Modeling the efficiency of TiO2 photocatalytic degradation of MTBE in contaminated water: a support vector regression approach

Methyl tert-Butyl Ether (MTBE) is one of the common environmental contaminants that pose severe risks to human health and the environment. Due to the detrimental effects of MTBE, many efforts have been devoted towards decontamination of wastewater containing MTBE. Most of the reports on the treatment of MTBE are experimental based, which requires studying the effect of several factors such as dosage of the photocatalyst, pH of the medium, the initial concentration of contaminant and the contact time on degradation efficiency in MTBE contaminated waters. Although, this approach is highly dependable and often leads to new insights. However, because the degradation efficiency is influenced by multiple factors, performing experiments to investigate the effect of these factors often increases the experimental burden, thus requiring more time and materials consumption to achieve desirable results. Herein, we propose a computational intelligent strategy to mitigate these challenges. In this contribution, the degradation efficiencies of MTBE in the presence of TiO2/UV were modeled under various experimental conditions using the support vector regression model. The model was built using experimental data comprising of inputs such as TiO2 dose, initial MTBE concentration, UV wavelength and contact time. Remarkably, the developed model exhibits significant accuracy as determined from the values of correlation coefficient (98.27%) and root means square error (5.53). In addition, it was determined that the achievable optimum conditions for degradation of 0.5-100 ppm MTBE-contaminated water were 2.4 g/L of TiO2 dose with UV radiation, a solution pH of 3 and treatment time of 2 h. This study will be useful in the experimental design of treatment for MTBE, consequently reducing the time spent on running experiments and at the same time ensuring efficient use of resources for treatment of MTBE contaminated water.