Determining major factors controlling phosphorus removal by promising adsorbents used for lake restoration: A linear mixed model approach.

Phosphorus (P) removal from lake/drainage waters by novel adsorbents may be affected by competitive substances naturally present in the aqueous media. Up to date, the effect of interfering substances has been studied basically on simple matrices (single-factor effects) or by applying basic statistical approaches when using natural lake water. In this study, we determined major factors controlling P removal efficiency in 20 aquatic ecosystems in the southeast Spain by using linear mixed models (LMMs). Two non-magnetic -CFH-12 and Phoslock- and two magnetic materials -hydrous lanthanum oxide loaded silica-coated magnetite (Fe-Si-La) and commercial zero-valent iron particles (FeHQ)- were tested to remove P at two adsorbent dosages. Results showed that the type of adsorbent, the adsorbent dosage and color of water (indicative of humic substances) are major factors controlling P removal efficiency. Differences in physico-chemical properties (i.e. surface charge or specific surface), composition and structure explain differences in maximum P adsorption capacity and performance of the adsorbents when competitive ions are present. The highest P removal efficiency, independently on whether the adsorbent dosage was low or high, were 85-100% for Phoslock and CFH-12, 70-100% for Fe-Si-La and 0-15% for FeHQ. The low dosage of FeHQ, compared to previous studies, explained its low P removal efficiency. Although non-magnetic materials were the most efficient, magnetic adsorbents (especially Fe-Si-La) could be proposed for P removal as they can be recovered along with P and be reused, potentially making them more profitable in a long-term period.