Polymers have been used as biomaterials for nearly a century and have recently become the material of choice for use in tissue engineering. However, the classes of biodegradable and biocompatible polymers available for use in biomedical devices and as tissue engineering scaffolds are limited. This lack of available polymers with suitable properties could inhibit the development of biomedical devices with improved biocompatibility and hinder the growth of the fledgling tissue engineering field. Researchers in the polymer and biomaterials fields have tried to remedy this problem by applying combinatorial and high throughput methods developed in drug discovery to the search for new polymers. A recent advance has been the development of combinatorial polymer libraries printed as microarrays. This format allows the polymers to be readily screened for their cell adhesion and differentiation properties, allowing ‘hit’ materials with ideal properties to be identified. However, without knowledge of the surface properties of these novel polymers it is impossible to rationalise their biological properties. The surface characterisation of such microarrays presents numerous practical problems included small sample size, sample number and even analysis of such large amounts of data. It is the aim of this thesis to develop methods for the characterisation of the surface chemistry, wettability and protein adsorption properties of polymers in situ in microarray format and within realistic timeframes. The thesis will explore multivariate statistics in the form of PCA and PLS as methods of analysing the large amount of data acquired. The first part of this thesis describes the surface chemical analysis of a polymer microarray using ToF-SIMS and XPS. A comparison of the polymers’ surface to bulk chemistries by XPS indicated that 64 % of the polymers had a surface chemistry which differed from the bulk. This reinforces the need for characterisation of the polymers’ surface chemistries, as it is obvious that this can not be inferred from their bulk chemistries. ToF-SIMS imaging was shown to be an ideal method of studying the distribution of specific ion species across the array and to confirm that the microarray was printed in the intended layout. Principal component analysis is shown to be an ideal technique to analyse both ToF-SIMS and XPS spectral data from the arrays, allowing similarities and differences in the surface chemistry of the polymers to be easily visualised. To estimate the surface energies of the arrayed polymers it is necessary to use picolitre volume droplets to make contact angle measurements. In Chapter 4 it is shown that contact angle measurements taken from picolitre volume water droplets are equivalent to those measured from more conventional microlitre droplets. In Chapter 5 picolitre contact angle measurements are used to estimate the polar and dispersive surface energies of a polymer microarray, which has been specifically designed to exhibit a maximum range of surface energy values. The analysis shows that there is indeed great variation in the WCA and polar surface energies of the polymers, demonstrating the power of intelligently designed combinatorial libraries. To understand the chemical basis of this large range of surface energies the results are compared to surface chemical data from ToF-SIMS and XPS. Surface atomic and functional data from XPS is unable to provide any definitive explanations for the range of surface energies observed. However, information about the molecular structure of the surface from ToF-SIMS gives an insight into what surface functionalities are responsible for high and low surface energies. In Chapter 6 PLS regression is investigated further as a method for investigating surface structure-property relationships in large polymer libraries. Specifically two issues are investigated: the influence of sample number on the results obtained and the ability of PLS to make quantitative predictions. The ToF-SIMS and surface energy dataset discussed in Chapter 5 is used for this task. It is demonstrated that the results obtained from PLS models of large polymer libraries are equivalent to those obtained from much smaller datasets, in terms of the ions identified in the regression vector. Using various test sets of polymers it is shown that there is a limit to the predictive ability of PLS: specifically, as the difference between the training and test sets increases, the quality of the predictions decreases. Potential problems with data pre-processing and re-scaling are also identified. In the final experimental chapter two methods are described for investigating protein adhesion and adsorption to micro-arrayed polymers using AFM and fluorescently labelled proteins. Both methods indicate a wide range of protein adsorption properties within the group of polymers analysed. A good correlation between the two sets of data was observed which appears to validate both methods. In summary the work described in this thesis has demonstrated the feasibility of the characterisation of the surface chemistry, energetics and protein adsorption properties of a micro-arrayed polymer library within realistic time-frames. PCA and PLS have been shown to be useful tools for analysing the data obtained. It is hoped that the methods described in this thesis will allow the biological data from polymer microarrays to be rationalised using the surface properties of the polymers, allowing the design of new biomaterials.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:514745 |
| Date | January 2009 |
| Creators | Taylor, Michael |
| Publisher | University of Nottingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://eprints.nottingham.ac.uk/10717/ |
Page generated in 0.003 seconds