Transport and separation through biological membranes exhibit fast, efficient and selective behaviors. These unique features have brought a significant interest recently in using artificial and biological nanopore membranes for separation and sensing pur- poses. Molecular transport through nanopores is a complex process that involves various interactions between the molecules and the channel. These interactions lead to a highly efficient and selective separation. The goal of this thesis is to under- stand this complicated phenomenon and develop a model that captures this behav- ior. Based on this modeling, the mechanism can be mimicked in different applications varying from molecular separation in nanopore sensing and drug delivery technologies to chemical and pharmaceutical separations. We construct a theoretical framework which models the general mechanism of selectivity in the translocation dynamics of molecular mixtures through a nanopore. By developing a discrete-state stochastic model, the effects of the molecule-channel interactions on molecular selectivity and flux are demonstrated. Our framework shows that the amount and the position of the interaction inside the channel has a significant effect on the separation efficiency. In addition, an efficient algorithm is developed to implement our theoretical framework for long channels. The proposed algorithm optimizes the magnitude and the position of the interaction in channels with two to ten binding sites.
Due to the exponential nature of the problem, it is computationally infeasible to apply our algorithm for channels with more than ten binding sites. To address this issue, we use machine-learning techniques and extend our results for longer channels and predict the best position for molecular-pore interaction. Our results demonstrate that the best interaction location depends on the length of the nanopore and the mass transport rates inside and outside of the channel. Depending on these factors, the best interaction location may vary between the entrance and the middle of the pore. Sensing is another important application of nanopores. Nanopores have been shown to have the ability of single-molecule sensing of various biological molecules in a rapid fashion and at a low cost. We discuss the recent progress in the nanopore- sequencing field with a focus on the nature of nanopores as well as sensing mechanisms during the translocation. The current challenges such as the fast translocation speed and low sensitivity of the sensing techniques, and alternative methods based on using Single Walled Carbon Nanotubes (SWCNTs) are also explored. Carbon nanotubes (CNTs) demonstrate enhanced water and ion flow, have wide variety of lengths and diameters, and manifest excellent electrical and optical properties. These features make CNTs one of the best candidates for artificial biochannels and nanopore devices. Our molecular dynamic simulation study shows that it is beneficial to use SWCNT for nanopore sequencing application. This is due to the possibility of slowing down the DNA transport through the tube by two orders of magnitude. Additionally, we propose to use the photoluminescence (PL) properties of SWCNTs as a sensing method. Therefore, the PL spectrum characteristics of SWCNTs at the presence of different defect densities are studied to investigate whether the carbon nanotube photoluminescence quantum yield can be enhanced at the single molecular level for future sensing technologies.