Synthesis and Characterization of ZnOS-x. The hydrothermal method was used for macroscopic preparation of the nanoparticles of ZnOS (Experimental section in Supporting Information), which is green and economic. The trithiocyanuric acid (Na3C3N3S3·9H2O) and reacted with Zn2+ to yield the white nanoparticles, upon further thermal annealing in air atmosphere to obtain ZnOS heterostructures (Huang et al. 2016). The obtained ZnOS samples were investigated by SEM, TEM, PXRD, and N2 adsorption-desorption isotherms. The microstructure and components of ZnOS can be modulated by controlling the addition of trithiocyanuric acid in reaction solution for 1, 2, 3, 4 mol, and the finally obtained ZnOS heterostructures were denoted as ZnOS-1, ZnOS-2, ZnOS-3, and ZnOS-4, respectively. Different trithiocyanuric acid content caused the different aggregation of ZnOS nanoparticles and resulted in the performance change. The SEM images of Fig. 1a-1d shows the details of physical aggregation of the ZnOS-x nanostructures. Among SEM images in Fig. 1c show that ZnOS-3 particles are loosely overlapped together.
For the XRD pattern of ZnOS-x, some intense diffraction peaks at 28.6°, 33.2°, 47.5°, and 56.4°, being well matched with the (111), (200), (220), and (311) planes of sphalerite ZnS (JCPDS # 05-0566), respectively (Jiang et al. 2022; Wu et al. 2009). In addition, partial diffraction peaks of PXRD can be distributed to ZnO (JCPDS # 36-1451) (Fig. 1e) (Dhiman et al. 2022; Zhang et al. 2018). The specific surface area of the ZnOS-x was investigated by nitrogen adsorption desorption isotherms studies, and the corresponding results are shown in Fig. 1f and Table S1. As showed, ZnOS-x displays the standard IV-type isotherm with H2 hysteresis loop (0.8 < P/P0 < 1.0). The BET specific surface areas are calculated to be 17 m2/g for ZnOS-1, 19.2 m2/g for ZnOS-2, 20.2 m2/g for ZnOS-3, 4 m2/g for ZnOS-4, respectively. It can be observed that ZnOS-3 has larger BET specific surface area than other samples, which matches well with the nature of dispersion shown in Fig. 1. It is also interesting to surmise that same quality ZnOS-3 has a higher average H2 evolution reaction rate in the following photocatalytic reactions.
Charge-carrier separation of photocatalyst is pivotal in photocatalytic process. The HRTEM image of ZnOS-3 shown that particles in the range of 20 to 80 nm (Fig. 2a), which indicate that charge-carrier have short paths to transfer from one surface to another, meanwhile, which can also mean that bring high charge-carrier separation efficiency (Dhiman et al. 2022). The high-magnification HRTEM image shows that ZnOS-3 heterostructures have not obvious coupling interface between ZnO and ZnS nanoparticles, meanwhile, which presents the lattice spacing of 0.261 nm can be assigned to the (102) plane of ZnO or the (220) plane of ZnS (Fig. 2b) (Wu et al. 2011; Zhao et al. 2018). Besides, Energy Dispersive Spectroscopy (EDS) for the characterization of the elemental composition within ZnOS-3. Element mapping confirm that O, S, and Zn elements are uniformly distributed on the nanoparticles (Fig. 2d and Figure S1), meanwhile, which further suggest that the phases of ZnO and ZnS are interweaved with each other in nanoscale rather than isolated. In addition, the fringes observed in the selected area electron diffraction (SAED, Fig. 2c) suggest the polycrystalline property of ZnOS nanoparticles.
To further investigate the surface compositions of ZnOS, X-ray photoelectron spectroscopy (XPS) survey spectra with the corresponding Zn 2p, O 1s, and S 2p spectra were studied. Two typical peaks of O components, each of which has the oxygen 2p1/2 and 2p3/2 doublet (Fig. 2e). The first one, located at 529.5 − 530.5 eV, corresponding to lattice oxygen (formation of ZnO), the second deconvoluted peak located at 531.8 − 532.8 eV eV is usually attributed to chemisorbed water on the ZnOS-x surface (Jiang et al. 2022; Zhao et al. 2018). Moreover and interestingly, it can be clearly observed that the binding energies of O 1s have systematically shifted towards low binding energies with 0.3–0.6 eV with an increase the addition of trithiocyanuric acid, which is usually ascribed to change of surface electron density (Jiang et al. 2022). The presence of C 1s peak confirms the residual carbonaceous components in ZnOS-x nanoparticles (Figure S3). In addition, the binding energies at 160.1, 1019.4, and 1044.5 eV are attributed to the S 2p peak, Zn 2p3/2, and Zn 2p5/2 peak (Figure S3), respectively, which indicated that S and Zn components exist mainly in the formation of ZnS on the nanoparticle skeleton (Xu et al. 2018). Besides, the Fourier transform infrared (FTIR) spectra of ZnOS-x. The broad bands at ca. 3423 cm− 1 for ZnOS-x can be assigned to chemisorbed water on the ZnOS-x surface, with a weak δOH-absorption at ca. 1608 cm− 1. The spectrum obtained for ZnO monitors a weak absorption centered at ca. 641 cm− 1 and 502 cm− 1 (Yein et al. 2018). These results indicating the successful preparation of ZnOS heterostructures.
Photoelectrochemical properties. The UV-vis diffuse-reflectance spectra was performed to rationalize light-harvesting ability of ZnOS-3. This result show that the ZnOS-3 possess remarkable light-harvesting ability in the range of 300–800 nm (Figure S4). In order to evaluate the charge-carrier separation efficiency of the photoexcited carrier, photoluminescence (PL) spectra was measured under 360 nm excitation (Figure S5). The weaker PL intensity of ZnOS-3 suggest the electron-hole recombination inhibited in heterostructure. Moreover, the photocurrent response of ZnOS-3 was performed (Figure S6) (Li et al. 2007). The photocurrent − time curve result further illustrate that ZnOS-3 possess efficient charge separation and transfer efficiency under the photoinduce.
Photocatalytic performance and stability. The photocatalytic H2 gerenation activity of ZnOS-x were investigated in the presence ascorbic acid (AA) as sacrificial regent under visible light (> 400 nm, 300 W Xe lamp). As seen from Fig. 3a-3b, ZnOS-3 shows a higher average H2 evolution reaction rate (788.7154 µmol h− 1 g− 1) than ZnOS-1 (243.0574 µmol h− 1 g− 1), ZnOS-2 (668.7388 µmol h− 1 g− 1), and ZnOS-4 (55.4427 µmol h− 1 g− 1). Notably, low agglomeration of ZnOS nanoparticles is beneficial to improve photocatalytic H2 production performance. Additionally, the formed ZnOS-3 herterojunction structure favor interfacial charge transfer and charge-carrier separation between ZnO and ZnS, thereby leading to higher photocatalytic H2 generation activity (Jiang et al. 2022; Huang et al. 2019; Zhao et al. 2018). Meanwhile, the photocatalytic H2 generation activity of ZnOS-3 with different catalyst dosages (5, 10, 15 mg) (Fig. 3c) or different sacrificial reagants (EDTA, triethanolamine (TEOA), and AA) (Fig. 3d) was evaluated at λ ≥ 400 nm. The high photocatalytic H2 generation performance of ZnOS-3 with AA as the sacrificial regent and compared with another sacrificial regents (Piǹa-Pérez et al. 2018). Besides, the photocatalytic H2 evoluation rate of ZnOS-3 increased firstly and then decraese gradually with increasing AA concentration in the reaction system (Fig. 3e).
The photostability of ZnOS-3 nanoparticles was inverstigated by running long-term H2 generation experiments, and the results are depicted in Fig. 4f. The reaction solution was not replaced but adding a fresh AA (0.04375g) in each next cycle. Clearly observed that H2 generation of ZnOS-3 display a slight decrease in 16h photocatalytic reaction through four cycles usage. Additionlly, the PXRD and ICP-MS results of the used ZnOS-3 also suggest no phase change after 16h photocatalytic reaction (Figure S7 and Table S2). The photocatalytic activity and photostability of ZnOS-3 nanoparticles can be attributed to the charge-separation efficiency and interfacial-charge transfer induced by the ZnOS-3 heterostructure (Huang et al, 2019; Jang et al. 2008).