Production of thymol from alkylation of m-cresol with isopropanol over ZSM-5 catalysts: Artificial Neural Network (ANN) modelling

Abstract

Thymol is one of the most widely used chemicals in the cosmetic and health industries and a known antioxidant. There are different methods for the preparation of thymol. One of the efficient methods for the synthesis of thymol is the alkylation of m-cresol using isopropanol as the alkylating agent and ZSM-5 acid catalysts. Different parameters such as SiO2/Al2O3 ratio, temperature and Weight Hourly Space Velocity (WHSV) can affect the conversion and selectivity of m-cresol alkylation. The application of tools to predict the effect of each of the effective factors and, thus, the proposition of a strong model of reaction conditions will lead to maximum conversion and selectivity. A connectionist tool has been used to evaluate the conversion and selectivity in thymol synthesis. The model simultaneously correlates the thymol selectivity and m-cresol conversion to independent variables. An appropriate statistical analysis has been carried out in order to guarantee the generalization and strength of the deterministic model. The developed network leads to coefficient of determination (R2) of 0.9890 and 0.9889 for the thymol selectivity and m-cresol conversion, respectively. Furthermore, the error analysis shows mean squared error (MSE) values of 2.2060 and 1.8126 for the thymol selectivity and m-cresol conversion, respectively.