Removal of dissolved natural organic matter by the Al/Fe pillared clay-activated-catalytic wet peroxide oxidation: Statistical multi-response optimization

Abstract The removal of dissolved Natural Organic Matter (NOM) using the Fenton-like Catalytic Wet Peroxide Oxidation (CWPO) activated by an Al/Fe-pillared clay catalyst (Al/Fe-PILC) was optimized in a semi-batch reactor. The Al/Fe-PILC was prepared from concentrated precursors and fully characterized by standard physicochemical techniques (DRIFTS, powder XRD, XRF, CEC, N2-adsorption, and TGA/DSC analyses). NOM mineralization (removal of dissolved organic carbon – DOC), color removal, and the fraction of H2O2 consumed were all successfully maximized by a multi-response, statistical Desirability function. Circumneutral pH, starting NOM concentration, and temperature were all assessed as non-controllable variables (covariates) within ranges typical of real surface water. Found overall optimal conditions of reaction were 0.45 mg H2O2/mg DOC (∼3.2 mg/dm3 for 7.0 mg DOC/dm3) and 5.97 g/dm3 of the Al/Fe–PILC clay catalyst, that allowed obtaining up to 75 % of DOC mineralization and 86 % of color removal from a synthetic surrogate of water; meanwhile, 37 % of DOC mineralization and 93 % of color removal were achieved when applied on real, raw surface water. Significant concentration of scavenging inorganic anions (mainly carbonates, sulfates, and chlorides) in surface water samples showed to deplete the NOM mineralization. Negligible loss of catalytic response (∼ 2.0 % in terms of NOM mineralization), iron leaching, or blocking of micropores were recorded in the catalyst over reuse through five sequential catalytic cycles. DRIFTS, N2-adsorption, and TGA/DSC revealed accumulation of high-molecular-weight fractions of NOM on the external surface, but the microporous content, where active Fe gets preferentially stabilized in Al/Fe-PILCs, remained pretty stable.