Understanding the fouling characteristics in ultrafiltration membrane for marine algae-laden seawater pretreatment: Focus on the role of algal extracellular organic matter

The increasing frequency and extent of harmful algal blooms (HABs) due to global warming and environmental pollution pose significant challenges to the ultrafiltration (UF)-seawater reverse osmosis (SWRO) desalination process. The UF process is a dominant pretreatment method to reduce HAB fouling potential in subsequent SWRO processes. Herein, we systematically evaluate the UF membrane fouling characteristics at different growth phases of HAB algae in terms of reversible/irreversible fouling and pinpoint the fouling mechanisms on UF membranes in the real seawater matrix. We employed Prorocentrum donghaiense as a model marine HAB specie to characterize algal derivatives, including extracellular organic matter (EOM) and granule particles. LC-OCD, MFIs, 3D-EEM, and GPC were used to analyze algae-derived matters. Results from the multiple fouling model revealed that algal solution in both stationary and decline phases exhibit severe membrane fouling, characterized by rapid cake layer formation, swift flux decline, and high membrane resistance. Soluble EOM (S-EOM) was the primary contributor to membrane fouling, linked to its elevated concentrations of dissolved organic matters, polysaccharides, proteins, and humic substances, which are predominantly hydrophobic. Additionally, granule particles mitigated organic membrane fouling, as evidenced by the smaller permeate volume required for S-EOM cake layer formation (<5 mL), resulting in a faster initial flux decline. These results revealed a shift in released organic matter from protein-centric to humic substance-centric during the transition from exponential growth to the decline phase, with correlation analysis further underscoring the substantial contribution of humic substances to membrane resistance. This research highlights the distinct UF membrane fouling characteristics at different marine algal growth stages, offering theoretical support for predicting and managing HABs.