The work in this thesis is focused on testing the accuracy of the gas sorption and NMRcryoporometry characterization techniques to estimate the key pore descriptors whichaffect the activity of porous materials used as catalyst supports and drug deliverysystems. Both techniques, though, assume independent pores, neglecting advancedadsorption and melting phenomena that can specifically skew the pore size distributionand subsequently lead to inaccurate predictions of catalytic or therapeutic efficiency ofthe porous system. Firstly, the independent domain theory for both processes wasstudied by breaking down the pore-filling process of a mesoporous catalyst support, intosteps. The system was partially saturated with water or cyclohexane at differentpressures, via adsorption and desorption, followed by a cryoporometry experiment ateach saturation fraction. Moreover, scanning curves and loops, together with PFG NMRand relaxometry were employed to ascertain the spatial arrangement of the liquidganglia at each partial saturation and for certain molten fractions. It was shown that theconfiguration of the liquid condensates varied with position around the hysteresis loop,deviating from the single pore hysteresis mechanism for both adsorbates. Advancedmelting of water was associated with a percolation-type transition in the connectivity ofthe ganglia, which could be curtailed to some extent by sample fragmentation. Also,some pores filled via advanced adsorption at lower pressures. On the contrary, advancedmelting of cyclohexane arose from the liquid bridging the pore cross-sections of thepartially filled pores. Secondly, an integrated nitrogen-water-nitrogen experiment wasemployed to test the source of sorption hysteresis and to compare the extent ofadvanced adsorption phenomena for nitrogen and water sorption, by isolating a subsetof pores. It was found that the Kelvin-Cohan equations and the DFT algorithmoverestimate the width of the sorption hysteresis in independent pores of the catalystsupport studied in this work. Moreover, the adsorption mechanism of nitrogen differs tothat of water, and advanced adsorption of nitrogen is less severe than that of water.Thirdly, cryodiffusometry and gas sorption techniques were used to estimate the porespace descriptors (surface area, pore size, tortuosity, porosity) of two different types ofmesoporous silicas, candidates for drug delivery. The structure-transport relationships inthese materials were investigated to interpret the drug release profiles obtained forrelease studies carried out in simulated gastrointestinal fluids. It was found that therelease rate was mainly controlled by the size of the silica particles and the silicasolubility itself in the environment present. Also, different synthesis routes were testedto optimize the drug loaded PLGA nanoparticles, for convection-enhanced drug deliveryinto the brain. Various model and real hydrophobic and hydrophilic drugs were tested.In-vitro and in-vivo studies showed that the dialysis method led to production of particleswith the desirable characteristics, which were successfully distributed in the mice brain.The sensitivity of the cryoporometry melting, gas sorption and imaging techniques wasfound inadequate to resolve the inner structure of the polymer matrix. Last, theexperimental time for the cryodiffusometry experiments in this work was long due to thehigh recycle delay times required to maximise the signal to noise ratio. It is though foundthat high delay times are unnecessary when BBP-LED pulse sequence is used, even whenthe fluid is imbibed in a mesoporous systems.
|Date of Award||31 Dec 2013|
|Supervisor||Marianne Ellis (Supervisor), Karen Edler (Supervisor) & John Lowe (Supervisor)|