Abstract
This study investigates the relationship between the concentrations of selected trace elements in coal and the oxide composition of ash, treated as a proxy for the mineral matter present in coal. Twenty-eight coal and corresponding ash samples were analysed using ICP-MS. Exploratory data analysis, correlation matrices, and regression models (linear and SVM) were employed to quantify metal-oxide relationships. Bituminous coals exhibited higher concentrations of Ag, Co, and Cu. Strong correlations were identified between Al₂O₃, CaO, and selected critical metals. SVM models achieved high predictive accuracy (R2 > 0.8 for several metals). Sample size limits model generalizability. Variability in coal lithotype and ash composition adds complexity. Findings support selective recovery of Co, Cu, and Mo, aligning with circular economy goals. The results show that ash oxide composition can provide indirect information on trace element distribution in coal and may serve as a useful proxy for assessing element associations with mineral matter.