Experimental study-based hybrid SVR-GWO modeling of copper sustainable reclamation from spent printed circuit boards (PCBs) induced by leaching followed by liquid–liquid extraction

Artikel i vetenskaplig tidskrift, 2024

Reclamation of copper from waste printed circuit boards (WPCBs) is critical in advancing eco-friendly manufacturing methods by considering viable secondary metal resources. Herein, this study introduced a novel hybrid intelligence model that relies on a support vector regression–grey wolf optimization (SVR-GWO) approach to predict, validate, and optimize the leaching of WPCBs in nitrate solution and copper purification using LIX 973 N. The hybrid model's performance surpassed that of the standalone SVR model due to fine-tuning hyperparameters using the GWO approach, as indicated by the lower values of MSE and narrow error distribution in the leaching, extraction, and stripping experiments. The modeling data indicated that 96.1 % of the copper in the WPCB material was dissolved during leaching at 75°C, using a pulp density of 7 % for 2 h. During the extraction phase, the hybrid model optimized the structure performance of equilibrium pH, extractant concentration, contact time, and O/A ratio, resulting in values of 2.5, 30 %, 20 min, and 1, respectively. The predicted isotherm data for the McCabe-Thiele diagram, derived from the developed model, suggested four operational stages for extracting copper from the leach solution. Under optimized conditions of 2 M H2SO4 and 0.3 A/O phase ratio at 25 °C for 20 min, the complete stripping process from the loaded organic phase required applying three counter-current stages. These developments highlight the capability of the SVR-GWO method to improve copper extraction from WPCBs, thereby making a substantial contribution to sustainable recycling efforts.