Figure 5 TEM images of A) CdS-Au hybrid nanorods with 1.5±0.2 nm and B) 7.1 ± 0.8 nm Au tip. C) Hydrogen evolution efficiencies ratio between small-tipped and large-tipped heterostructures as a function of excitation fluence presented as peak intensity of the irradiation beam (red circles) (insert: normalized hydrogen generation %QY of small-tipped hybrids (red) and the large-tipped hybrids (blue) as a function of excitation fluences presented as peak intensity of the irradiation beam. Green circles represents the corresponding peak intensities used in the TA experiments. D) Photocatalytic quantum efficiency for the water reduction of Ni-CdSe@CdS hybrids with different tip size. Experimental quantum efficiency in dark green bars, and the light green bars are quantum efficiency corrected for metal absorption. The inserts is the TEM image of corresponding Ni-CdSe@CdS nanorods, from left to right: 2.3 nm, 3.1 nm, 5.2 nm, 8.9 nm, and 10.1 nm. E) Hydrogen production rate (blue) and Cd normalized rate (red) curves as a function of Au size domain in the hybrid nanoparticles. Negligible rates are measured for the CdS nanorods. The red curve is normalized to total Cd content to better express the basic metal size effect during hydrogen reduction. F) Measured QY (black dashed line) along with the non-monotonic kinetic model behaviour (blue solid line). Green and red dotted lines present limiting behaviours of the model for zero and infinite metal domain sizes, respectively. Error bars show the size distribution of the Au tip and the hydrogen production rate uncertainty. The scale bar is 20 nm. A-C) Reproduced with permission.[18] Copyright 2018, American Chemical Society. D) Reproduced with permission.[19] Copyright 2017, American Chemical Society. E-F) Reproduced with permission.[20] Copyright 2015, Springer Nature.
Ben-Shahar et al. found that CdS-Au hybrid nanorods with large Au tips of a diameter of 7.1 ± 0.8 nm (Figure 5B) can better facilitate the separation and transference to metal domains of multiple excitons generated by high excitation fluence than small Au-tipped (1.5 ±0.2 nm) hybrid nanorods (Figure 5A). Figure 5C compares the quantum yield (QY) ratio of the large-tip to small-tip hybrid nanorods, indicating that the small tip favors hydrogen evolution in the single exciton region, and while in the multi-exciton region this ratio drops sharply to near 1:1. This is mainly attributed to the transcendence of Auger recombination over electron transfer from the semiconductor to the metal tip for small tip hybrid nanorods, resulting in the loss of most of the excited electrons. Although the increasing the intensity of excitons at the initial stage of excitation can improve the QY, it eventually reduces the QY in the small-tip hybrid nanorods. For large-tip hybrid nanorods, the photocatalytic performance can be enhanced distinctly because the ultrafast electron transfer plays a dominant role than the Auger recombination and can extract all the excited electrons.[18] Nakibli et al. prepared CdSe@CdS hybrid nanorods with different sizes of Ni tips, and the characterizations showed that the lowest emission QY, maximum amplitude value (amplitude reflected charge transfer probability) and the faster depopulation process of conduction band were obtained on the 5.2 nm-sized Ni tips, which is consistent with the results of the highest quantum efficiency achieved at this size (Figure 5D).[19] This is reasonable given the fact that the two processes of Coulomb blockage, which has a greater influence on the small-tip, and the Schottky barrier height, which affects the large-tip more, compete with each other, thereby producing an optimum radius in the middle.[19] Selective small metal island growth on the CdS nanorods was obtained by varying the irradiation time and the Au3+/nanorod ratio. Figure 5E (blue curve) shows the relationship between hydrogen production rate and Au tip size. The weak dependence observed on the two smallest dimensions while the photocatalytic performance of the larger Au domain is significantly reduced. Figure 5 F shows that the non-monotonic relationship between QY and Au tip sizes which is consistent with the standardized experiment QY (dotted line connected square). The opposite behavior of QY at the small and large size of Au domain could be attribute to the different main factors influencing the charge transfer rate. For the former, the precipitation of hydrogen QY is mainly determined by the electron injection rate at the metal tip, while the on the larger Au tips the water reduction on the metal surface will play the dominant role. Therefore, the intermediate Au tip size provides the best balance between charge separation rate and efficiency.[20]
3.3. Surface location of metal nanoparticles
The place where metal was decorated on the semiconductor nanorods also affects the catalytic performances of the photocatalysts, owing to the variation of the rate of charge transfer and effectiveness of electron-hole seperation. Besides, the loading in different positions will change the amount of metal requirements, which can affect the cost of the noble metal-semiconductor photocatalyst.[21]