Optimization of the Nitroxide Mediated Copolymerization of Styrene and Alpha-Methyl Styrene

FORTUNATTI, C.; BRANDOLIN, A.; SARMORIA, C.; ASTEASUAIN, M.

Pittsburgh, PA

Congreso; 2012 AIChE Annual Meeting; 2012

AIChE - American Institute of Chemical Engineers

Free radical polymerization is widely used in industry. However, due to the high reactivity of the polymer radicals it is very difficult to control the molecular structure of the polymer. As an alternative, controlled radical polymerization (CRP) has been developed since the mid-90s, offering enormous possibilities for creating materials with unique properties. CRP is based on the idea of establishing a balance between active and inactive chains. The interactions between living growing chains are limited by the addition of an agent that can deactivate them reversibly. Thus, in a CRP the total effect of termination is reduced because there is a very small number of active chains. CRP allows obtaining (co) polymers with controlled molecular weight and polydispersity, as well as composition and chain architecture. In this way manufacturers are enabled to improve product properties currently on the market or create polymers with unique properties (Matyjaszewski and Spanswick, 2005). One of the most studied mechanisms of CRP is the nitroxide mediated polymerization (NMP), in which the reaction is controlled through a reversible termination with a stable radical. The propagation and termination rates are reduced by the balance between active and inactive species, which allows all chains to be initiated and grow at the same speed on average. Copolymers may be characterized by chain length, composition and sequence length distributions. In previous works we presented a mathematical model of the NMP of styrene and alpha-methyl styrene in tubular and semibatch reactors (Asteasuain et al., 2011; Fortunatti et al., 2012). This model predicted average molecular properties, such as average molecular weights and composition, as well as the full bivariate MWD of the copolymer. Average properties were modeled using double index moments and the bivariate MWD by means of the 2-D probability generating function technique. In this work, this model is used for the optimization of the process aiming at obtaining molecular weight distributions and other molecular properties tailored according to the requirements of the different product applications. The optimization problem was formulated and solved in gPROMS (Process Systems Enterprise, Ltd.), in order to determine optimum design conditions of the process to minimize the difference between the desired molecular properties and the ones actually obtained in the process. The optimization space included reactor type (semibatch or tubular), reactor design and operating policies (initial charge, feeding policy, temperature profile).