Fig. 1. Schematic diagram for the general electronic processes occurring in photocatalytic reaction. The electronic processes occurring for forward reaction are shown by green arrows; while the red arrows show the electronic processes related to the back reaction.

Fig. 2. Light-intensity-dependent SC band alignment and interfacial electron transfer (a-c) and the related photocatalytic behaviors (d,e). The photocatalytic hydrogen evolution (HE, squares) rates (d) generally match the simulated SC interfacial electron transfer (Jiet, solid lines) after correction (dash lines) with electron utilization (η = 0.294). Under weak illumination (a), the potential barrier (qVB) is high and the interfacial transfer of electrons to cocatalyst (cat) is dominated by trap-assisted charge recombination (CR). By intensifying irradiation (b), thermionic emission (TE) emerges to dominate SC interfacial electron transfer, making the photocatalytic reaction rate grows quadratically with the light intensity (d). Under very intense irradiation (c), qVB can be very low and photocatalytic photon utilization can be nearly constant (e). Adapted with permission from Ref. [5] (a-c) and Ref. [30] (d,e). Copyright 2023, American Chemical Society.

Fig. 3. Light-intensity-dependent HE rates (a) and the related gross quantum yields (b) for CdS-based photocatalysts with influence of manganese dopant concentration. Surface doping with manganese can be realized by hydrothermal treatment of CdS nanorods in Mn2+-containing solution. Mn-0, Mn-0.5, Mn-1, and Mn-3 respectively represent the molar ratios of 0, 0.5%, 1%, and 3% of manganese to cadmium of CdS in the hydrothermal dispersion. Reprinted with permission from Ref. [24]. Copyright 2022, Royal Society of Chemistry.

Fig. 4. Schematic diagrams for the three-channel electron transfer (a) and the shrinking of dispersed semiconductor (b) and photocatalyst (c) into porous films for OCP measurements. Adapted with permission from Ref. [5] (a). Copyright 2023, American Chemical Society. Adapted with permission from Ref. [13] (b,c). Copyright 2021, American Chemical Society.

Fig. 5. The OCP decay from the biased potential in dark or the photoinduced potential for TiO2 and Pt/TiO2 electrodes. By taking the derivation of OCP with respect to time (a), we can obtain the time constants (b) for transfer of electrons from semiconductor to solution with one channel (τ1C), two channels (τ2C), and three channels (τ3C). The time constant for SCS interfacial electron (τSCS) and side/back reaction (τSCS-SB) can be calculated from τ1C, τ2C, and τ3C (c). By these parameters, we can estimate photogenerated electron utilization (c), which is very close to photon utilization (~38%) of the photocatalytic reaction under very intense irradiation. The abscissa values for GQY (c) can refer the equilibrium potential of the photocatalyst (Pt/TiO2) under irradiation (d). Adapted with permission from Ref. [5]. Copyright 2023, the American Chemical Society.

Fig. 6. Schematic diagrams (a) and time constants (b,c) for interfacial transfer of electrons from semiconductor to the argon (b) and oxygen (c) saturated solution via catalytic sites. For the conventional construction loaded with metallic cocatalyst like Pt/CdS (left panel, a), SC interfacial electron transfer can be realized by the combination of trap-assisted charge recombination (CR) and thermionic emission (TE). For cocatalyst-free photocatalyst like W-CdS (right panel, a), in addition to transport/transfer of electron to the catalytic sites by conduction band affected by trapping/detrapping events, the levels in the bandgap can also transport charges. When the required bias is less negative (e.g.: oxygen reduction, c), the trap states in the deeper levels sensitively affects the electron transport in the levels in the bandgap (c). Reprinted with permission from Ref. [41]. Copyright 2022, the American Chemical Society.

Fig. 7. Influence of bridging metal on band alignment and interfacial electron transfer of SC contacts. For the semiconductor loaded with cocatalysts (cat) typically like platinum with large work function, the very high potential barrier (qVbi0) severely slows SC interfacial electron transfer by thermionic emission (TE, shown by the thin blue arrow, a). Although some electrons can be transferred to the cocatalyst by trap-assisted charge recombination (CR), the utilization of incident photons will be undesirably reduced as the light becomes intense. By introducing intermediate metal (mim) with small work function (b), the lowering of potential barrier (qVbi) can facilitate the interfacial transfer of electrons to the cocatalysts by thermionic emission (shown by the green thick arrow, b). Reprinted with permission from Ref. [50]. Copyright 2022, Elsevier.

Fig. 8. Schematic diagrams for common (bulk) metal-semiconductor contact (a), nanostructuring induced band bending confinement for the semiconductor nanostructures without (b) and with (c) ultrathin interlayers and the related interfacial electron transfer. The band structures for contacts and semiconductor-metal (SC) interfacial electron transfer (bottom panels of (a)-(c)) are sensitively influenced by the size of the semiconductor (upper panels of (a)-(c)). Reprinted with permission from Ref. [34]. Copyright 2020, the American Chemical Society.