Laser recycling of silver in bulk and nanoparticle form from silicon solar cells and deep learning for process automation

Abstract

This study advances the green recycling of silver from silicon solar cells by selectively removing silver from electrical contact lines through laser ablation. The laser ablation process, conducted in the air and in the water medium, provided microparticles and higher-value silver nanoparticles, respectively. Optical microscopy and energy dispersive X-ray spectroscopy (EDS) analysis confirmed the successful removal and recovery of silver. A basic understanding of laser removal of Ag is provided. Comprehensive characterization revealed the nanoparticles' size, shape, and elemental composition, with optimized laser parameters achieving 93 % purity by weight, with the remaining 7 % primarily silicon. Additionally, convolutional neural networks (CNNs) trained with TensorFlow accurately detected silver lines on broken silicon solar cells. A comprehensive training dataset enabled high accuracy across diverse geometries and conditions, with validation confirming real-world applicability. Integrating CNN models with laser ablation automated silver recovery processes, enhancing efficiency and sustainability in photovoltaic recycling. A preliminary cost analysis highlights the process's cost-effectiveness and potential for recycling other materials. This demonstrates the efficacy of laser ablation as a sustainable method for selective silver removal.