Biosurfactant-driven desorption and remediation of heavy oil contaminated soils underpinned by molecular simulations and microbial dynamics.

This study integrates molecular dynamics simulations and bench-scale experiments to investigate the adsorption and desorption behaviors of heavy oil on five mineral substrates: SiO2, kaolinite, muscovite, and Ca2+-/Na+-montmorillonite. Adsorption followed Langmuir isotherms, with montmorillonite exhibiting the highest capacities (0.061-0.062 molecules per Å2 for aromatics in simulations; 0.086-0.091 g g-1 in bench-scale tests) and SiO2 the lowest (0.027 pcs per Å2; 0.013 g g-1). Among four biosurfactants evaluated-rhamnolipid, sophorolipid, trehalose lipid, and mannosylerythritol lipid-sophorolipid consistently achieved the greatest desorption efficiency, removing up to 99.63% of adsorbed oil from Na+-montmorillonite and 96.04% from field-contaminated soil. 16S rRNA and metagenomic sequencing revealed an increased abundance of hydrocarbon-degrading bacteria within the soil microbial community, highlighting a synergistic effect between biosurfactant-induced desorption and biodegradation. These findings underscore the critical roles of mineralogical properties, oil fraction characteristics, and biosurfactant selection in soil washing treatment. This work presents a viable and eco-friendly strategy for remediating crude oil-contaminated soils, with important implications for optimizing large-scale environmental restoration efforts.