Dicamba Degradation Using a Low-Cost Chlorine/Ferrous-Based AOP: ANN-PSO Model Development, Intermediate Identification, and Toxicity Assessment Using Microalgae

Abstract

In this study, the degradation of Dicamba methyl ester (DME) was investigated using a low-cost chlorine/ferrous process. The degradation yield was determined by examining the influence of several factors, such as NaClO concentration, dicamba concentration, FeSO4 catalyst mass, and initial solution pH, over a 15 min period. To determine the optimal conditions for DME degradation, an artificial neural network (ANN) model with 4-5-1 architecture was developed. The particle swarm optimization (PSO) algorithm was then utilized in conjunction with the ANN model to identify the optimal factor levels predicted yield of 88%. The following optimal conditions were identified: [NaClO] = 422.3 μM, [Dicamba] = 4.4 mg/L, [FeSO4] = 9.5 mg/L, and a pH of 2.56. GC/MS analysis was conducted to identify the byproducts that were generated during DME degradation. Benzene, 1,2,4-trichloro-3-methoxy DME-BP (m/z 210) was the only identified byproduct that contained chlorine in its structure. A proposed reaction pathway for the DME degradation was suggested based on the obtained mass spectra. In the final stage of the study, total organic carbon (TOC) removal was analyzed using a Fenton-like process under optimized conditions for a duration of 195 min. To confirm the effectiveness of DME and its byproduct degradation, toxicity assessments were performed using the Chlorella vulgaris microalgae as a model organism. The results indicated a low toxicity of 20% when the DME mineralization reached 62.52%. These findings provide strong evidence that support the effectiveness of the proposed low-cost system for DME removal.