Machine Learning\nAssisted Wavelength Recognition in\nCu<sub>2</sub>O/Si Self-Powered Photodetector Arrays for Advanced\nImage Sensing Applications

The ability of a photodetector array (PDA) to detect\nmultiple wavelengths\nsignificantly expands its range of potential applications. However,\neffectively detecting and distinguishing between different wavelength\nbands remains a challenge for these arrays. This study introduces\nan approach for wavelength recognition in PDAs by integrating machine\nlearning techniques with solution-processed Cu<sub>2</sub>O/Si heterojunction\nphotodetectors. We propose a simple solution-processing method to\nfabricate a PDA consisting of a 4 × 4 array of p-Cu<sub>2</sub>O/n-Si photodiodes. This method involves low-power UV irradiation\nof a molecular precursor film containing Cu (II) complexes to produce\na p-type Cu<sub>2</sub>O thin film on a Si substrate. A UV-shielding\nglass plate is used as a patterning mask, and water is used to wash\naway the UV-shielded areas. Using machine learning techniques, we\neffectively classify various wavelengths of light, including UV, visible,\nand near-infrared, and accurately predict their corresponding photocurrents\nin the Cu<sub>2</sub>O/Si heterojunction. Notably, the PDA enables\nclear identification of images across different light wavelengths.\nThis PDA paves the way for advanced applications in multispectral\nimaging and sensing technologies.