Design and synthesis of complex polymer architectures is a promising field in polymer chemistry to produce new materials with unprecedented macroscopic properties. Recent advances in 'living' radical polymerization and polymer coupling chemistries has facilitated the fabrication of new polymer topologies. The main goal of this thesis is to develop novel methods to fabricate cyclic polymers with pendent functional groups, and use them as building blocks in the synthesis of complex polymer topologies. The 'click' reaction used to couple the cyclic polymers was by copper-catalyzed azide/alkyne cycloaddition reaction (CuAAC). The thermal properties of the resultant complex structures were investigate by Differential Scanning Calorimetry to determine the effect of topology on the glass transition temperature (Tg).
First, the combination of RAFT polymerization and CuAAC reaction were used for the synthesis of cyclic polymer with pendent hydroxyl group. An alkyne functional RAFT agent was used for the synthesis of linear polystyrene (PSTY), in which the RAFT moiety was then converted to an azide moiety and a free OH group via a two step synthetic reaction. This linear polymer was cyclized in high yield and considered as a highly efficient method, and has the potential to be applied to a wide range of polymers made by RAFT. Although the monocyclic polymer was synthesized in high yield, we observed ester cleavage during the synthesis of more complex topologies from the monocyclic precursor building blocks. A detail degradation study was conducted using different catalyst/ligand complexes, and finally the methodology for the synthesis of different topologies of cyclics was amended to reduce this degradative side reaction. However, for the fabrication of more complex topologies, the synthetic methodology was redirected towards a more stable synthetic approach.
A modular approach was followed for the synthesis of multifunctional linear polymer precursors through modulating the Cu(I) activity towards the click reaction over radical formation. The post-modification approach allowed for the synthesis of α, ω-heterotelechelic linear polymer precursors which was cyclised by using a modified CuAAC cyclization reaction in which the hydroxyl functional groups were equally spaced. The hydroxyl groups were converted to azides or alkynes and then further coupled together through the CuAAC reaction to produce complex structures, including a spiro tricyclic and 1st generation dendritic structures. All these structures were produced in high yields with good 'click' efficiencies. The purity and ‘click’ efficiencies were calculated by fitting the experimental SEC traces with a log-normal distribution (LND) model based on fitting multiple Gaussiun functions for each polymer species. The crude polymer was purified by preparative SEC that essentially removes all the unreacted species and by-products.
In the follow up work, a range of different topologies of cyclic homo and copolymers were synthesized by combining of ATRP, SET-LRP and CuAAC coupling reactions. The homopolystyrene architectures ranged from di-block to 3-armed star polymers, consisting of both linear and cyclic polymer building blocks. Additionally, the di-block, AB, miktoarm AB2 and A2B type of amphiphillic copolymers consisting of PSTY and polyacrylic acid (PAA) and their cyclic analogues were successfully synthesized. All these topologically diverse polymers were purified by preparative SEC to remove any impurities formed during ‘click’ reaction. To investigate the topology effect on thermal property such as Tg, the polymers were characterized by differential scanning calorimetry (DSC). The results revealed that the topologies which possessed higher number of cyclic units (i.e., lower number of chain ends) showed higher Tg values. The thin film self-assemblies of block copolymers of both linear and cyclic analogues were also characterized by AFM to investigate the effect of cyclic topology on the morphology. The thin film domain spacing of cyclic block copolymer decreased by ~50% compared to the linear analogue due to the structural compactness.
Finally, a range of complex polymer architectures such as linear, cyclic, spiro di and tricyclic, star tricyclic, G1 star tetracyclic and dendrimer pentacyclic were used to investigate the effect of placing knots in different locations in a cyclic polymer on the glass transition temperature. The molecular weight of all these polymers was kept essentially the same to avoid the influence of molecular weight effect on Tg. To form a knot, we used covalent linkages that produce irreversible knots. The experimental results revealed that the Tg for this series of polymers was not only affected by the number of knots but also the type and location of the knots.