) offers a promising pathway to reverse entropy and recycle carbon [1-4]. However, the electrochemical CO₂ reduction reaction (ECRR) is hindered by the kinetically favorable hydrogen evolution reaction (HER) and the challenge of achieving a highly selective reduction of CO₂ to specific products with similar thermodynamic reduction potentials [5, 6]. Among the ECRR products, methane (CH₄) is the main component of petroleum with the highest heat of combustion of 56 kJ/g [7], considered a promising energy to solve the global energy crisis. Various electro-reduction catalysts have been extensively investigated to achieve a highly selective ECRR to CH₄ [8-12]. For example, a carbon-encapsulated copper-doped cerium oxide composite (Cu/CeO₂@C) achieves a CH₄ Faraday efficiency of 80.3% due to the synergistic effect of Cu/CeO₂ and in situ-formed carbon with robust channels to assist charge transfer [13]. However, the complicated multiple proton and electron transfer steps, high overpotential, and poor selectivity of intermediate generation still block efficient CO₂ utilization by electrochemical reduction [14-16].

Among studied catalysts, copper is the only metal that simultaneously possesses positive hydrogen and negative oxygen adsorption energy, allowing the formation of the following deep intermediates [17, 18]. The anchoring of Cu NPs on ZIF-8 (Cu/ZIF-8) achieves a maximum C₂₊ FE of 61.0% at a current density of 400 mA/cm² [19]. The single-atom Ir/CuFeS₂ QDs catalyst exhibited a suitable redox potential and active catalytic components, tailoring the chemical microenvironment for precisely engineered single-atom active sites [20]. Ag₂Cu₂O₃ nanowires with Cu···Ag Lewis acid-base dual sites formed the highly stable Ag···C = O···Cu adsorption of the critical intermediate for CO₂ to CH₄ [21]. As demonstrated experimentally, the high-valence Cu²⁺ species would facilitate the hydrogenation and formation of hydrocarbon intermediates, showing more preference for the selectivity of CH₄ [22]. For example, recent reports have shown that Cu²⁺ species in the solid solution (e.g., Cu-Ce-O_x) maintain high valence to benefit the adsorption of the ^*CO intermediate, further the hydrogenation for high CH₄ faradaic efficiency [22]. In addition, a novel layered coordination polymer (CuPEDOT) keeps the Cu²⁺-EDOT coordination stable and promotes the hydrogenation of ^*CO intermediates, facilitating CH₄ formation and prohibiting the dimerization of ^*CO to C₂₊ products [23]. The proof-of-concept catalysts of SrTi_1-xCu_xO₃ also unraveled that the Cu²⁺ can promote activation/adsorption of reaction intermediates to produce CH₄ [24]. However, Cu²⁺ species are thermodynamically unstable with cathodic potential applied and inevitably reduced to Cu⁰ [22, 23, 25]. The strategy to stabilize Cu²⁺ in its high valence is speculated to facilitate the generation of CH₄ in ECRR.

Perovskite oxides (ABO_3-δ) possess a flexible electronic structure and adjustable elemental components, which enable them to meet the meticulous requirements under different catalytic environments, including HER, oxygen evolution reaction (OER), nitrogen electroreduction reaction (NRR) [26-29]. By incorporating or substituting different metals, the electronic structure and charge density distributions of catalysts can be altered to formulate a suitable allotropic environment [26, 30-32]. Sun et al. designed Cu-MOF and CuBi double perovskite catalysts, CuBi@Cu-MOF-15 achieved a FE_HCOOH of 93% at −1.1 V and a current density of 91.0 mA/cm², for CuBi@Cu-MOF-15 is more favorable to stabilize *HCOO [33], while other CuBi@Cu-MOF achieved 56% of FE_HCOOH [34]. In the ECRR catalyst studies with CH₄ as the product, the combination of Cu with oxyphilic metal La can lower the energy barrier from ^*CO to *CH_xO, facilitating the C—O cleavage for CH₄ formation [35]. The perovskite catalyst La₂CuO₄ has been applied to the electrochemical methanation of CO₂ via the structural evolution that generates the Cu/La₂CuO₄ heterostructure [36]. Furthermore, by introducing different A-site metals into perovskite catalysts, Ca₂CuO₃ and Sr₂CuO₃ showed different performance in terms of either selectivity or electrochemical activity, with uncoordinated copper sites on Ca₂CuO₃ contributing to the hydrogenation of ^*CO and ^*CHO intermediates to ^*CH₂O [37]. There has also been a deepening of the realization that the A-site metal affects the electroreduction process according to its physicochemical property rather than a single supporting function, as previously thought [29, 37].

Here, we propose that the A-site components exert certain effects on the electronic structure of perovskites, which in turn affect the stability of the B-site metal, leading to distinctive intermediate preference and electrochemical properties. We introduced alkaline earth metal into the copper-based perovskite to test this conjecture, synthesizing LaCa_0.4CuO_3-δ for ECRR. After the introduction of Ca, an electrochemical test and in situ spectroscopic analysis revealed that the Cu²⁺ species in LaCa_0.4CuO_3-δ remain stable during ECRR, while that of La₂CuO₄ and commercial CuO are reduced to Cu⁰. The LaCa_0.4CuO_3-δ with stable Cu²⁺ species exhibit a FE_CH of 59.6% ± 3.8% at −1.3 V vs. reversible hydrogen electrode (RHE) in H-cell, and 38.7% ± 5.7% FE_CH in membrane electrode assembly (MEA) configuration with a partial current density of 155.0 mA/cm². In situ Raman and density functional theory (DFT) calculations unraveled that the hydrogenation of ^*CH₂O to ^*CH₃O was promoted on LaCa_0.4CuO_3-δ, promoting the formation of CH₄.

The LaCa_0.4CuO_3-δ perovskite catalyst was synthesized by a combined sol-gel and calcination method (Fig. 1). The quantitative elements component of LaCa_0.4CuO_3-δ perovskite catalyst was measured by inductively coupled plasma optical emission spectrometry (ICP-OES), which indicated that the atomic ratio of Ca/La/Cu was 0.4:1:1 (Table S1 in Supporting information), and thus the obtained electrocatalysts were named to LaCa_0.4CuO_3-δ.

X-ray diffraction (XRD, Fig. 2a) analysis showed that the synthesized samples possessed orthorhombic phases with Cmca space group for both LaCa_0.4CuO_3-δ (PDF #82–1444) and La₂CuO₄ (PDF #89–8845). Compared to La₂CuO₄, the main peak corresponding to (131) plane of LaCa_0.4CuO_3-δ was slightly shifted to a higher angle (Fig. 2b), demonstrating the successful incorporation of Ca through doping (Fig. 2c).

Scanning electron microscopy (SEM) images reveal that commercial CuO, La₂CuO₄, and LaCa_0.4CuO_3-δ show a similar nanoparticle morphology (Figs. S1a-f in Supporting information). Transmission electron microscopy (TEM) image shows that LaCa_0.4CuO_3-δ gives a particle size of ca. 100 nm (Fig. 2d). The lattice spacings of 0.286 and 0.267 nm in the high-resolution TEM (HR-TEM) images corresponding to the (131) and (200) planes of LaCa_0.4CuO_3-δ (Fig. 2e and Fig. S2 in Supporting information). Energy-dispersive spectra (EDS) element mapping confirmed the uniform distribution of La, Cu, Ca, and O in LaCa_0.4CuO_3-δ (Fig. 2f), further confirming the successful introduction of Ca into the perovskite lattice.

The ECRR catalytic performance was firstly carried out in H-cell (Fig. S3 in Supporting information) with 0.1 mol/L KHCO₃ aqueous as electrolyte under ambient conditions using a three-electrode system. All potentials shown in the H-cell have been converted to RHE with iR correction unless otherwise stated. The linear sweep voltammetry (LSV) curves of LaCa_0.4CuO_3-δ, La₂CuO_4, and commercial CuO are similar, especially at large current densities (Fig. S4 in Supporting information). The electrochemical surface areas (ECSA) of the catalysts were investigated using double-layer capacitance measurements, and these three catalysts also showed no significant difference in ECSA (Figs. S5a-d in Supporting information).

However, for the ECRR performance, with the modulation and elemental introduction of A-site, the CH₄ selectivity showed an obvious increase. LaCa_0.4CuO_3-δ exhibited a FE_CH of 59.6% ± 3.8% at −1.3 V (vs. RHE) (Fig. 3a and Fig. S6 in Supporting information), while La₂CuO₄ yielded a ca. 18.3% FE_CH and commercial CuO had less than 10% (Fig. 3b and Fig. S7 in Supporting information). In addition, the commercial CuO favored C₂₊ products, and La₂CuO₄ had H₂ as the main product, whereas LaCa_0.4CuO_3-δ showed a high preference for CH₄ instead of these two types of co-products (Fig. 3c and Fig. S8 in Supporting information), whose CH₄ selectivity is also better than other previously reported catalysts (Table S3 in Supporting information). Other doped samples with different Ca contents were also conducted for ECRR performance, which showed the Ca doping influenced the selectivity of CH₄ better compared to La₂CuO₄ (Fig. S9 in Supporting information), and the LaCa_0.4CuO_3-δ performed best among these samples. Tafel slopes were also performed to understand the catalytic kinetics (Fig. S10 in Supporting information). LaCa_0.4CuO_3-δ gave a lower Tafel slope value than La₂CuO₄ and CuO, reflecting the fast catalytic kinetics of LaCa_0.4CuO_3-δ to produce CH₄. Furthermore, the stability of LaCa_0.4CuO_3-δ under a constant potential of −1.3 V (vs. RHE) was evaluated in the H-cell for more than 9200 s (about 2.5 h) with an average FE_CH of about 54.6% (Fig. S11 in Supporting information).

The performance of LaCa_0.4CuO_3-δ, La₂CuO_4, and commercial CuO were also evaluated in a two-electrode MEA with 1 mol/L KHCO₃ as the anolyte to understand the potential for industry application. Our A-site modulation also improved the ECRR performance of the catalysts for CH₄ production under high current conditions in MEA (Fig. S12 in Supporting information). Commercial CuO still has ethylene as the main product, and CH₄ production is negligible (Fig. S13 in Supporting information). The FE_CH of La₂CuO₄ was slightly increased to 8.8% ± 5.6% by introducing La in the A-site of perovskite structure (Fig. 3d). Then, by doping Ca to share the A-site that La possessed, LaCa_0.4CuO_3-δ reached a FE_CH of 38.7% ± 5.7% (Fig. 3e and Fig. S14 in Supporting information), corresponding to a j_CH of −154.9 mA/cm² (Fig. 3f), showing an explicit CH₄ preference in MEA (Table S4 in Supporting information). In addition, the voltage of the catalysts under MEA electrochemical reaction also showed no extinguished difference (Figs. S15a-c in Supporting information).

The samples loaded on carbon paper after electroreduction were subjected to XRD analysis. The characteristic peaks at 31.2° and 33.5°, corresponding to (131) and (200) lattice planes, respectively, are more distinctly retained in LaCa_0.4CuO_3-δ than La₂CuO₄ (Fig. 4a). Peaks at 48.1° and 58.5° were observed in LaCa_0.4CuO_3-δ but nearly disappeared in La₂CuO₄. Meanwhile, the main peaks in commercial CuO totally disappeared. Overall, the characteristic peaks in LaCa_0.4CuO_3-δ remained largely intact compared to those in La₂CuO₄, demonstrating that the introduction of Ca enhanced the stability of the perovskite structure without causing structural transformation during reduction. In addition, the atomic ratio of La/Ca/Cu in LaCa_0.4CuO_3-δ determined by ICP-OES also remained almost unchanged compared to that before the electroreduction (Table S2 in Supporting information).

After electroreduction, the morphology of LaCa_0.4CuO_3-δ revealed by TEM remained similar to the original sample mentioned before (Fig. S16 in Supporting information and Fig. 2d). The lattice fringe attributed to (131) and (200) facets of LaCa_0.4CuO_3-δ can be observed in the TEM images of LaCa_0.4CuO_3-δ after ECRR (Fig. 4b and Fig. S17 in Supporting information). The elemental mapping also showed a uniform distribution of elements as before (Fig. S18 in Supporting information). X-ray photoelectron spectroscopy (XPS) showed that Cu 2p in LaCa_0.4CuO_3-δ exhibited a main peak at 933.6 eV corresponding to Cu²⁺ 2p_3/2 (Fig. 4c). After electroreduction, this peak attributed to Cu²⁺ remained stable, whereas the peak corresponding to lower valence Cu species was detected in La₂CuO₄ (Fig. 4d), demonstrating the effect of Ca doping in preventing the electrochemical reduction of Cu²⁺.

Cyclic voltammetry (CV) was performed to determine the specific valence state change at negative potentials and to understand the stability of the catalysts. Commercial CuO exhibited typical peaks for the electrochemical reduction of Cu²⁺ (Fig. S19 in Supporting information), with the reduction peaks at 0.5 and 0.2 V attributed to Cu²⁺ to Cu⁺ and Cu⁺ to Cu⁰, respectively [38, 39]. CV for La₂CuO₄ showed smaller reduction peaks corresponding to Cu²⁺ to Cu⁺ and Cu⁺ to Cu⁰ (Fig. 4e) [40]. For LaCa_0.4CuO_3-δ, no significant reduction peak was observed in CV curves, indicating that the introduction of Ca can stabilize the structure of LaCa_0.4CuO_3-δ and maintain the valence state of Cu in +2 (Fig. 4f).

Subsequently, in situ Raman spectroscopy was performed to investigate the stability of Cu species and the CO₂ to CH₄ mechanism in LaCa_0.4CuO_3-δ during the ECRR process. As shown in the ex situ Raman of the catalyst, the peak at 606 cm^-1 in commercial CuO is related to Cu-O (Fig. S20a in Supporting information) [41]. La₂CuO₄ displayed Raman peaks at 695, 904, 1165, 1214, and 1412 cm^-1 corresponding to the structure of La₂CuO₄ (Fig. S20b in Supporting information) [36, 42]. The peaks at 530 and 606 cm^-1 in LaCa_0.4CuO_3-δ can be attributed to the characteristic stretching vibrations of the Cu-O bond, and the later peaks at 1164 and 1214 cm^-1 can be attributed to the typical perovskite structure (Fig. S20c in Supporting information) [36, 43].

In situ Raman spectroscopy of the three catalysts were collected at the negative potential of −0.5 V vs. RHE to investigate the structure stability during electroreduction (Fig. S21 in Supporting information) [44]. The commercial CuO was reduced to metallic Cu with the peak at 606 cm^-1 fading (Fig. S21a). In contrast, the other two perovskite catalysts were more stable, with the characteristic peak still present (Figs. S21b and c) [41, 45]. With more negative ECRR potentials applied, the peak at 1412 cm^-1 representing La₂CuO₄ also gradually disappeared, and the peak corresponding to Cu-O in LaCa_0.4CuO_3-δ remained stable in contrast, indicating a more stable structure of LaCa_0.4CuO_3-δ than La₂CuO₄ (Fig. 5a and Fig. S22 in Supporting information) [46, 47].

Moreover, in situ Raman spectroscopy of LaCa_0.4CuO_3-δ exhibited a peak at 1398 cm^-1 [38] corresponding to the ^*CH₃O intermediate during the ECRR process (Fig. 5b) [35, 48]. It suggested that the hydrogenation process and the intermediate pathway starting from ^*CHO for CH₄ formation were more favorable on the surface of LaCa_0.4CuO_3-δ compared to La₂CuO₄ (Fig. 5c). Combined with where the Cu²⁺ stability was achieved, it can be assumed that the Cu²⁺ species in LaCa_0.4CuO_3-δ facilitates the hydrogenation process and stabilizes the ^*CH₃O intermediates for higher CH₄ selectivity.

In order to further elucidate the reaction mechanism of the catalyst in the reaction process, DFT calculations were performed on La₂CuO₄ and LaCa_0.4CuO_3-δ (Figs. S23-S27 in Supporting information). The density of states (DOS) of Cu d-band and O p-band of La₂CuO₄ and LaCa_0.4CuO_3-δ were analyzed for the electronic property changes due to the doping (Figs. 5d and e). Since the center of the Cu p-band of LaCa_0.4CuO_3-δ was more approaching to the Fermi energy, more absorbed electronics on the surface of LaCa_0.4CuO_3-δ achieved better catalytic performance with higher electron occupation number, forming a new electronic state for intrinsic activity [49, 50]. As displayed in Fig. 5f, the rate-determining step of the ECRR on La₂CuO₄ was the hydrogenation of ^*CH₂O to ^*CH₃O, requiring 1.8 eV to overcome the energy barrier. By contrast, the rate-determining step on LaCa_0.4CuO_3-δ was the hydrogenation of ^*CO to ^*CHO, only 1.2 eV was required to overcome the energy barrier for further processing, making it easier for LaCa_0.4CuO_3-δ to form the subsequent intermediates ^*CH₂O and ^*CH₃O, which was in agreement with in situ Raman spectroscopy. The DFT calculation results indicated that LaCa_0.4CuO_3-δ lowered the energy barrier of hydrogenation of ^*CH₂O in the electroreduction process and facilitated the intermediate hydrogenation pathway to CH₄.

In summary, we developed a strategy for A-site modulation of La₂CuO₄ to stabilize Cu²⁺ for efficient electrochemical CO₂ reduction (ECRR) to produce CH₄. Specifically, a perovskite catalyst of LaCa_0.4CuO_3-δ was prepared by a sol-gel procedure. The structure of LaCa_0.4CuO_3-δ remains unchanged during ECRR. The LaCa_0.4CuO_3-δ exhibits a FE_CH of 59.6% at −1.3 V vs. RHE in H-cell and a FE_CH of 38.7% with a large partial current density of −155.0 mA/cm² in MEA. In situ Raman spectra and DFT calculations confirmed LaCa_0.4CuO_3-δ facilitates the hydrogenation from ^*CH₂O to ^*CH₃O to produce CH₄. This study provides an efficient strategy to stabilize Cu²⁺ and guides the further study of perovskite in ECRR.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Yuhan Zheng: Writing – original draft, Visualization, Validation, Methodology, Investigation, Conceptualization. Yunzhen Jia: Writing – original draft, Software. Xuelei Lang: Methodology. Dazhong Zhong: Writing – review & editing, Writing – original draft, Supervision, Methodology, Funding acquisition, Formal analysis. Jinping Li: Supervision, Resources. Qiang Zhao: Writing – review & editing, Supervision, Funding acquisition, Formal analysis, Data curation.

Acknowledgments

The authors gratefully acknowledge financial support from the National Natural Science Foundation of China (Nos. 22308246, 22478278), Central Government Guides the Special Fund Projects of Local Scientific and Technological Development (No. YDZJSX20231A015), and the Fundamental Research Program of Shanxi Province (No. 202203021212266).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111193.