Sensitivity analysis and multi-objective optimization of gas-phase polymerization of propylene using Ziegler-Natta catalysts

Abstract

Gas-phase polymerization of propylene via 4th generation Ziegler-Natta (ZN) catalysts was studied and optimized using intelligent data-based methods. The effect of co-catalyst (triethylaluminium (TEA)) and external donor (ED) compositions on the activity and PP-isotacticity for two commercial MgCl2/TiCl4/phthalate(ID)/TEA/silane(ED) catalyst systems with different Ti-contents were investigated using design-of-experiment (DOE) combined with response-surface-methodology (RSM). Sobol’s sensitivity analysis was employed to determine the quantitative impact of catalyst compositions on the target variables. Both catalysts performance was optimized via multi-objective genetic algorithm (GA) to maximize the activity and PP-isotacticity. Results revealed that the activity and PP-isotacticity for both catalysts will be enhanced at low co-catalyst and high ED amounts, respectively. Increasing the ED improved the activity of low-Ti catalyst, whereas it reduced the activity of high-Ti one. Sensitivity results for both catalysts depicted the comparable influence of co-catalyst and ED on the activity and the strong influence of ED on PP-isotacticity. GA optimization offered the activity of 3.7 – 4.65 gPP/mgCat and PP-isotacticity of 95.8 – 97.5 % for low-Ti catalyst at the optimal TEA/Catalyst = 7 – 7.2 and TEA/ED = 6.0 – 7.2 and the activity of 3.9 – 4.7 gPP/mgCat and PP-isotacticity of 93 % for the high-Ti one at the optimal TEA/Catalyst = 7 – 11.2 and TEA/ED = 12.