Molecular Simulations Of Phase Transitions In Multi-component Polymer Systems

In multi-component polymers, their crystallization and phase separation are two basic phase transitions. These phase transitions will interplay with each other if they take place within the same temperature window, so as to modulate the morphologies and properties of polymer materials. In the engineering process of polymer materials, polymer chains are frequently forced to strain. As a result, the nucleation mechanism and crystalline morphologies are changed. Phase separation between different components is induced by strain as well. Therefore, the interplay between strain-induced crystallization and phase separation in multi-component polymers becomes extremely complicated.Block copolymers, in which two or more polymers with different chemical composition are combined together by chemical bonds, can be regarded as a special kind of multi-component polymers. The linked blocks are difficult to mix with each other. Therefore microphase separation on molecular level takes place. The interplay between crystallization and microphase separation in crystalline block copolymers results various ordered materials, which are applied to many areas. The directed self-assembly of block copolymers on chemical pattern is believed to be the most promising technology to fabricate electrical devices with a sub-20nm size. Due to the mismatching between the directed period of chemical pattern and natural period of block copolymers, polymer chains are either stretched or compressed during directed self-assembly, so as to impact the directed self-assembly.Since the strain-induced crystallization and phase separation in multi-component polymers are so complicated that micro-level details are necessary to understand these process and their interplay. The development of computer simulations provides a good tool to study these problems. Especially, the recently developed dynamic Monte Carlo simulations based on lattice model behave magnificently in dealing with polymer crystallization and phase separation, thus gained fruitful results. In our thesis, we performed dynamic Monte Carlo simulations to study several problems, including the melting point of solution polymers under strain, the crystallization and segregation of polymer blends under strain, as well as the asymmetry between stretching and compressing of microdomains in directed self-assembly of triblock copolymers by chemical substrate.In chapter one, we introduced progress of crystallization and phase separation in multi-component polymers. Basic concepts of polymer phase transitions are presented, with polymer crystallization and phase separation emphasized. For polymer crystallization, we described the prediction of polymer melting points and the influencing factors. For phase separation, we mainly illustrated the binodal decomposition and spinodal decomposition. After that, we gave the research progress on interplay between crystallization and phase separation in static polymer systems. The influences of strain on crystallization and phase separation are separately presented. Finally, we introduced in detail the interplay of crystallization and microphase separation in block copolymers and progress on directed self-assembly of block copolymers.In chapter two, we described the simulation methods. After a simple description of computer simulation, we mainly introduced two molecular simulations on polymer crystallization, that is, molecular dynamic simulations and Monte Carlo simulations. Since our work is based on dynamic Monte Carlo simulations of lattice model, we demonstrated its simulation details, including the motion of polymer chains in the lattice, the unit cell and periodical conditions, the energy parameters in chain motion as well as the Metropolis important sampling method.In chapter three, we combined two Flory's thermodynamic equations of polymer melting points, one on solution polymers and another on stretched polymers, to predict the melting point of stretched solution polymers. The results were validated by the onset strains for crystallization of solution polymer networks at variable temperatures, polymer volume fractions and solvent qualities observed in our dynamic Monte Carlo simulations under a constant strain rate. In addition, we found that in poor solvent, a calibration of polymer concentration to the polymer-rich phase appears as necessary owing to phase separation prior to crystallization.In chapter four, we realized stretching of binary blends of constraint and free polymers, reproduced their strain-induced segregation and investigated its influence to crystal nucleation. We firstly confirmed an entropic driving force for strain-induced segregation between two components. We then observed its competition to strain-induced crystallization with the decrease of strain rates, which results in variable compositions in the crystal precursors. The scenario may settle down the arguments on the different compositions in the flow-induced nucleation precursors of shish-kebab crystallites between polymer solutions and melt.Microdomains of ABA triblock copolymers can be stretched in a larger extent by the chemical-patterned substrate than being compressed. In chapter five, we performed dynamic Monte Carlo simulations to reproduce this behavior. We observed that the middle part of bridge chains can be released to fill the space gap between two against loops, which optimizes the total free energy and thus allows further stretching. Such a delicate inhomogeneity in the bridge-chain conformation has not yet been considered by the conventional approach of the mean-field self-consistent theory.In the last chapter, a summary of this thesis and a perspective of further study were given.