New NMR methods applied to carbohydrate analysis

Abstract

This thesis concerns the analysis of mixtures by solution state nuclear magnetic resonance (NMR), more specifically the analysis of carbohydrates as these tend to be very challenging samples for NMR. The most common method for the analysis of intact mixtures by NMR is diffusion-ordered spectroscopy (DOSY), which is able to distinguish components based on molecular shape and size (i.e. hydrodynamic radius), through the diffusion coefficient, which derives the rate at which molecules diffuse. For carbohydrates, and many other compounds, it is not unusual for DOSY to fail when some of the solution components have comparable diffusion coefficients, or may not be distinguishable by their NMR signal frequencies due to signal overlap. Therefore, to overcome these difficulties, alternatives to conventional DOSY methods are employed here in the context of carbohydrate analysis. An investigation of the compositions of a few commercial beers was undertaken, in order to evaluate some of the methods presented here. These beverages (Clausthaler Classic premium low alcohol lager and Mackeson stout 2.8% ABV) were chosen because they are known to contain a range of carbohydrates at high concentration, and need no special sample preparation procedure, such as lyophilisation (also known as freeze-drying) or filtration. Similarly, a sample of peppermint oil was analysed. Peppermint (and other mint) oils are dominated by monoterpenoid alcohols named terpineols, which present similar complications to carbohydrates in solution state NMR analysis. Additionally, this thesis contains the publications of two original piece of software, based on Matlab®, for processing and analysing NMR data. These focus on user-friendliness for non-NMR experts, and are very robust. Both software packages are open-source under the GNU General Public License, and are distributed free of charge. They include a calculator to estimate molecular weight or hydrodynamic radius from diffusion coefficient (extracted from DOSY data, for example), and vice-versa, which is also described in this thesis.

Date of Award	1 Aug 2019
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Gareth Morris (Supervisor) & Mathias Nilsson (Supervisor)