Selective Detection of SO₂ from Mixed SF₆ Decomposition Products via Au/ZnO Nanorods with (101−0) Plane Exposure

SF6 decomposition composition analysis (DCA) is an effective method for identifying partial discharge (PD) fault types in gas-insulated switchgear (GIS). Among various SF6 decomposition byproducts, the SO2 concentration exhibits a strong correlation with PD types, making it a key indicator for PD detection. However, detecting SO2 in an oxygen-free SF6 environment using semiconductor sensors presents challenges, such as low response values, slow recovery, and poor repeatability. To address these limitations, three ZnO micro/nanomaterials are synthesized via the hydrothermal method in this work. The ZnO nanowire sensor demonstrates optimal selectivity for 50 ppm of SO2 (S = 9.18), indicating that reduced grain size and increased exposure of the (101−0) facet significantly enhance SO2 sensing performance. Furthermore, loading 2 at.% Au nanoparticles onto ZnO nanorods increases the response value to 27.43, while simultaneously reducing the response and recovery times to 21.86 and 28.38 s, respectively. In situ diffuse reflectance infrared Fourier transform (DRIFT) experiments reveal that SO2 molecules directly exchange electrons with Au/ZnO material in the oxygen-free SF6 background, independent of surface-adsorbed oxygen. The doping of Au nanoparticles facilitates electron exchange and accelerates the SO2 desorption processes. To assess the real-time detection capability of the proposed Au/ZnO sensor, an SF6 partial discharge simulation platform is established, and discharge experiments are conducted at 21.60 and 24.00 kV for 2 h. The SO2 concentration detected by the Au/ZnO sensor deviates by less than 3.62 ppm from gas chromatography (GC) results, demonstrating its strong potential for real-time SO2 detection in GIS equipment.