光催化制氢作为一种具有前景的能源转化方式，受到了广泛关注。但是光催化过程中的三个步骤(光吸收、载流子分离、表面反应)效率较低，目前难以实现工业应用。研究者们对光催化的机理进行了深入研究，并提出了多种策略来调节半导体光催化剂的物理化学性质，以期有效提高光催化剂对可见光的吸收，降低光生载流子的复合，加速表面反应。上述策略包括：制造缺陷、局域表面等离子体共振、元素掺杂、异质结构建、助催化剂负载等。深入研究上述改性策略能够为设计制备高效稳定的光催化剂提供指导。因此，本综述聚焦于优化光吸收、载流子分离、表面反应的机理和改性光催化剂的制氢应用，并对构建高效制氢光催化剂的趋势做出了展望。
The photocatalytic hydrogen evolution reaction (PHER) has gained much attention as a promising strategy for the generation of clean energy. As opposed to conventional hydrogen evolution strategies (steam methane reforming, electrocatalytic hydrogen evolution, etc.), the PHER is an environmentally friendly and sustainable method for converting solar energy into H2 energy. However, the PHER remains unsuitable for industrial applications because of efficiency losses in three critical steps: light absorption, carrier separation, and surface reaction. In the past four decades, the processes responsible for these efficiency losses have been extensively studied. First, light absorption is the principal factor deciding the performance of most photocatalysts, and it is closely related to band-gap structure of photocatalysts. However, most of the existing photocatalysts have a wide bandgap, indicating a narrow light absorption range, which restricts the photocatalytic efficiency. Therefore, searching for novel semiconductors with a narrow bandgap and broadening the light absorption range of known photocatalysts is an important research direction. Second, only the photogenerated electrons and holes that migrate to the photocatalyst surface can participate in the reaction with H2O, whereas most of the photogenerated electrons and holes readily recombine with one another in the bulk phase of the photocatalysts. Hence, tremendous effort has been undertaken to shorten the charge transfer distance and enhance the electric conductivity of photocatalysts for improving the separation and transfer efficiency of photogenerated carriers. Third, the surface redox reaction is also an important process. Because water oxidation is a four-electron process, sluggish O2 evolution is the bottleneck in photocatalytic water splitting. The unreacted holes can easily recombine with electrons. Sacrificial agents are widely used in most catalytic systems to suppress charge carrier recombination by scavenging the photogenerated holes. Moreover, the low H2 evolution efficiency of most photocatalysts has encouraged researchers to introduce highly active sites on the photocatalyst surface. Based on the abovementioned three steps, multifarious strategies have been applied to modulate the physicochemical