An experimental and computational study of two state of the art living free radical polymerisation techniques

Download files
Access & Terms of Use
open access
Copyright: Chaffey-Millar, Hugh William
Altmetric
Abstract
This thesis describes the research conducted by t he author in completion of a Doctor of Philosophy in the Centre for Advanced Macromolecular Design(CAMD), Univcrsity of New South Wales (UNSW) , Sydney, Australia; under the supervision of Professor Christopher Barner-Kowollik and Doctor Michelle L. Coote (Australian National University). The research has led to the creation of new knowledge in the fields of free radical polymerisation and chemical kinetics. Research was conducted in two main thrusts: (1) investigation into the governing kinetic processes behind star polymer synthesis via what has become known as a reversible addition fragmentation chain transfer (RAFT), R-group approach and (2) an entirely new mode of living free radical (LFR) polymerisation which has been named thioketone mediated polymerisation (TKMP). In the first broad area of the described research, a novel kinetic modelling scheme has been developed in which only the reactions of a single arm star are simulated explicitly. Subsequently, the molecular weight distributions (MWDs) arising from the single arm star simulation are convolved, using probabilistic calculations, to generate the MWD appropriate to a multi-arm star polymerisat ion bearing t he same kinetic parameters as the single arm one. This model is validated against experimental data, enabling, for the first time, the use of rigorous theoretical reasoning to distill a set of synthetic guidelines for star polymer synthesis via a RAFT, R-group approach. Subsequently, the product spectra resulting from RAFT, R-group approach polymerisations of para-acetoxystyrene have been analysed via mass spectrometry. This has led to direct evidence for many of the complex species whose existence had, up until this point, been inferred from gel permeation chromatography (GPC) measured MWDs. The menagerie of species identified includes, but is not limited to, star-star couples, initiator fragment terminated stars, initiator fragment terminated star-star couples and linear chains -- both living and terminated. Using a kinetic model devised specifically for application in mass spectrometry analysis, the experimentally observed abundances of each of the above species have been compared to t hose predicted by simulation. The qualitative agreement between the predicted and observed abundances has provided additional evidence that t he proposed mechanism for RAFT, R-group approach polymerisations is correct and operative. Further, it seems unlikely that significant, undiscovered kinetic phenomena exist. Due to (a) long simulation times encountered using the state of the art, commercial partial differential equation solver for polymerisation kinetics (i.e. PREDICI, Computing in Technology (CiT), GmbH; see http://www.cit-wulkow.de) and (b) the limited flexibility this software provides with respect to the types of chemical species that can be simulated, fundamental research has been conducted into the kinetic Monte Carlo method to (i) examine fundamental aspects of this simulation approach; (ii) determine the maximum speed attainable through a combination of optimisations including run-time generation of problem specific code and parallelisation; and, therefore (iii) find out what the potential of this method may be as a replacement for t he existing methods. In terms of speed, the developed code outperforms previous Monte Carlo benchmarks in the literature by a factor of 2.6 and the latest developments in the commercial tool, PREDICI that took place during the author's Ph.D. candidature give it similar performance to the herein described Monte Carlo code; however, the latter is required to run on multiple processors in order to compete with the serial algorithm implemented in PREDICI. The Monte Carlo method does, however, provide complete freedom with respect to the chemical species whose kinetics can be simulated, allowing for complex species with many chain lengths and, in principal copolymer compositions and branched structures. The Monte Carlo approach is the method of choice for these types of simulations and for the first time competes with the commercial tool in terms of speed. In the second broad area of the described research, an experimental investigation has been conducted into the applicability of thioketones, S=C (R1) (R2), as mediating agents for free radical polymerisations. The compound di-tert-butyl thioketone (DTBT), S=C-(C(CH3)3)2, has been chosen as a model reagent and this, when incorporated into a free radical polymerisation of styrene has led to a linear increase of the average molecular weight as conversion of monomer into polymer takes place - demonstrating control. A reversible radical trapping mechanism has been proposed and evidence for this has been provided in the form of an ab initio calculation of the equilibrium constant for the trapping of a styryl dimer radical by DTBT. This equilibrium constant was approximately K = 105 L mol-1 and is close to the value which is expected on the basis of the experimental results. To aid future experimental investigations intoTKMP, a quantum chemical survey has, been conducted with the aim of discovering the radical affinities of a large range of thioketones. It has been demonstrated that there is ample scope within this class of compound for potent radical trapping - far above that of DTBT. The affinities of various thioketone substrates for radicals have been understood in terms of the radical stabilising and thioketone destabilising effect of the two substituents R1 and R2 on, respectively, the adduct radical, R-S-C•(R1) (R2), and the parent thioketone. All results appearing in this thesis have been published previously in peer-reviewed scientific journals.
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Chaffey-Millar, Hugh William
Supervisor(s)
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2008
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download Chaffey-Millar-014178230.pdf 22.09 MB Adobe Portable Document Format
Related dataset(s)