Leaching state prediction for battery black mass using online sensor data

Abstract

This study investigates the leaching behavior of metal elements from pyrolyzed battery black mass using a combination of experimental and data analysis methods. A linear regression model, utilizing pH, conductivity, and temperature as input features, effectively predicted leaching states with an average prediction error below 1.1%. Compared to calculating leaching states using the Arrhenius equation and traditional kinetic models, the proposed approach demonstrated improved generalizability and accuracy. Kinetic analysis using the shrinking core model indicated that the leaching of nickel and cobalt is primarily controlled by diffusion after an initially chemically controlled reaction phase. This diffusion-controlled mechanism explains the observed strong linear relationship between sensor data and the leaching states of these metals. The calculated activation energies for nickel and cobalt are 29.8 kJ mol⁻¹ and 22.6 kJ mol⁻¹, respectively. The rapid leaching kinetics of lithium and manganese are attributed to their unique physiochemical properties, likely influenced by the thermal treatment process.