Solar-driven water splitting to produce hydrogen is a promising renewable energy solution. Titanium dioxide (TiO₂) is a widely used photocatalyst due to its low cost, stability, and efficiency under UV light. However, TiO₂ suffers from high charge-carrier recombination and slow proton reduction kinetics. Strategies to improve TiO₂'s performance include doping, defect engineering, heterojunction formation, and cocatalyst loading. Loading noble metal cocatalysts like Pt, Au, and Pd enhances charge separation and proton reduction, but these metals are expensive and the efficiency remains moderate. Earth-abundant transition metals, such as Cu, offer a cost-effective alternative. Single-atom catalysts (SACs) maximize atom utilization, but achieving high loading amounts of SACs (>0.5 wt%) remains challenging due to aggregation and leaching. The easily changeable valence states of Cu nanoparticles make them promising for efficient charge separation and transfer, as demonstrated by a recent wrap-bake-peel process achieving 45.5% AQE at 340 nm. This study aims to develop a more efficient strategy to stabilize Cu SACs and create an in-situ self-heal approach for continuous H₂ production.

Existing literature highlights the challenges in maximizing the efficiency and cost-effectiveness of photocatalytic hydrogen evolution. While TiO₂ is a promising photocatalyst, its limitations in charge carrier recombination and proton reduction kinetics require improvement. Noble metals have been successfully employed as cocatalysts, but their high cost necessitates the exploration of earth-abundant alternatives. Transition metals, particularly copper, have shown potential due to their readily adjustable oxidation states, yet achieving high loading amounts of single-atom catalysts (SACs) without aggregation or leaching remains a significant hurdle. Recent studies using various methods to synthesize Cu SACs demonstrate improvements in apparent quantum efficiency (AQE), motivating the search for a more effective and stable approach.

The researchers developed a novel bottom-up approach to synthesize the Cu SACs on TiO₂, distinct from typical post-treatment methods. This method utilizes metal-organic framework (MOF) MIL-125 as a precursor. Metal ions (Cu²⁺ in this case) were anchored into the MOF MIL-125, forming metal-oxygen-titanium bonds. This strategy ensures uniform immobilization of metal SACs. The metal-MIL-125 intermediates were subsequently calcined to obtain the final photocatalysts. The amount of Cu was optimized to achieve a high loading of approximately 1.5 wt%. The photocatalytic H₂ evolution activity was evaluated using a full glass automatic on-line trace gas analysis system (Lab solar-6A) and a Multichannel photochemical reactor (PCX-50C) under simulated solar light irradiation. AQE was determined using a 1 W UV LED (365 nm) light source. Characterization techniques included XRD, FE-SEM, TEM, HAADF-STEM, BET, UV-vis spectroscopy, XPS, EPR, PL, time-resolved fluorescence decay spectroscopy, electrochemical impedance spectroscopy (EIS), transient absorption spectroscopy (TAS), DFT calculations, and isotopic experiments. The isotopic experiments employed deuterated methanol and water (CD₂OD/H₂O or D₂O/CH₃OH) to determine the hydrogen source. The DFT simulations modeled the CuSA-TiO₂ system with 1.5 wt% Cu replacing Ti, tracking charge density changes under irradiation to understand electron transfer.

The optimized CuSA-TiO₂ photocatalyst exhibited a significantly high H₂ evolution rate of 101.7 mmol g⁻¹h⁻¹ under simulated solar light, exceeding the performance of other reported photocatalysts, including PtSA-TiO₂ (95.3 mmol g⁻¹h⁻¹). Remarkably, it achieved an unprecedented apparent quantum efficiency (AQE) of 56% at 365 nm, a significant improvement over state-of-the-art TiO₂-based photocatalysts (AQE of 4.3–45.5%). The high H₂ evolution rate was attributed to the highly dispersed Cu SACs (1.5 wt%) and their efficient charge separation and catalytic effects. Characterization studies confirmed the successful incorporation of Cu single atoms into the TiO₂ lattice, forming Cu-O-Ti clusters. HAADF-STEM imaging showed the atomic dispersion of Cu atoms within the Ti vacancies. PL spectroscopy and time-resolved fluorescence decay measurements demonstrated the effective reduction of charge recombination and faster electron transfer in CuSA-TiO₂ compared to pristine TiO₂. In-situ XPS and EPR analyses revealed a reversible redox cycle between Cu²⁺ and Cu⁺ during the photocatalytic reaction, suggesting an in-situ self-healing mechanism. Isotopic experiments confirmed that water was the primary source of protons for H₂ evolution, with methanol acting as a hole scavenger. DFT calculations supported the experimental findings, illustrating the electron accumulation on Cu atoms upon irradiation and the subsequent reduction of Cu²⁺ to Cu⁺. Photoelectrochemical measurements indicated enhanced charge separation in CuSA-TiO₂. Long-term stability tests showed consistent performance even after 380 days of storage.

The superior performance of CuSA-TiO₂ is attributed to several factors: the high loading of atomically dispersed Cu atoms, the efficient charge separation facilitated by the Cu²⁺-Cu⁺ redox cycle (in-situ self-healing), and the strong interaction between Cu and the TiO₂ support. The bottom-up synthesis method using MOF MIL-125 as a precursor is crucial for achieving a high concentration of uniformly distributed Cu SACs. The in-situ self-healing mechanism ensures continuous photocatalytic activity without the need for regeneration, unlike some previously reported systems. The results demonstrate the potential of this novel strategy for designing highly active and stable photocatalysts for H₂ production, surpassing the performance of traditional noble metal-based catalysts.

This research successfully demonstrated a highly efficient and stable photocatalyst for H₂ production using a novel bottom-up synthesis strategy. The resulting CuSA-TiO₂ material exhibited exceptional performance with an AQE of 56% at 365 nm, exceeding current state-of-the-art catalysts. The observed high activity and long-term stability are attributed to the synergistic effects of the high loading of atomically dispersed Cu SACs, efficient charge separation, and an in-situ self-healing mechanism. Future work could explore other metal SACs using this method and investigate the optimization of reaction conditions for further performance enhancement.

While the study demonstrates exceptional performance, the use of methanol as a hole scavenger in the photocatalytic water splitting experiments should be noted as it is not a practical approach for large-scale hydrogen production. Further research should explore the use of alternative hole scavengers or strategies to improve the efficiency of hole utilization. Also, although long-term stability was observed, further prolonged studies under diverse operational conditions are beneficial to fully assess the long-term durability and robustness of the developed material.