Rare earths leaching from Philippine phosphogypsum using Taguchi method, regression, and artificial neural network analysis

Abstract

The Philippines produce some 2.1–3.2 million t phosphogypsum (PG) per year. PG can contain elevated concentrations of rare earth elements (REEs). In this work, the leaching efficiency of the REEs from Philippine PG with H₂SO₄ was for the first time studied. A total of 18 experimental setups (repeated 3 times each) were conducted to optimize the acid concentration (1–10%), leaching temperature (40–80 °C), leaching time (5–120 min), and solid-to-liquid ratio (1:10–1:2) with the overall goal of maximizing the REE leaching efficiency. Applying different optimizations (Taguchi method, regression analysis and artificial neural network (ANN) analysis), a total REEs leaching efficiency of 71% (La 75%, Ce 72%, Nd 71% and Y 63%) was realized. Our results show the importance of the explanatory variables in the order of acid concentration > temperature > time > solid-to-liquid ratio. Based on the regression models, the REE leaching efficiencies are directly related to the linear combination of acid concentration, temperature, and time. Meanwhile, the ANN recognized the relevance of the solid-to-liquid ratio in the leaching process with an overall R of 0.97379. The proposed ANN model can be used to predict REE leaching efficiencies from PG with reasonable accuracy.