A fundamental understanding of the processes affecting fluid transport in the voids of porous materials constitutes a key step in numerous emerging applications in nanotechnology, materials design, membrane science and biology. By a purely hard sphere treatment of the diffusant, for over a century the Knudsen-based method has been the primary technique in our craft for investigating the transport of fluids in confined space, particularly in narrow disordered nanoporous adsorbents, catalysts and membranes. However, recent theoretical results and simulations, as well as experimental data on transport in well characterised narrow nanoporous materials indicate that the apparent success of the correlation is not necessarily a vindication of the Knudsen theory. Instead, these simulations, and the unrealistically high tortuosities obtained on application of the correlation, demonstrate the Knudsen model to substantially overestimate the diffusivity due to significant dispersive fluid-solid interaction. Besides affecting the molecular trajectories, which are no longer linear, the presence of the fluid-solid interaction leads to strong density profiles and adsorbate inhomogeneity in the direction normal to the pore walls. In this work, contributions are made involving such fundamental understanding of fluid transport in nanopores, and novel approaches are developed to facilitate the comprehensive analysis of transport in supported silica membranes.
Firstly, an analysis of the transport of single gases in macroporous α-alumina substrates having a mean pore size of around 500 nm was conducted, and the results indicate that the tortuosity is dependent on the gas, and varies with operating conditions in the slip flow regime. A new effective medium theory (EMT) approach for modelling the transport in the substrate was developed to include the entire pore size distribution and the pore aspect ratio effect due to finite pore length. Theoretical results of the EMT provide an improvement on existing models that are based on empirical correlations using a representative pore size. The dependence of tortuosities on operating conditions in macroporous networks in slip flow regime is caused by the difference in dependence of Knudsen and viscous flow-based permeability on pore size, temperature and pressure; these yield different tortuosity limits for the pure Knudsen and viscous flow, respectively, in the presence of pore size distribution. Besides these, the importance of the choice of representative pore radius in determining the apparent tortuosity trends with temperature is also extensively provided in different nanoporous network.
Moreover, a mesoporous γ-alumina membrane having a mean pore size around of 10.4 nm was synthesized on the surface of the macroporous α-alumina substrate by dip-coating. The transport mechanism of single gases in the mesoporous γ-alumina layer was investigated by the EMT approach to predict the macroscopic flow rate, using the classical slip flow model and a version corrected for finite molecular size, as well as the recent-developed Oscillator model in this laboratory. The analysis results indicate that all the three diffusion models describe the experimental data accurately and the interfacial pressure is correctly resolved in the approach, without the artifacts observed with the methodology using a single pore size.
In addition, using literature data on the diffusion of N2, Xe and i-C4H10 in mesoporous Shell silica spheres (mean pore size 14.2 nm), the transport of gases was also analyzed to predict the pore coordination number for various diffusion models. It has found that both the Knudsen model and the Oscillator model adequately interpret the data in conjunction with EMT approach due to this large pore size. A mesoporous amorphous silica layer having mean pore size of around 3.7 nm was further synthesized on the asymmetric support comprising a macroporous substrate and mesoporous interlayer. The transport mechanism of single gases in the mesoporous silica layer was investigated to predict the membrane thickness. The most satisfactory results were obtained with the Oscillator model, in which the fitting error was significantly reduced using an acceptable membrane thickness, indicating that the Knudsen model fails to represent the transport for the mesopores in silica.
Finally, a microporous amorphous silica layer was also synthesized on the asymmetric support, having a mean pore size around 1.5 nm. The adsorption and transport mechanism of single gases in the microporous silica membrane was examined, with the pore resistance represented by a combination of pore mouth and internal pore diffusion resistances. It has found that the pore mouth barrier dominates the overall transport resistance; and the internal diffusion resistance in the relatively smaller pores is significant, especially for weakly adsorbed gases at higher temperature. Overall, this thesis explored gas transport in different porous materials under low pressure limits, using modification of established diffusion models. The results provide fundamental understanding about how the adsorptive and diffusive behaviour of adsorbate is affected by the adsorbent structure at the nanoscales. These results are accompanied by detailed investigations to explain fluid-dependence of tortuosity under different transport mechanisms. While this work has predominantly focused on silica membranes, the results and models developed are generally applicable to transport in other nanoporous materials.