Search

Two lab-scale biofilters packed with manganese ore sand and quartz sand were constructed to reveal the behavior in removing iron and manganese during the start-up period. Meanwhile, the removal mechanism of the two sands was also investigated by means of EDS, XPS, and SEM. With the influent iron (2-3 mg·L -1 ) and manganese (0.3-0.6 mg·L -1 ), the start-up operational results indicated that the quartz sand biofilter needed 15 and 30 d to achieve the removal of iron and manganese, respectively. The manganese ore sand only required 10 d to remove iron, while the effluent manganese was always below of 0.1 mg·L -1 . The results confirmed that the natural iron and manganese oxides coated on the manganese ore sand surface could explain its better removal behavior as compared to quartz sand. However, the generated iron oxide could also act as the adsorbent and catalyst like natural iron oxide, only when iron removal occurred in the quartz sand biofilter. The final product of iron removal was a complex consisting of divalent and trivalent iron, with a specific value of 1:1.44-1:1.54. Moreover, during the start-up period, manganese ore sand transformed manganese from divalent to trivalent by the catalytic effect, while the latter tended to be converted to the quadrivalent state under the bioactivity. The quartz sand could adsorb manganese but easily became saturated, and then the removal was dominated by bioactivity. The product generated by the manganese removal process was also a complex with the three valences. Moreover, the two complexes could coat onto the surface of the sands, but most of the iron complex was easily washed out of the filtering layer. Conversely, the manganese complex tended to coat onto the manganese ore sand surface or accumulate between the pores of quartz sand.

In order to remove trace amounts of phosphorus from water bodies, a lab-scale biofilter was constructed to investigate the capacity of in situ oxidation products of iron or manganese for phosphorus adsorption. SEM, EDS, BET, and zeta technologies were employed to reveal the adsorption mechanisms. The results indicated that phosphorus could be removed by the oxide products generated from the iron or manganese removal process, at 106.28 μg·mg -1 and 77.98 μg·mg -1 , respectively, as shown by the linear relationships between phosphorus removal and the two oxides. SEM, EDS, and BET analysis demonstrated that the BET specific surface areas for the iron- and manganese-rich oxides were 96 m 2 ·g -1 and 67 m 2 ·g -1 , respectively, with the former accumulated between the pore spaces of the filtering sand and easily washed out of the layer by backwashing, whereas the latter coated the surface of the filtering sand. Thus, backwashing was favorable for phosphorus adsorption in the iron oxidation process to avoid overaccumulation. Moreover, the zero point of charge of the two oxides indicated electrostatic attraction may have occurred between iron-rich oxide and phosphorus; however, inner-sphere complex reactions obviously occurred for the two oxides because the zero point of charge after phosphorus adsorption decreased to a lower level. In addition, other anions were negatively complexed with the phosphorus on the surface of the oxides, it demonstrated that phosphorus adsorption on the surface of the two oxides seemed to be a specific adsorption.