Full text
| |
| Author(s): |
Andre Rodriguez Murari
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | Campinas, SP. |
| Institution: | Universidade Estadual de Campinas (UNICAMP). Faculdade de Engenharia Química |
| Defense date: | 2007-12-18 |
| Examining board members: |
Liliane Maria Ferrareso Lona;
Reinaldo Giudici;
Reginaldo Guirardello
|
| Advisor: | Liliane Maria Ferrareso Lona |
| Abstract | |
Living Free-Radical Polymerization (LFRP) has been target of much attention in the last years, because it allows getting polymers with a highly controlled microstructure, with narrow distribution of the average molecular weights and polydispersity (PDI) close to unit, and with complex molecular structures (e.g. pure block copolymers), with different architectures with complex and varied functionality. Ionic polymerization also generates controlled microstructure, however the reaction conditions for ionic polymerization demand high pureness, what makes economically impracticable its the application in some industrial processes. Nitroxide Mediated Living Free Radical Polymerization (NMRP) is one of alternative techniques of the LFRP, which is more robust for impurities compared to the ionic polymerization. Many researchers have developed kinetic models that predict the behavior of the NMRP. However, its mechanism is not yet fully known. It is believed that there are other parallel reactions not foreseen in the mechanisms presented in literature. This explains the fact that the model predict conversions maximum of 100 %, whereas experimentally lower conversions are obtained at the end reaction; some models are insensitive to the controller/initiator rate; of the models under predict average molecular weights. It was intended a greater kinetic mechanism understanding of NMRP processes through the use of a mathematical model which considers the controlled polymerization of styrene using TEMPO (2,2,6,6 - tetramethyl-1-piperidinoxyl) as controller and BPO (benzoil peroxide) as a chemical initiator. Therefore, the problem has been addressed on three fronts: The first approach assumes that there are diffusional effects occurring in the NMRP process. The diffusional effects are extremely important in Free-Radical Polymerization - FRP, but were not considered in the NMRP first models reported in the literature. After a while, some authors have started to consider them. Despite the many articles published in this area, there are still many controversies over the existence or not of diffusional effects in NMRP; therefore the first contribution of this work is to elucidate this fact, with the proposed new correlations that predict the diffusional effects in NMRP of styrene with theoretical basement (the free volume theory). Empirical correlations and semi-empirical (based on the free volume theory) used both in FRP as in LFRP were tested for NMRP of styrene in a mathematical model of literature. Some of them (based in the free volume theory had been modified with new estimated parameters that better were adjusted in the study system. The second approach developed in this work makes a parametric study of all rate kinetic constants used in the model. It was observed that the kinetic constants parameters differ quite in the literature (sometimes have different magnitude orders). This second work phase aims not only a better model fit, but also identify the steps most important and infer the possibility of any steps not considered in the kinetic mechanism. The third approach considers being present a reaction between the controller and irreversible chemical initiator, this stage, not considered in the mathematical models presented in the literature. In the literature, is described a reaction irreversible in the beginning of styrene polymerization between the controller - TEMPO and chemical initiator - BPO. Based on this information, has been inserted a controller efficiency factor - fc, similar to that already consolidated factor in the initiator efficiency - f, and it was estimated in order to improve the agreement between the model and experimental data. It was possible to observe expressive agreement with experimental data using this factor - fc. In such a way, it could be shown that irreversible side reactions in the stage of initiation are significant, and that the correlations semiempirical based on the free volume theory do not improve the model significantly, which can be more one indication the diffusional effect in NMRP are very little significant, contradicting some works in literature. (AU) | |