Underestimation of phosphorus fraction change in the supernatant after phosphorus adsorption onto iron oxides and iron oxide–natural organic matter complexes

The phosphorus (P) fraction distribution and formation mechanism in the supernatant after P adsorption onto iron oxides and iron oxide–humic acid (HA) complexes were analyzed using the ultrafiltration method in this study. With an initial P concentration of 20 mg/L (I = 0.01 mol/L and pH = 7), it was shown that the colloid (1 kDa–0.45 μm) component of P accounted for 10.6%, 11.6%, 6.5%, and 4.0% of remaining total P concentration in the supernatant after P adsorption onto ferrihydrite (FH), goethite (GE), ferrihydrite–humic acid complex (FH–HA), goethite–humic acid complex (GE–HA), respectively. The < 1 kDa component of P was still the predominant fraction in the supernatant, and underestimated colloidal P accounted for 2.2%, 55.1%, 45.5%, and 38.7% of P adsorption onto the solid surface of FH, FH–HA, GE and GE–HA, respectively. Thus, the colloid P could not be neglected. Notably, it could be interpreted that Fe^3 + hydrolysis from the adsorbents followed by the formation of colloidal hydrous ferric oxide aggregates was the main mechanism for the formation of the colloid P in the supernatant. And colloidal adsorbent particles co-existing in the supernatant were another important reason for it. Additionally, dissolve organic matter dissolved from iron oxide–HA complexes could occupy large adsorption sites of colloidal iron causing less colloid P in the supernatant. Ultimately, we believe that the findings can provide a new way to deeply interpret the geochemical cycling of P, even when considering other contaminants such as organic pollutants, heavy metal ions, and arsenate at the sediment/soil–water interface in the real environment.