Abstract
This thesis details work carried out over a period of three years on the two gelling carbohydrates agarose and carrageenan. The major part of the work deals with agarose. Two approaches have been used which yield information from different angles; these are the experimental (laboratory) and the simulation (computational) approaches. There is a large field of interest in gelling carbohydrates from the point of view of the food industry. Their extraordinary ability to form stable gels and emulsions incorporating other food ingredients makes them important in many deserts and dairy products. In the present work, models for agarose and carrageenan carbohydrates were developed using structural x-ray data and related carbohydrate literature. The models were treated with two different solvent simulation methods. It was found that the inclusion of individual solvent molecules (the closest approximation to a real solution) was extremely uneconomical when the demands on computing time were taken into account, and in fact the long term outcome of the simulation was the same for both methods. Inclusion of solvent simply reduces diffusion rates and the time constant for chain flexing. Gel permeation chromatography and differential scanning calorimetry were used to prepare samples of agarose molecules of known size, and to probe temperature dependent phase transitions. This work was done at the UNILEVER laboratories at Colworth House, Sharnbrook in Bedfordshire. It was found that only molecules longer than fifteen residues displayed the molecular ordering transition typical of agarose polymer, and a value for the enthalpy of the transition of -1.5kcal per mole of residues was measured. It was predicted that in agarose itself, helical regions of a size of approximately 40 residues should exist. Simulations were then done on several agarose molecules of different sizes in order to parallel the experimental work. The differences in energy between molecules in various conformations were compared. These results were also related to helix-coil transition theory. The modelling predicts an enthalpy per mole of residues for the agarose coil to helix transition of approximately -2kcal, and indicates that single agarose coils may be of some importance in agarose gel structure. The work illustrates the difficulty in modelling such complex systems, and in fact it remains impossible to observe agarose molecules undergoing the transition between a random coil and a helical conformation.