Nitrogen-Defect-Driven PtCu Dual-Atom Catalyst for Photocatalytic CO₂ Reduction

Abstract

Owing to global energy demands and climate change resulting from fossil fuel use, technologies capable of converting greenhouse gases into renewable energy resources are needed. One such technology is photocatalytic CO₂ reduction, which utilises solar energy to transform CO₂ into value-added hydrocarbons. However, the application of photocatalytic CO₂ reduction is limited by the inefficiency of existing photocatalysts. In this study, we developed a nitrogen-deficient g-C₃N₄-confined PtCu dual-atom catalyst (PtCu/V_N-C₃N₄) for photocatalytic CO₂ reduction. Aberration-corrected high-angle annular dark-field scanning transmission electron microscopy and X-ray absorption fine structure spectroscopy confirmed the atomic-level anchoring of PtCu pairs onto the nitrogen-vacancy-rich g-C₃N₄ nanosheets. The optimised PtCu/V_N-C₃N₄ exhibited superior photocatalytic performance, with CO and CH₄ evolution rates of 13.3 µmol/g/h and 2.5 µmol/g/h, respectively, under visible-light irradiation. Mechanistic investigations revealed that CO₂ molecules were preferentially adsorbed onto the PtCu dual sites, initiating a stepwise reduction pathway. In situ diffuse reflectance infrared Fourier-transform spectroscopy identified the formation of a key intermediate (HCOO*), whereas interfacial wettability studies demonstrated efficient H₂O adsorption on PtCu sites, providing essential proton sources for CO₂ protonation. Photoelectrochemical characterisation further confirmed the enhanced charge-transfer kinetics in PtCu/V_N-C₃N₄, which were attributed to the synergistic interplay between the nitrogen vacancies and dual-atom sites. Notably, the dual-active-site architecture minimised the competitive adsorption between CO₂ and H₂O molecules, thereby optimising the surface reaction pathways. This study establishes a rational strategy for designing atomically precise dual-atom catalysts through defect engineering, achieving concurrent improvements in activity, selectivity, and charge carrier utilisation for solar-driven CO₂ conversion.

1. Introduction

The dual challenges of escalating global energy demands and anthropogenic climate change driven by relentless fossil fuel consumption and consequent CO₂ emissions underscore the critical need for sustainable technologies capable of converting greenhouse gases into renewable energy resources [1,2]. Photocatalytic CO₂ reduction, a process that utilises solar energy to transform CO₂ into value-added hydrocarbons (e.g., CO and CH₄), has emerged as a possible solution for addressing these intertwined crises. By simulating natural photosynthesis, this approach not only mitigates atmospheric CO₂ levels, but also produces storable chemical fuels, thereby closing the carbon cycle [3]. However, the practical implementation of this technology remains constrained by the inefficiency of existing photocatalysts, which are limited by inadequate photoabsorption, fast photogenerated charge–carrier recombination, and a lack of active sites for CO₂ activation and adsorption [4,5]. Among the various semiconductor materials explored, graphitic carbon nitride has attracted considerable interest because of its visible-light responsiveness, tuneable band structure, and robust thermal and chemical stabilities. Despite these advantages, the performance of pristine g-C₃N₄ is fundamentally limited by its narrow optical absorption spectrum (460 nm) and inefficient charge separation owing to the strong Coulombic interactions between the photogenerated electrons and holes. These intrinsic shortcomings necessitate innovative strategies for reengineering g-C₃N₄ at the atomic and electronic levels to unlock its full potential for solar-driven CO₂ conversion [6].

To overcome the limitations of conventional photocatalysts, recent advances have focused on defect engineering and atomic-scale metal incorporation to tailor the optical, electronic, and catalytic properties of the catalyst. Introducing nitrogen vacancies into g-C₃N₄ (denoted V_N-C₃N₄) represents a paradigm shift. These vacancies act as electron-deficient regions, creating mid-gap states that extend photoabsorption into the visible and near-infrared wavelengths, while simultaneously serving as electron traps to impede charge recombination. Furthermore, the undercoordinated carbon atoms adjacent to the vacancies can function as preferential anchoring sites for metal species, enhancing interfacial charge transfer. Parallel to defect engineering, the advent of dual-atom catalysts (DACs) has revolutionised heterogeneous catalysis by leveraging the synergistic interactions between adjacent metal sites [7,8]. DACs combine the multifunctional reactivity of bimetallic systems with the high atom-usage efficiency of single-atom catalysts, enabling precise control over reaction pathways. In this context, the strategic pairing of platinum (Pt) and copper (Cu) atoms offers a compelling synergy: Pt, with its strong CO₂ adsorption capability and low activation energy for C=O bond cleavage, complements the exceptional electron transfer kinetics and ability of Cu to stabilise key intermediates (e.g., *COOH and *CO). By immobilising PtCu dual atomic species on N-deficient g-C₃N₄, the resultant catalyst harmonises defect-mediated light harvesting with bimetallic cooperativity, creating a multifunctional platform for CO₂ reduction. This proposal not only optimises the adsorption-energy landscape for CO₂, but also establishes dual reaction centres to accelerate proton-coupled electron transfer, a rate-limiting step in photocatalytic CO₂ conversion [9].

Herein, we present a nitrogen-deficient g-C₃N₄-supported PtCu dual-atom catalyst (PtCu/V_N-C₃N₄) engineered for high-efficiency photocatalytic CO₂ reduction. This study not only establishes a blueprint for integrating defect engineering with dual-atom catalysis, but also elucidates the atomic-level interplay between vacancies and bimetallic sites, offering transformative insights for the rational design of photocatalysts. N-deficient g-C₃N₄ nanosheets were synthesised using a controlled alkali etching protocol, which selectively removed nitrogen atoms to generate vacancies while preserving the structural integrity of the material. This defect-rich framework not only broadened the photoabsorption edge to 620 nm, but also provided a high density of anchoring sites for the atomic-level dispersion of Pt and Cu. Through a codeposition strategy involving spatially confined reduction, Pt and Cu atoms were uniformly embedded in the N-deficient g-C₃N₄ lattice, forming stable PtN_X and CuN_X sites, as verified by the extended X-ray absorption fine structure (EXAFS). Aberration-corrected HAADF-STEM imaging further corroborated the atomic dispersion of PtCu diatoms, indicating direct metal–metal interactions. Under visible light (λ > 420 nm), the optimised PtCu/V_N-C₃N₄ demonstrated a CO evolution rate of 13.3 µmol/g/h, which was three to five times faster than those of the majority of g-C₃N₄-based photocatalysts that have been reported. In situ diffuse reflectance infrared Fourier-transform spectroscopy (DRIFTS) revealed the reaction mechanism: dual-atom species preferentially adsorb and activate CO₂ molecules while stimulating the C=O bond to form *COOH intermediates, thus lowering the energy barrier for its protonation to *CO, which subsequently desorbs as CO.

2. Results and Discussion

A defect confinement strategy was employed to deposit the PtCu dual-atomic species on the V_N-C₃N₄ nanosheets. As illustrated in Figure S1, the surface of V_N-C₃N₄ exhibits a negative charge (zeta potential of −8.2 mV). The PtCu dual-atomic species and V_N-C₃N₄ nanosheets can form a solid interface more easily because of their opposing surface charges [10]. The morphology of the as-prepared PtCu/V_N-C₃N₄ composite was investigated using aberration-corrected transmission electron microscopy (AC-TEM). As shown in Figure 1a, the V_N-C₃N₄ exhibits a surface-exposed nanosheet structure, which provides an ideal environment for anchoring bimetallic Pt and Cu atomic sites. As shown in Figure 1b,c, the presence of uniformly distributed bright spots on the PtCu/V_N-C₃N₄ nanosheets indicates the homogeneous dispersion of Pt and Cu atoms across the nanosheet. Furthermore, the coexistence of Cu and Pt species was observed on the same nanosheets. This observation confirms that the PtCu obtained through the impregnation method exists in a dual-atomic configuration, rather than as isolated Cu and Pt NPs. Finally, as illustrated in Figure 1d, the homogeneous distribution of C, N, Cu, and Pt across the surface of the PtCu/V_N-C₃N₄ nanosheets was revealed by elemental mapping of the PtCu/V_N-C₃N₄ composite using energy-dispersive X-ray spectroscopy (EDS). The successful loading of PtCu onto the PtCu/V_N-C₃N₄ nanosheets was further confirmed by the corresponding high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images and quantitative elemental analysis presented in Figures S2 and S3.

The Fourier-transformation expanded X-ray-absorbing microstructure (FT EXAFS) was analysed in the R space for the determination of Pt and Cu atoms (Figure 2a,b). Referencing the standard Pt foil, the Pt L₃-edge k³-weighted FT EXAFS of the PtCu/V_N-C₃N₄ sample exhibited a dominant peak at 1.57 Å, with no peaks corresponding to Pt–Pt bonds (2.57 Å). This observation ruled out the formation of metallic crystalline Pt clusters or nanoparticles and confirmed the atomic dispersion of Pt [11]. Similarly, referencing the standard Cu foil, the PtCu/V_N-C₃N₄ sample displayed a dominant peak at 1.44 Å, with no peaks corresponding to Cu–Cu bonds. This excludes the formation of metallic crystalline Cu clusters or nanoparticles, confirming that Cu also exists in the atomic form [12]. These findings are consistent with the HAADF-STEM results. Furthermore, the existence of defects and the interaction between the PtCu dual-atomic species and V_N-C₃N₄ in the mixed specimen were studied by X-ray photoelectron spectroscopy (XPS). XPS (Figure S4) showed that the feature peaks in the PtCu/V_N-C₃N₄ specimens can be attributed to four elements: C, N, Pt, and Cu, suggesting that PtCu/V_N-C₃N₄ was successfully synthesised by defect confinement. XPS was performed on the as-prepared PtCu/V_N-C₃N₄ samples to investigate their elemental composition and valence states. The Cu 2p XPS profiles were separated into two peaks, as illustrated in Figure 2c. The peaks of Cu 2p_3/2 and Cu 2p_1/2 of Cu²⁺ in PtCu/V_N-C₃N₄ are found at 934.7 eV and 954.5 eV, respectively [13]. As illustrated in Figure 2d, decompressing the C 1s XPS spectrum into three peaks at 288.1 eV, 286.6 eV and 284.8 eV is possible, which correspond to NC=N, CNHₓ, and sp² hybridised carbon (CC). The N 1s XPS spectrum can be matched with three peaks at 398.6 eV, 400.2 eV and 401.1 eV, as illustrated in Figure 2e, and these values are attributable to CN=C, N(C)₃, and CNHₓ. The Pt 4f XPS spectrum exhibits two characteristic peaks at 73.09 eV and 76.19 eV, as illustrated in Figure 2f, and they are attributed to Pt 4f_7/2 and Pt 4f_5/2 of Pt4+ in PtCu/V_N-C₃N₄ [14].

Figure 3a shows the X-ray diffraction (XRD) patterns of the as-synthesised specimens. The samples exhibited similar crystalline structures. Two distinct diffraction peaks were observed in all specimens at approximately 12.7° (100) and 27.6° (002), which correspond to the in-plane repeating units of the conjugated aromatic system and the interlayer stacking of g-C₃N₄, respectively [15]. Compared with g-C₃N₄, the (002) diffraction peak of V_N-C₃N₄ shows a slight leftward shift, indicating an increased interlayer distance according to Bragg’s law nλ = 2dsinθ [16]. However, no detectable diffraction peaks corresponding to Pt or Cu are observed for the PtCu/V_N-C₃N₄ composite, which may be attributed to their relatively low loading. Electron spin resonance (ESR) measurements conducted at room temperature (25 °C) on both g-C₃N₄ and V_N-C₃N₄ demonstrated that both samples exhibited a characteristic signal near g = 2.004 (Figure 3b), which was attributed to unpaired electrons in the heptazine rings [17,18,19]. Although both g-C₃N₄ and V_N-C₃N₄ possess carbon atoms with unpaired electrons within their heptazine ring structures, the successful incorporation of nitrogen vacancies into V_N-C₃N₄ induces the localisation of additional unpaired electrons on adjacent carbon atoms. This phenomenon was further substantiated by the significantly enhanced ESR signal intensity observed in V_N-C₃N₄ compared with that in g-C₃N₄. ESR spectroscopic analysis provided definitive evidence of the successful formation of nitrogen defects in V_N-C₃N₄. Furthermore, the stronger ESR signal of V_N-C₃N₄ indicated its enhanced electron delocalisation capability, thereby promoting a more efficient separation of the photogenerated charge carriers [20].

The textural properties of the samples, including specific surface area and pore characteristics, were systematically investigated through N₂ adsorption–desorption measurements. As illustrated in Figure 3c, g-C₃N₄, V_N-C₃N₄, and PtCu/V_N-C₃N₄ exhibited characteristic Type IV isotherms with H3-type hysteresis loops, indicating the formation of a mesoporous structure [21]. The specific surface areas were determined to be 30.488, 56.906, and 63.825 m² g⁻¹ for g-C₃N₄, V_N-C₃N₄, and PtCu/V_N-C₃N₄, respectively. The enhanced surface area provides an increased number of active sites for water molecule interactions, thereby improving photocatalytic activity. Furthermore, Barrett–Joyner–Halenda (BJH) analysis revealed that PtCu/V_N-C₃N₄ possessed a pore size distribution ranging from 1 to 70 nm, with particularly abundant pores in the 10–20 nm range (Figure 3d). For a comparative analysis, the specific surface areas, pore volumes, and pore size distributions of g-C₃N₄, V_N-C₃N₄, and PtCu/V_N-C₃N₄ are summarised in Table S1 (Supporting Information).

The photocatalytic properties of CO₂ were tested using a mixed quartz glass reactor in a closed gas circulation system. Figure 4 illustrates the photocatalytic CO₂ reduction activity of all synthesised catalysts under simulated solar irradiation. Among them, PtCu/V_N-C₃N₄ demonstrated enhanced CO production, reaching up to 13.3 µmol/g/h after 4 h of irradiation, which is significantly higher than that of single-atom Pt/V_N-C₃N₄ (5.1 µmol/g/h), single-atom Cu/V_N-C₃N₄ (6.0 µmol/g/h), V_N-C₃N₄ (3.3 µmol/g/h) and g-C₃N₄ (2.2 µmol/g/h) (Figure 4a). Notably, the anchoring of single atoms (Pt or Cu) resulted in a relatively modest enhancement in CO production, approximately 1.5 and 2 times that of V_N-C₃N₄, respectively, whereas PtCu/V_N-C₃N₄ exhibited a CO production rate approximately 4.1 times that of V_N-C₃N₄ [22]. The results demonstrate that the synergistic effect induced by the incorporation of dual-atomic PtCu can significantly suppress photo-induced electron and hole recombination, thus increasing the photocatalytic CO₂ reduction activity. In addition, after five successive photocatalytic cycles (20 h), the PtCu/V_N-C₃N₄ complex showed a negligible decline in CO and CH₄ yields and strong cyclic stability under long illumination conditions (Figure 4b). Additionally, to systematically evaluate structural stability, the post-reaction PtCu/V_N-C₃N₄ catalyst was analysed by XRD. As shown in Figure 4c, the crystalline structure did not change after 20 h of reaction, indicating that PtCu/V_N-C₃N₄ exhibited a stable structure.

The photocatalytic activity of semiconductor photocatalysts is predominantly governed by their capability to generate electron–hole pairs upon photoexcitation, which is fundamentally determined by the bandgap size and photo-absorption capacity of the semiconductor. As illustrated in Figure 5a, the composite photocatalyst (g-C₃N₄, V_N-C₃N₄, Pt/V_N-C₃N₄, Cu/V_N-C₃N₄, and PtCu/V_N-C₃N₄) was measured by diffusion reflectivity spectrometry (DRS). The g-C₃N₄ specimen exhibited an absorbing edge extending to 460 nm in the UV-visible region. In addition, the V_N-C₃N₄ sample exhibited a red-shifted absorption edge from 460 to 470 nm, accompanied by a broadened photoabsorption spectrum compared with that of g-C₃N₄ [23]. Notably, the Cu/V_N-C₃N₄, Pt/V_N-C₃N₄, and PtCu/V_N-C₃N₄ samples exhibited similar absorption edges. The respective bandgap energies, calculated using the Kubelka–Munk function, decreased from 2.76 eV to 2.58 eV (Figure 5b). These results suggest that a PtCu dual-atomic scheme can improve the photoabsorbing ability of the g-C₃N₄ defects. Notably, minimal shifts or even blue shifts were observed for the single-atom Pt/V_N-C₃N₄ and Cu/V_N-C₃N₄ samples. This observation suggests a synergistic effect between the Pt and Cu species, which optimises the photoabsorption properties of PtCu/V_N-C₃N₄, further confirming the successful preparation of dual-atom PtCu. This observation suggests a synergistic effect between the Pt and Cu species, which optimises the photoabsorption properties of PtCu/V_N-C₃N₄, further confirming the successful preparation of dual-atomic Pt–Cu.

Through photoelectrochemistry, the ability to separate and migrate photogenerated electrons and holes was studied, which demonstrated that PtCu/V_N-C₃N₄ has excellent catalytic properties. Figure 5c shows the instantaneous photocurrent response curves of specimens subjected to intermittent radiation. In general, a positive relationship exists between the optical response strength and the electron–hole separation efficiency, which indicates improved photocatalysis. The experiment demonstrates that V_N-C₃N₄ has a photocurrent density approximately five times that of g-C₃N₄, which suggests an improvement in photogeneration owing to the introduction of defect energy levels. The maximum photocurrent strength of PtCu/V_N-C₃N₄ was higher than those of V_N-C₃N₄ and g-C₃N₄ [24]. In addition, Figure 5d shows the periodic cyclic voltammetry (CV) profile of the specimen, which shows clear anodising and cathodic peaks for every specimen. The current density achieved on PtCu/V_N-C₃N₄ is double or triple those of V_N-C₃N₄ and g-C₃N₄, suggesting that PtCu/V_N-C₃N₄ has a remarkable catalytic effect in increasing the efficiency of electron–hole pair separation. The overpotential of hydrogen evolution was studied using linear sweep voltammetry. As shown in Figure 5e, the PtCu/V_N-C₃N₄ sample exhibited the lowest hydrogen evolution overpotential compared with V_N-C₃N₄ and g-C₃N₄, providing strong evidence that PtCu/V_N-C₃N₄ more effectively supplied hydrogen sources for the CO₂ reduction process [25]. As illustrated in Figure 5f, the electrochemical impedance spectroscopy (EIS) results of these specimens demonstrate that the PtCu/V_N-C₃N₄ composite exhibits a lower resistance after the introduction of the PtCu dual-atomic species.

To further explain the mechanism for the improved photocatalytic properties of the PtCu/V_N-C₃N₄ composite, the behaviour of the photogenerated charges in the photocatalyst was investigated through photoluminescence (PL) spectroscopy [26]. The intensity of the PL peak reflects the extent of electron–hole recombination, with a lower peak intensity indicating a reduced rate of electron–hole recombination. The PL spectra of g-C₃N₄, V_N-C₃N₄, Cu/V_N-C₃N₄, Pt/V_N-C₃N₄, and PtCu/V_N-C₃N₄ were recorded at an excitation wavelength of 300 nm. As illustrated in Figure S5, all samples exhibited similar peak shapes, with g-C₃N₄ demonstrating the highest peak intensity, indicating the most rapid electron–hole recombination rate among the samples. The peak intensity of V_N-C₃N₄ is significantly lower than that of g-C₃N₄, suggesting that the recombination of photogenerated charges in V_N-C₃N₄ is effectively suppressed. The PtCu/V_N-C₃N₄ sample exhibits the lowest peak intensity, indicating the highest charge-separation efficiency for the photogenerated carriers. In summary, the photochemical characterisation results suggest that the synergistic effect of PtCu dual-atomic species can accelerate the migration and transfer of photogenerated electrons, thereby enhancing the multifunctional photocatalytic conversion of CO₂ [27].

The water contact angles of g-C₃N₄ and V_N-C₃N₄ are 48.3°and 45.0°, respectively. Upon the introduction of PtCu atoms, the contact angle of PtCu/V_N-C₃N₄ decreased to 42.3° (Figure S6), indicating enhanced surface hydrophilicity, which facilitated the absorption of water and accelerated protonation reactions. The interaction of CO₂ molecules with the catalyst was studied by CO₂-TPD. As shown in Figure 6a, both PtCu/V_N-C₃N₄ and V_N-C₃N₄ exhibit a strong CO₂ desorption peak near 175 °C. However, for PtCu/V_N-C₃N₄, an enhanced CO₂ desorption peak appears at 450 °C as the temperature increases, which is attributed to the physical adsorption of CO₂ on the sample. In contrast, the CO₂ adsorption peak of V_N-C₃N₄ was weak and negligible [28]. The results indicate that, after loading PtCu dual-atom species, the multifunctional active sites on the surface of PtCu/V_N-C₃N₄ facilitate the adsorption of CO₂.

In situ DRIFTS spectroscopy revealed distinct characteristic vibrational signals corresponding to various intermediate species in the photocatalytic process. As illustrated in Figure 6b–f, CO²⁻ (1630 cm⁻¹), b-CO₃²⁻ (1599 cm⁻¹, 1325 cm⁻¹), and m-CO₃²⁻ (1514 cm⁻¹) exhibit negative signals [29]. The accumulation of these carbonate and carboxylate compounds over a long irradiation time (60 min) suggests that a significant intermediate HCOO⁻ is formed, which is responsible for photocatalytic CO₂ conversion. Furthermore, HCOO⁻ acts as the rate-determining intermediate for the photocatalytic reduction of CO₂ to CO and CH₄ [30]. The subsequent consumption of HCOO⁻ effectively participates in the photoreduction of CO₂, which is closely associated with the high selectivity toward methane and the remarkable reduction efficiency of CO₂. The results further corroborate the synergistic effect of the platinum–copper (PtCu) dual-active sites, which stabilise the HCOO^- intermediate and facilitate its protonation to form *CO, thereby ultimately driving the selective formation of hydrocarbon products. In contrast, the DRIFTS peak strengths of the intermediates in the in situ DRIFTS of g-C₃N₄, Cu/V_N-C₃N₄, and Pt/V_N-C₃N₄ were significantly lower, indicating that their ability to convert CO₂ in the presence of light was inferior to that of PtCu/V_N-C₃N₄.

As shown in Figure 7, the introduced PtCu dual-atomic species exhibited a synergistic photocatalytic effect, where the dispersed Pt sites functioned as hydrogen storage centres participating in hydrogenation reactions, while Cu acted as an active adsorption centre for CO₂. Both active centres promoted long-lived charge separation. Research indicates that copper species have a strong binding effect on the surface adsorption CO₂ molecules by means of a strong hybridisation of Cu and the 2 p orbital of CO₂. During the photocatalytic process, evidence from the dynamic trajectories of the charge carriers highlights the synergistic oxidation facilitated by the dispersed PtCu active sites. The bimetallic PtCu atoms act as active centres, taking advantage of their respective strengths. On the other hand, the scattered PtCu atoms offer an active site for ultrafast photogeneration of electron transfer, thus maximising the lifetime of charge separation on the V_N-C₃N₄ nanosheets. The main reaction mechanism is as follows [31].

CO₂ adsorption and activation:

CO₂ + PtCu sites → CO₂* (* denotes adsorbed species)

(1)

Formation of HCOO* intermediate:

CO₂* + H⁺ + e⁻ → HCOO*

(2)

Protonation to *COOH:

HCOO* + H⁺ → *COOH

(3)

Reduction to CO:

*COOH + H⁺ + e⁻ → CO + H₂O

(4)

Further reduction to CH₄ (minor pathway):

CO + 6H⁺ + 6e⁻ → CH₄ + H₂O

(5)

3. Experimental Section

3.1. Chemicals

Melamine (AR, 99%), methanol (AR, 99%), absolute ethanol (AR, 99.5%), tetrabutylammonium hydroxide (AR, 40%), and ammonium sodium borohydride (NaBH₄, 98%) were obtained from the Sinopharm Chemical Reagent Co., Ltd. Copper chloride (CuCl₂·2H₂O, ≥99.0%) and chloroplatinic acid hexahydrate (H₂PtCl₆·6H₂O, Pt ≥ 37.5%) were sourced from the Shanghai Macklin Biochemical Technology Co., Ltd. In the present study, all chemical agents used were of analytical quality and used as obtained, with no further purification processes.

3.2. Synthesis of g-C₃N₄

Pure g-C₃N₄ was prepared via calcination. First, 5 g of melamine was placed in a 50 mL ceramic crucible, which was heated to 550 °C at 5 °C per minute, followed by a 3 h isothermal treatment [32]. After cooling naturally, the product was ground into a fine powder. The resulting yellow powder was dispersed in a mixed solvent containing methanol, ethanol, and tetrabutylammonium hydroxide, followed by magnetic stirring for 24 h and ultrasonication for 8 h. The finished product was separated by centrifugation, cleaned thoroughly with deionised water and alcohol, and dried for 12 h at 80 °C. The resulting yellow powder was crushed and collected for subsequent use.

3.3. Synthesis of Defect State g-C₃N₄ Nanosheets (V_N-C₃N₄)

The synthesis commenced with the homogeneous blending of melamine and sodium borohydride (NaBH₄) in an equimass ratio (100:1) through mechanical grinding. The thoroughly mixed composite was then loaded into an alumina crucible and subjected to controlled thermal treatment in a tube furnace. The temperature was accurately increased from the ambient temperature (20 °C) to 550 °C at a rate of 5 °C per minute and then subjected to isothermal annealing treatment for 3 h in a continuous stream of argon [33]. The samples were allowed to cool to room temperature naturally. The obtained yellow powder was mechanically crushed and dispersed in a ternary solvent system containing methanol, ethanol, and tetrabutylammonium hydroxide (volume ratio of 10:10:1). The mixture was subjected to continuous magnetic stirring for 24 h, followed by a 24 h ultrasonication treatment. The final product was isolated via centrifugation, thoroughly washed with deionised water, and dried overnight under ambient conditions. The resulting yellow powder was collected for further characterisation.

3.4. Synthesis of PtCu/V_N-C₃N₄

A PtCu/V_N-C₃N₄ composite catalyst was prepared using a wet impregnation method. First, aqueous solutions of CuCl₂·2H₂O and H₂PtCl₆·6H₂O with Cu and Pt mass concentrations of 10 mg/mL were prepared. Subsequently, a precisely weighed quantity of 0.1 g of V_N-C₃N₄ nanosheets was homogeneously dispersed in 30 mL of deionised water. Precisely measured volumes of 0.267 mL of CuCl₂·2H₂O solution and 0.263 mL of H₂PtCl₆·6H₂O solution were sequentially introduced into the aforementioned dispersion. The resulting mixture was subjected to continuous magnetic stirring for 2 h and stilling for 24 h to facilitate the uniform adsorption of metal ions onto the surface of the V_N-C₃N₄ nanosheets. Finally, the powder was isolated by centrifugation, thoroughly washed with deionised water and anhydrous ethanol, and dried to obtain the PtCu/V_N-C₃N₄ photocatalyst. The synthesis of the PtCu/V_N-C₃N₄ photocatalyst is shown in Figure S7. Pt/V_N-C₃N₄ and Cu/V_N-C₃N₄ photocatalysts were prepared by the same method.

3.5. Characterisation

XRD images of the specimens were recorded in the 10–80° scanning range by means of a Bruker D8 Advance Apparatus with Cu Kα supply, working at 40 kV and 40 mA. The shapes and crystalline properties of the specimens were investigated using scanning electron microscopy (STEM, JEM-ARM 200p) and energy-dispersive X-ray spectrometry (EDS) (Bruker Xflash 5030) at an acceleration of 200 kV. The optical absorption characteristics were examined using UV-Vis DRS (UV2450 spectrophotometer, Cary 5000, Varian, USA) with BaSO₄ as the standard. An Escalab 250 apparatus (Escalab 250, Thermo Fisher, Waltham, USA) was used to analyse the specimen by means of XPS with a single-colour Al Kα irradiation source, and the C 1s peak binding energy was taken as a benchmark at 284.8 eV. The EPR (ESR) was measured with a JES-FA 200 spectrometer at 25 °C. A Zetasizer (Nano Z90, Malvern Instruments, Malvern City, United Kingdom) was used to measure the ζ-potential of the specimens.

3.6. Photocatalytic CO₂ Reduction Measurements

A Meery Change automated gas analyser (MC-SPB10) was employed to measure photocatalytic CO₂ reduction in H₂O. Using continuous mechanical agitation, 100 mg of catalyst, 100 mL of deionised water, and 100 μL of acetonitrile were used to create the catalyst system. A glass lid was used to seal the reaction container and to keep the system intact. Typically, the pressure inside the reactor is maintained below the normal value. Through the injection of cooling circulation water on the outside of the reactor, the reactor was kept at 6 °C. A gas chromatograph fitted with two detection systems was used to quantify the evolution of the gas: a heat conduction detector (TCD) and a flame ionisation detector (FID). Automatic sampling was carried out at 60 min intervals with a 5 A. Ultra-high purity argon (99.999%) was the carrier gas at a velocity of 20 mL/min. The GC was run with the injector held at 180 °C, the probe was kept at 360 °C, and the column was first set at 50 °C. Photocatalysis was performed using a 300 W xenon arc lamp (PLS-SXE300D, Beijing Perfect Light, Beijing, China) fitted with a 420 nm cut-off filter to provide 150 m W/cm² radiation.

3.7. Photoelectrochemical Test

Electrochemical characterisation, including transient photocurrent response spectroscopy, linear sweep voltammetry, the volt–ampere cycle method, and impedance spectroscopy, was performed using a conventional three-electrode configuration on a CHI660E electrochemical workstation (China). The electrochemical cell included a 0.5 M Na₂SO₄ aqueous electrolyte solution, with a platinum wire counter electrode and an Ag/AgCl reference electrode. The working electrode was fabricated by depositing the photocatalyst onto fluorine-doped tin oxide (FTO) conductive glass following the protocols established in our previous investigations. Photocurrent measurements were acquired under an applied bias of 0.5 V, with illumination provided by a 300 W Xe arc lamp. EIS was measured from 10⁻² to 10⁵ Hz, and the AC disturbance amplitude was 10 mV.

4. Conclusions

In summary, a new photocatalytic system was developed using a simple wet impregnation method to introduce PtCu atoms into nitrogen-deficient nanosheets. The PtCu dual-atomic species on the nitrogen-deficient carbon nitride nanosheets show synergy in the photocatalytic CO₂ reduction process, which facilitates tight interactions and quicker reaction dynamics. The synergistic action of the dual-atom metallic sites in PtCu/V_N-C₃N₄ not only combines the desired merits of Pt and Cu, but also accelerates the transport of photoproduced carriers in nitrogen-deficient carbon nitride nanosheets, thus improving the efficiency of CO₂ conversion. Furthermore, we demonstrated that the bimetallic PtCu atomic anchoring policy could decrease the loading of Pt while maintaining the maximum Pt mass activity. This new research holds significant importance for designing efficient photocatalytic materials with dispersed noble metals and improving effective atomic utilisation.