Nonionic surfactant Tween 80-facilitated bacterial transport in porous media: A nonmonotonic concentration-dependent performance, mechanism, and machine learning prediction
Dong Zhang,Jiacheng Jiang,Huading Shi,Li Lu,Ming Zhang,Jun Lin,Ting Lü,Jingang Huang,Zhishun Zhong,Hongting Zhao
DOI: https://doi.org/10.1016/j.envres.2024.118670
IF: 8.3
2024-06-01
Environmental Research
Abstract:The surfactant-enhanced bioremediation (SEBR) of organic-contaminated soil is a promising soil remediation technology, in which surfactants not only mobilize pollutants, but also alter the mobility of bacteria. However, the bacterial response and underlying mechanisms remain unclear. In this study, the effects and mechanisms of action of a selected nonionic surfactant (Tween 80) on Pseudomonas aeruginosa transport in soil and quartz sand were investigated. The results showed that bacterial migration in both quartz sand and soil was significantly enhanced with increasing Tween 80 concentration, and the greatest migration occurred at a critical micelle concentration (CMC) of 4 for quartz sand and 30 for soil, with increases of 185.2% and 27.3%, respectively. The experimental results and theoretical analysis indicated that Tween 80-facilitated bacterial migration could be mainly attributed to competition for soil/sand surface sorption sites between Tween 80 and bacteria. The prior sorption of Tween 80 onto sand/soil could diminish the available sorption sites for P. aeruginosa, resulting in significant decreases in deposition parameters (70.8% and 33.3% decrease in K<sub>D</sub> in sand and soil systems, respectively), thereby increasing bacterial transport. In the bacterial post-sorption scenario, the subsequent injection of Tween 80 washed out 69.8% of the bacteria retained in the quartz sand owing to the competition of Tween 80 with pre-sorbed bacteria, as compared with almost no bacteria being eluted by NaCl solution. Several machine learning models have been employed to predict Tween 80-faciliated bacterial transport. The results showed that back-propagation neural network (BPNN)-based machine learning could predict the transport of P. aeruginosa through quartz sand with Tween 80 in-sample (2 CMC) and out-of-sample (10 CMC) with errors of 0.79% and 3.77%, respectively. This study sheds light on the full understanding of SEBR from the viewpoint of degrader facilitation.
environmental sciences,public, environmental & occupational health