The macroscopic properties of porous systems, which are ubiquitous in nature, are determined by their internal microstructure. For example, the insulating properties of building materials or the transport properties of chromatography columns is determined by the size, shape, and connectivity of their internal pore structure. Therefore the key to understanding the behaviour of porous materials is determining their microstructure. One method of determining the microstructure of porous systems is to observe the translational diffusion of molecules within the pore space. Diffusing molecules can probe the size, shape, and connectivity of porous systems via collisions with the pore walls. These collisions impart a spatial signature on the motion of the molecules which, when measured, can be used to estimate the pore structure. Nuclear magnetic resonance (NMR) diffusometry, also known as pulsed gradient spin-echo (PGSE) NMR, is a powerful non-invasive technique commonly used to measure the translational diffusion of molecules. For molecules diffusing in unconfined isotropic environments, PGSE NMR measurements provide information on the molecule itself (i.e., hydrodynamics). PGSE NMR measurements of molecules within confined geometries also contain information on the surrounding geometry. The non-invasive nature of PGSE NMR and its ability to directly measure the average diffusion propagator has stimulated great interest in applying it to determine the structural properties of porous systems. This interest has lead to significant theoretical and experimental developments in determining structural information from PGSE NMR restricted diffusion experiments. Still, there are significant complications and limitations which restrict the applicability of PGSE NMR to the study of porous materials. This thesis is primarily concerned with improving NMR diffusometry techniques for probing porous systems. PGSE NMR probes the mean square displacement (MSD) of molecules on a timescale defined by the change in nuclear spin phase which is induced by two magnetic gradient pulses separated by a delay I". In order for PGSE NMR to provide structural information, the MSD of the molecules during the delay I" must be of the order of the characteristic distance(s) of the pores being probed such that the structure can have an impact on the measured MSD. But to achieve this, I" cannot be arbitrarily increased and is limited by the spin relaxation of the probe species. Thus, to date, PGSE NMR has been limited to probing porous systems with characteristic distances with an upper limit ~ 100 Im. In this research, the favourable relaxation properties of singlet spin states were used to probe significantly larger pore sizes. Singlet spin states were created using 2-3 dibromothiophene in CDCl3 and used to probe the structure of rectangular and cylindrical pores. Diffusive diffraction coherence features, obtained from PGSE NMR restricted diffusion experiments, using a singlet states pulse sequences were similar to the results obtained from a pulsed gradient stimulated echo (PGSTE) pulse sequence for 2-3 dibromothiophene diffusing between planes separated by 250 Im and I" set up to 15 s. Singlet spin state restricted diffusion experiments done using 2-3 dibromothiophene diffusing within a cylindrical pore of diameter 0.8 mm show that structural information can be obtained from macroscopic pores with = 70 s. In conventional PGSTE measurements I" times this long are impossible. Hence this research shows that the range of pore sizes able to be studied using PGSE NMR can be significantly extended by using singlet spin states. The diffusive diffraction coherence features produced by PGSE NMR can be distorted or smeared out due to a distribution of characteristic distances (polydispersity). This is because the average weighted signal acquired in PGSE NMR superimposes nodes corresponding to the different characteristic sizes thus damping the coherence features. Previous attempts of characterising polydispersity have incorporated distribution functions into analytical expressions for the PGSE NMR signal attenuation however the validity of this method has not been checked systematically. Here, PGSE NMR experiments were carried out on a system of parallel planes prepared to give different amounts of polydispersity. Samples prepared with a higher degree of rugosity clearly show a damping of the coherence features. A Gaussian distribution of characteristic distances incorporated into the expression for diffusion between planes showed good agreement with the experimental data. Fourier inversion of the experimental data provided additional insight into the pore size distribution and could explain possible inconsistencies between the fitting and experimental results. For emulsions, droplet concentration, size, polydispersity, and interfacial properties/interactions are key features which influence emulsion characteristics such as stability, texture, and appearance. PGSE NMR is particularly favourable for studying the characteristics of emulsions because, in addition to its non-invasive nature, it can measure highly concentrated and opaque emulsions. Here, a concentrated emulsion was studied in order to investigate the impediments to the accurate quantification of emulsions. Several complications can arise when studying concentrated emulsions using NMR diffusometry. These include background gradients, radiation damping, overlapping resonances from different molecules or molecules in different motional environments, polydispersity, contact between adjacent droplets, a loss of magnetisation due to relaxation and/or permeation during the pulse sequence, exchange, and finite gradient pulse effects. For a concentrated water-in-oil emulsion, this study found that background gradients are minimal and the effects of radiation damping can be suppressed easily. Overlapping resonances from molecules in different motional environments were able to be modelled by summing the different population fractions within the emulsion. Also methods for characterising polydispersity developed from studies of the planes were applied to the emulsion. The pore size distribution obtained using the damping of the coherence features was similar to those obtained from established techniques. Finally, the pore structure parameters of a sandstone core were estimated. For sedimentary rocks, pore structure parameters such as surface area-to-pore volume ratio, pore size distribution, porosity, and permeability are key to understanding transport processes. These transport processes determine the viability of rock beds for commercial applications such as hydrocarbon and heat mining. Sedimentary rocks are one of the most challenging systems to study using NMR diffusometry. They contain multiple deviations from ideal NMR diffusometry experimental conditions. These deviations include irregular pore shapes, polydispersity, and background magnetic field gradients arising from differences in magnetic susceptibility. The methods developed in this thesis were used to characterise the sandstone core. The distribution of background gradients were used to estimate pore size and pore size distribution. These estimates gave a good fit to the experimental data and correlated to electron micrographs. The transport characteristics of the sandstone were estimated from changes in the measured diffusion coefficient as a function. Models fitted to the NMR diffusometry data gave the porosity of the sandstone as 0.21. This compared to a value of 0.19 measured using helium porosimetry. Permeability estimated from diffusion measurements was 6.6 Darcy which compared to 1.8 Darcy from helium porosimetry measurements. The difference is a result of NMR diffusion measurements being more sensitive to larger pore spaces which have higher permeability. Importantly however, it is these larger pores which are important for determining the transport properties of porous systems. From changes to the apparent diffusion coefficient, the approximate time for a molecule to move into adjacent pores was estimated to be 0.29 s. This time matched exchange experiments which observed the time taken for molecules to move into regions with different background gradient magnitudes.
| Date of Award | 2009 |
|---|
| Original language | English |
|---|