High-performance triple-band ultraviolet photodetector based on mixed-phase InGaO thin film for material identification

The wavelength-dependent nature of material reflectance in the ultraviolet (UV) range makes multi-band UV response analysis an effective technique for material identification. In this work, a triple-band UV photodetector was constructed on the In2O3-Ga2O3 (InGaO) mixed-phase thin films. By incorporating In into Ga2O3 via plasma-enhanced chemical vapor deposition method, the bandgap of the mixed-phase InGaO thin films was modulated to 4.64 eV, extending the spectral response range to cover UVA (365 nm), UVB (295 nm), and UVC (255 nm) bands. This broadband photoresponse enables quantitative analysis of material-specific UV reflectance signatures, offering a reliable basis for material differentiation. Integrating a 5 × 4 photodetector array with convolutional neural networks, a precise identification and an accurate imaging of geometrically identical materials (Al, Ag, Si, and Au) were realized. For Al, Ag, and Si, this system achieves an overall accuracy exceeding 95%. This work proposes an efficient strategy for nondestructive surface inspection and material recognition by combining multi-spectral sensing with intelligent data processing.