Machine learning exploration of the direct and indirect roles of Fe impregnation on Cr(VI) removal by engineered biochar

• Cr(VI) removal by iron-biochar (Fe-BC) was evaluated machine learning. Direct impact of impregnated-Fe content on limited. Indirect impacts Fe were critical due to variable Fe-BC properties. Impregnated-Fe improved SOFGs contents for removal. C−O more than C=O Data mining and knowledge discovery learning (ML) have recently come into application in environmental remediation, especially the exploration multifactorial process such as hexavalent chromium [Cr(VI)] composite (Fe-BC). The capacity concurrently controlled impregnated iron species (Fe 0 /Fe 2+ 3+ ), carbon properties, iron-carbon interactions, while current lab-scale research could hardly untangle overall relationships with experiments one or several Fe-BCs under different frameworks. Herein, we investigated various microscopic material properties aqueous ML approach highlighted variations biochar after impregnation. Our results suggested that direct impregnated-iron limited, possibly related undistinguished models, which roles might be counteracted. Instead, embodied indirectly altering Surface oxygen-containing functional groups (SOFGs) played a pivotal role according models. condensed polyaromatic matrices BC high non-polar also proved conducive models developed this study consider surface functionalities information offer accurate prediction removal, behind can act vital reference rational design engineered remove Cr(VI).