Polydopamine-incorporated MOF membrane with hydrophilicity for effective oil/water separation and fouling-resistant model analysis and prediction

• Published in: Journal of molecular liquids 2024-12, Vol.415, p.126432, Article 126432
• Main Authors: Wang, Lu, Cao, Langyuan, Fan, Jianhua, Li, Zonghao, Han, Zhiwu, Liu, Dan
• Format: Article
• Language: English
• Subjects: Membrane fouling; MOF and MOF@PDA membrane; Oil–water separation; Predicting model; XDLVO theory
• ISSN: 0167-7322
• DOI: 10.1016/j.molliq.2024.126432

[Display omitted] •Prepared MOF@PDA membranes exhibit excellent oil–water separation performance.•Membrane fouling mechanisms are analyzed using the XDLVO theory.•The molecular modes of interaction with oil–water emulsions by membrane are compared.•The optimal machine learning model for oil–water separation efficiency was developed.•Prepared MOF@PDA membrane presents prospects in the remediation of oily wastewater. A large number of oil spills and oily wastewater are discharged, in the modern industrial production process, resulting in serious water pollution. Oily wastewater can be extremely harmful to ecosystems and human health. Oil-water separation has become a major challenge at present, and membrane separation has aroused more and more concern in recent years due to its high economic efficiency. This paper fabricated PES filtration membranes using the novel nanoparticlesbased on both metal–organic frameworks-5 (MOF-5) and polydopamine (PDA) layers as dopants, and then adequately explored the porosity, morphology, separation, hydrophilicity, and fouling-resistant performance of the resultant membranes. In addition, the prepared membranes were used for oil–water separation, including soybean-in-water, petroleum ether-in-water, and gasoline-in-water emulsions. MOF@PDA membranes exhibit high separation efficiency of up to 99.8%. Subsequently, membrane fouling mechanisms were investigated using the Derjaguin-Landau-Verwey-Overbeek (XDLVO) theory, and the molecular mode of interaction with the three oil–water emulsions with the membrane was compared. Newly prepared MOF@PDA membranes have better stability, fouling-resistant, and self-cleaning properties. Finally, an optimal machine learning model for oil–water separation efficiency was developed with a high prediction accuracy of 98%. The obtained results indicate that mixed matrix membranes exhibit excellent oil–water separation performance, demonstrating great application prospects in the remediation of oily wastewater.