Abstract
[Display omitted]
•The use of the Fe3+-Oxalate complex under UV light was studied.•This study is done with an unadjusted pH, with a result, in the end, of a pH final to neutrality.•The using an approach progressive and gradual of the oxalate.•An artificial neural network optimized and modeled is used to simulate the degradation and mineralization of amoxicillin.
In this work, the Ferrioxalate complex was used to mineralize the widely used antibiotic amoxicillin (AMX). The effect of operational parameters such as Fe3+ concentration, the molar ratio (Oxalate/Fe3+), and initial pH was studied for achieving high efficiency of degradation and mineralization. The effect of inorganic ions in the degradation of AMX was also investigated. Optimization of an Artificial Neural Network (ANN) to predict the performance of the process was realized. Neural networks were applied to model AMX degradation using 75 experimental data. Analysis of the results shows that the oxalate: iron ratio is limited to 3 with a Fe3+ concentration of 0.35 mM. In order to avoid this limitation, an alternative solution was adopted to increase the molar ratio with the gradual and progressive use of oxalic acid ligand. Almost total degradation and mineralization of AMX was achieved at the free initial pH of 2.8 and final pH of 6. The study of the effect of inorganic ions showed that bicarbonate and sulfate ions play an important role in contrast to Cl− ions with no noticeable influence. Neural network application resulted in a MSE of 1.1543 × 10-5 and a correlation coefficient r of 0.99 for AMX degradation and MSE of 1.12162 × 10-5 with r of 0.99 for mineralization. The model can describe successfully the percentage of degradation and mineralization of AMX under various conditions. ANN and NSGA-II (Non-dominant Sorting Genetic Algorithm -II) hybrid method to solve a multi-objective problem was proposed to determine the optimal (Fe3+/Ox/UVA) operating and process parameters.