A Bias-Tunable Dual Band Photodetector for Plastic Material classification

This work introduces a novel technique for efficient and accurate classification of plastic materials in recycling process. The approach is based on a voltage-tunable Germanium on Silicon photodetector that operates within the 400 nm - 1600 nm spectral range. Utilizing a lab bench setup with a broad-spectrum light source and standard components, material properties are analyzed through photocurrent measurements. By employing the K-Nearest Neighbors algorithm, the analysis achieved a high accuracy of 94.6% for classifying 6 different types of plastics. This work underscores the critical role of advanced photo detection techniques in enhancing the efficiency and accuracy of classification for plastic recycling, contributing to more effective waste management and environmental sustainability.