Figure 3 A) Scheme of 1D nanocrystals growth by the SLS method. B-C) TEM images of B) Ag nanoparticles (the inset is its high magnification image), C) Ag-ZnS hybrids (the inset is its high-magnification image). The scale bar is 20 nm. A) Reproduced with permission.[1f] Copyright 2019, Wiley-VCH. B-C) Reproduced with permission.[9] Copyright 2012, The Royal Society of Chemistry.
3. Engineering strategies
Metal-semiconductor hybrid nanoparticles are composed of at least two components made from different materials and exhibit not only unique characteristics that are intrinsic to each component but also new properties that are not possessed by each component due to synergistic effects. State-of-art synthetic approaches enable the precise control of the size, shape, composition and spacial location of metal-semiconductor hybrid nanorods so one can regulate their functionalities through the engineering strategies to realize their full potential in photocatalysis. At present, engineering strategies for leveraging the photocatalytic activity of metal-semiconductor hybrid nanoparticles mainly include changing metal types, adjusting the metal domain size, growing metal nanoparticles in different locations, adjusting the length of nanorods, introducing of co-catalyst, regulating band gap structure, using various type of surface ligands or hole scavenger and so on. In this section, we will discuss and summarize the emerging engineering strategies.
3.1. Types of metal nanoparticles and semiconductors
Metal-semiconductor hybrid nanorods can be used as efficient catalysts for photocatalytic hydrogen evolution processes due to the enhanced charge separation capability and the reduction of the activation energy for the formation of hydrogen in a photocatalytic water splitting reaction. In the hybrid nanorods, the photoexcited electron generated by the light absorption of semiconductor is transferred to the metal domain to promote the water reduction reaction of hydrogen evolution. Each type of metals has different Fermi energy levels with diverse capabilities in separating charges and facilitating electrons to be transferred to the metal domain, so that metal-semiconductor hybrid nanorods may have diverse photocatalytic activities in hydrogen production.[10] Figure 4A is a schematic diagram of the Fermi energy levels of metals and band alignments of CdS. As shown in the Figure 4A, platinum has a relatively good reductive activity among these metals for catalytic water splitting reaction and has been widely studied.
Stone et al. prepared pristine CdS nanorods, hybrid Au-CdS and Pt-CdS heterostructures and compared their photocatalytic activities (Figure 4C and D).[11] The kinetics of hydrogen generation measured by gas chromatography showed that the hydrogen evolution efficiency of Pt-CdS hybrid nanorods was higher than that of Au-CdS hybrid nanorods, and both of them were larger than the bare CdS nanorods which is consistent with our previous discussion. Besides, they revealed that hydrogen peroxide and hydroxyl radical are produced more effectively by Au-tipped hybrid nanorods, which can be explained by d-band theory. Compared to the fully occupied d-band of Au metal, Pt tends to form strong Pt-O bond thus improves the dissociation of O-O bond. Tongying et al. performed hydrogen production by using CdSe/CdS and Pt-CdSe/CdS hybrid nanoparticles. Under broadband (UV/Visible) illumination, an obvious improvement of hydrogen production efficiency of Pt-CdSe/CdS hybrid nanoparticles can be clearly seen, and the hydrogen production rate is increased to 434.29 μmol h-1g-1 from 80.92 μmol h-1g-1 when Pt metal is introduced into the CdSe/CdS nanorods that forms Pt-CdSe/CdS hybrid nanoparticles.[12] Like other noble metals with excellent hydrogen evolution performances, Pd is considered to be an efficient catalyst for hydrogen production in previous research.[10h] However, Aronovitch et al.shows that the efficiency of Pd-CdS/CdSe photocatalyst for hydrogen production is not such satisfactory, which is even worse than the Au-tipped hybrid nanorods due to the severe etching of nanorods (shortened about 21%) during Pd deposition process in its deposition process.[13] In this process, partial substitution of Cd ions by Pd ions in the nanorods produces more defect sites on the surface of the nanorods and consequently reduces the photocatalytic efficiency (Figure 4E). Furthermore, this type of hybrid nanorods is more likely to break down in the photocatalytic process.[13]
Bimetallic or polymetallic components sometimes can achieve better catalytic activity than single metal-based hybrid nanorods due to the increased adjustability of the catalyst properties, optimized width and energy position of the surface d band, and the synergistic effect.[14] Aronovitch et al. synthesized Au@Au/Pd tip CdS/CdSe nanorods (Figure 4F), which greatly improved hydrogen generation efficiency.[13] This is attributed to a more favorable hydrogen release at the bimetallic tip. On the other hand, Pd tends to form oxidation layer on the surface which increases the time of light induction, while more inert Au can reduce the formation of it. In addition, Au reduces migration of Pd by forming a physical barrier, making the structure of hybrid nanorods less susceptible to change than Pd tip nanorods.[13]Kalisman et al. prepared a combination of the same nanorods with different metal tips (Figure 4G). The results show that when the Pt is coated on the outside of the Au tip to form the Au@Pt core-shell structure, the hydrogen generation rate is twice as high as that of pure Pt-tipped hybrid nanorods. This can probably attributed to the fast injection of photoexcited electrons to Au and the ability of fully extraction of photoexcited electrons of Pt. When photodeposition of Au and Pt is carried out simultaneously on the nanorods, a gold core decorated with a platinum island is formed, which has a hydrogen generation rate four times higher than that of the pure Pt tip. It is believed that many exposed interfaces introduced between the two phases can achieve improved hydrogen evolution activity.[15]
Generally speaking, noble metals can better extract charge for redox reaction at metal domain. However, it is difficult to achieve high-scale production due to the high cost of precious metals. Therefore, it is a trend of current research to seek abundant metal elements in the earth’s crust as the metal components for metal-semiconductor hybrid nanorods. The non-noble metal elements that are currently studied are nickel, cobalt, etc. Simon et al. firstly used the Ni decorated CdS nanorods for visible light photocatalysis and they proposed a two-step mechanism to promote hole transfer from CdS to the hole scavenger by a hydroxyl anion/radical redox shuttle to improve the efficiency of hydrogen evolution.[4d] The basic principle of the process is shown in the Figure 4B. In Ni-CdS hybrid nanorods, an external quantum yield of up to 53% and internal quantum yield up to 71% at high -OH ion concentrations with the system prevented photooxidation at a high pH are achieved.[4d] Maynadié et al. reported an example of heterogeneous growth of Co on CdSe nanorods (Figure 4H). With a Fermi level similar to Au, the band alignment of such hybrid nanorods will allow the transfer of electrons from the excited semiconductor to the metal tip within a shortened time period, which is consistent with the PL quenching observations in the photophysical studies. Due to ferromagnetism of Co, this type of hybrid nanorods are more commonly used in biological labelling and optoelectronic devices.[16]