Use this URL to cite or link to this record in EThOS:
Title: Computational modelling of self-assembling triblock copolymer surfactants for use in the synthesis mesoporous silica
Author: Sobek, Olivia Nile
ISNI:       0000 0004 8509 6071
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Block copolymers find applications in many fields, including adhesives, plastics, drug delivery and photonics. Several of these rely on the ability of block copolymers to self-assemble into ordered mesophases in solution. One such application of particular interest to our research group is their use as templates in the synthesis of porous silica materials, such as SBA-15. Because of their highly ordered pores, high surface areas, high functionality and low cost, mesoporous silicas have been of great interest for an increasing variety of applications and research. Understanding the synthesis mechanism for this class of materials, however, models that can predict how block copolymer templates self-assemble in aqueous solution. This study aims to produce an accurate coarse-grained, CG, model of self-assembling block copolymers, including those used in the synthesis of SBA-15 mesoporous silica (i.e., Pluronic surfactants). Such a model will enable us to probe the large time and length scales that are needed to describe the mesostructure formation from solution, thus clarifying themechanisms by which these materials are formed. Our approach was based on the established MARTINI CG force field,which has been previously applied to model these systems. We have found that existing models were unable to accurately describe micelle aggregation self-assembly of Pluronic surfactants, although they are designed to match single-chain properties. To parameterise and validate new self-assembly models for Pluronics,we performed a a series of intermediate CG model parameterisations;alkane models, PEO and PPO models, and finally, Pluronic surfactant models. We have thus systematically tested the existing MARTINIparameters for the alkane solvent basis of these systems againstexperimental values such as density, Gibbs free energies of solvation,enthalpies of vaporisation, self-diffusion coefficients, and radii of gyration. By noting where the MARTINI model was lacking, we were then able to parameterise 1-bead, 2-bead, 3-bead, and 4-bead 2:1, 3:1,4:1, and 5:1 mapped alkane models, methodically matching Lennard-Jones and bonded parameters to known density, Gibbs free energies of solvation, and enthalpies of vaporisation properties, and validating the models against experimental self-diffusion coefficients and shearviscosity, and simulated radii of gyration. These models were then parameterised for non-bonded interaction parameters with MARTINI water, specifically for free energy of solvation, and further validated for use in polyethylene, a polymer with a similar carbon backbone to alkanes, by matching densiity and radii of gyration. We then validated seven existing coarse-grained PEO and four existing coarse-grained PPO models, for 1-bead and 2-bead model simulations,against known experimental and simulated data such as Gibbs free energies of solvation, enthalpies of vaporisation, densities, and selfdiffusion coefficients. The two models that best matched the 2-beadsimulation results, one for PEO and one for PPO, were chosen and then calibrated for free energy of solvation in our 4:1 and 2:1 mapped hexadecane and hexane models, and a parameterised 3.5:1 heptanemodel that was interpolated from our original alkane model results. Longer chains of these two models were then validated against simulated end-to-end distances, relaxation times and radii of gyration data, and were observed in polar and nonpolar solvents to ensure their behaviour adhered to theory for self-assembly (i.e. the modelsdisplayed the correct hydrophobicity and hydrophilicity typical of PEOand PPO within these solvents). When we had established alkane and PEO and PPO models that had the best adherence to experimental and simulated data, we finally moved on to Pluronic surfactant simulations. We chose suitable PEO-PPO interaction parameters from the same study as our chosen PPOmodel, and we create eight Pluronic models for Pluronic L31, L35, L44,L62, L64, P85, P123, and F38. Melts of seven of these models were compared to experimental and simulated properties including densities, shear viscosities, heat capacities, and radii of gyration, and larger chain simulations of three of the models, Pluroincs L35, L44, and P123 were run in MARTINI water. These larger simulations, as well as another of Pluronic P123 in our 2:1 mapped hexane model, were evaluated for self-assembly and aggregation, and the micellisation free energies and aggregation numbers of the resulting aggregates were calculated against experimental data and theory. This process has lead to improved mapping schemes and adjusted parameters for alkane solvent models, PEO and PPO models with the most accurate properties in themselves and in those alkane solvents,and ultimately, Pluronic models with self-assembly behaviour. It is unfortunate, however that this process also inevitably ended up being more time-consuming and difficult than expected and we were unable to progress any farther with this thesis. We expect that in the future our models will be combined with existing models of silica precursors to effectively model and analyse SBA-15 synthesis.
Supervisor: Jorge, Miguel Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral