Removal of a quaternary ammonium compound by electrocoagulation: Mechanistic analysis and multi-response optimization...

The widespread use of quaternary ammonium compounds (QACs), intensified by the COVID-19 pandemic, has led to their increasing presence in aquatic environments, thereby demanding effective treatment strategies for shock loads from industrial discharges. This study presents a hybrid modeling and multi-response optimization (MRO) framework to optimize the electrocoagulation (EC) process for removing cetyltrimethylammonium bromide (CTAB) as a model QAC. A central composite design was employed for efficient experimentation and data collection regarding the effects of four variables on CTAB removal efficiency, energy consumption, and electrode consumption. While the response surface methodology (RSM) model yielded the best prediction for energy demand, a machine learning (ML)-based linear regression (LR) model showed superior accuracy for CTAB removal and electrode consumption. Optimization with Pyomo (compared with RSM, PyTorch, and Scikit-learn) proved most effective, yielding a Pareto-optimal solution (100 mg L^-1 CTAB, 1.5 A, 2.5 g L^-1 Na₂SO₄, and 15 min) that achieved a 94.95 % removal with energy and electrode consumptions of 26.33 kWh kg^-1 and 12.28 kg Fe kg^-1 CTAB removed, respectively. This optimum reduced the treatment cost by 38.5 % (to 5.59 USD per kg of CTAB removed) compared with the RSM-derived solution. Mechanistic studies involving kinetics, sludge characterization, and density functional theory (DFT) calculations revealed that CTAB removal proceeds through the electrostatic adsorption of CTA⁺ onto in-situ generated Fe-Al hydroxide flocs via electrostatic interaction, followed by sweep flocculation. DFT analysis further localized the reactive site on the quaternary ammonium head. The developed ML-MRO framework provides a scalable strategy for designing intelligent, cost-effective electrochemical systems to mitigate emerging contaminants.