Free radical polymerization has a wide range of applications and continues to attract research interest as the demand for tailored specialty polymers grows. Sequence distribution, tacticity and composition of copolymers play a key role in understanding the reaction kinetics and the properties of the polymers 1. One of the chief challenges in unraveling the relationship between sequence and properties is that experimental techniques allow only some aspects of sequence to be uncovered, but do not yet measure explicit sequence distributions. Research efforts have been focused on modeling and simulation of free radical polymerization (FRP) in order to complement experimental efforts by revealing the full sequence information of all polymer chains. Among multiple modeling techniques, kinetic Monte Carlo (KMC) has been exploited for simulating polymeric reactions, because of its capability to record the explicit sequence of every polymer chain, enabling more detailed analysis of polymer properties. Nevertheless, compared to deterministic models, KMC is computationally much more expensive, limiting its application for efficient design of sequence properties. In this thesis, a general KMC simulation framework that simulates free radical copolymerization was developed, with the ability to keep track of the explicit sequence of polymer formed at any instant of the reaction, from which the statistics related to sequence distribution can be easily extracted, such as dyad or triad fractions and sequence length distribution. In order to make KMC computationally more competitive, two novel methods were developed: the scaling concept and the hybrid algorithm. These innovations reduced the simulation time by approximately a factor of 100 while keeping accurate tallies for polymer sequence distributions and other properties including molecular weight distribution and copolymer composition. The scaling concept successfully tackled the issue that the minimal number of molecules used in KMC simulations for free radical systems is too large. The hybrid method reduced the effort spent on simulating propagation events which consume the largest fraction of time in KMC simulations. With this novel KMC algorithm that provided significant acceleration in hand, it was integrated into a simulation-based optimization algorithm that designs molecular weight and sequence properties of polymers. As demonstrated through case studies, it was shown that the optimization framework can efficiently and precisely search for the synthesis conditions to achieve certain targets for molecular weight and sequence distributions.

Last modified

• 04/02/2018

Creator

• Hanyu Gao

DOI

• https://doi.org/10.21985/N2NH5G

Subject

• Chemical and Biological Engineering

Keyword

• Polymerization
• Process Optimization
• Kinetic Monte Carlo

Date created

• 2017-01-01

Resource type

• Dissertation

Rights statement

• In Copyright