Development of RSM Statistical Model for Methanol Carbonylation Rate for Acetic Acid Synthesis by Using Cativa TM Technology

Abstract

Abstract In recently developed CativaTM process, acetic acid is produced by methanol carbonylation reaction in which a complex interaction among all the major reaction species including iridium, ruthenium, methyl acetate, methyl iodide and water are observed. In this study, a statistical technique of response surface method (RSM) (which is called historical data design algorithm) is applied to investigate the concentration effects of these species on the carbonylation reaction rate. A quartic equation is fitted to the experimental data, and its suitability is examined by several statistical tests. Lack of fit, model F-value, coefficient of determination (R2), adjusted R-squared (Adj.R2) and predicted R-squared (Pre. R2) are respectively equal to 1.65, 182.73, 0.9822, 0.9768 and 0.9263. The validation of the proposed model is investigated by numerical optimization of the design-expert software. The desirability value of the model prediction is 0.94 that indicates the high agreement between the model prediction and the experimental results. The individual and binary effects of the considered parameters on the carbonylation rate are also investigated through the developed model. The steep slope/ curvature of Ir, Ru and water concentrations in perturbation plot indicates the high sensitivity of carbonylation rate to these species. The interaction effects of Ru and water on carbonylation rate show that at water concentration of 7 w/w %, a steep increase occurs in the reaction rate by increasing Ru promoter. Investigating the simultaneous effects of Ru and Ir concentrations on the carbonylation rate indicates that the increase of Ru to Ir molar ratio promotes the reaction rate by enhancing the lability of [Ir(CO)2I3Me]- complex and preventing the production of inactive species of [Ir(CO)2I4]- in the catalytic cycle.