Molecular simulations of light gas separations in metal-organic framework nanosheets

Carbon dioxide/methane gas mixture is present in natural gas and biogas. Sequestration of carbon dioxide from methane to avoid pipe corrosion in industry and increase the energy density of the mixture is needed. Commonly the separation is performed with energetically demanding methods or techniques where hazardous chemicals are used. Utilizing gas adsorption phenomenon in nanomaterials, for the aforementioned gas separation could prove a safer and more energy-efficient alternative. In the past two decades a class of adsorbent nanomaterials, Metal-Organic Frameworks (MOFs) have received increasing research attention for gas separations. A novel approach, growing MOFs in 2D structures as Metal-Organic Nanosheets (MONs) has shown promise in literature to outperform their 3D bulk MOF counterparts for gas separations. Although MOFs show enhanced performance for gas separations when compared to other nanomaterials such as zeolites or activated carbon, MONs demonstrate greater performance still. For example, MOF CuBDC and CuBDC nanosheet have shown great promise for CO2/CH4 gas separation, with the MON outperforming the MOF. Incorporating MONs for gas sequestration is a very new field with a limited number of publications and their potential for gas separations is not fully explored. This thesis is an entirely computational research work on CO2/CH4 gas separation with MOF CuBDC and CuBDC nanosheets. The aim of this work is to understand the carbon dioxide/methane separation with CuBDC nanosheet, identify and explore the separation mechanism and origin of the increased performance of the MON, and make the first steps for modelling MONs for gas separations. Initially, bulk CuBDC is characterized and the calculated properties are compared to available experimental data. Based on previously published experimental observations, the most likely form of the CuBDC nanosheet structure is generated computationally. The adsorption properties of bulk CuBDC and CuBDC nanosheet are estimated and compared to experiments. The observed experimental adsorption preference of CuBDC (both bulk and nanosheet) for carbon dioxide is confirmed computationally via Monte Carlo simulations. The kinetic properties of the system were also evaluated under equilibrium conditions (NVT molecular dynamics) and it was found that methane molecules exhibit larger self-diffusion coefficients in bulk CuBDC. Experimentally, the system is typically studied under a pressure drop, rather than under equilibrium. As such, non-equilibrium molecular dynamics (NEMD) simulations were employed and reveal that the effective gas separation happens due to a pore blocking mechanism, where carbon dioxide molecules preferentially adsorb into the material and interact strongly with the framework. In this regime, carbon dioxide molecules limit the penetration of methane molecules into the nanosheet and although methane molecules diffuse faster through the material once within it, this pore blocking ensures that the permeability of CO2 is greater than methane. For a defect-free CuBDC nanosheet and an equimolar CO2/CH4 gas mixture the simulated permeability values ranged from 93703-116716 Barrer for carbon dioxide and 48834-76869 Barrer for methane, while the selectivity was around 2. The simulations confirm what was observed experimentally, that CuBDC nanosheet is more efficient for the gas separation when compared to bulk CuBDC. It is proven that the MON's surface is the reason the nanosheet outperforms the bulk MOF. Methane molecules accumulate on the MON's surface, facing the mass resistance while trying to penetrate the CuBDC nanosheet structure. The same mass resistance is not faced as extensively in the bulk MOF case as the external surface area per unit volume of the MON is significantly higher than that of the bulk MOF. The need for multicomponent simulations instead of pure component simulations when studying gas separations is underlined. When pure components were studied, methane demonstrated higher permeability values in the CuBDC nanosheet when compared to carbon dioxide, which was in direct contrast with the experimental results found in literature. When the binary mixture was considered a reversal in the permeability values was observed. Because of the surface importance on the CO2/CH4 gas separation with CuBDC nanosheet, different chemical groups were tested as capping groups on the MON's surface. It was found that the adsorption capacity of the nanosheet does not change with various surface capping groups. Also, the separation performance of CuBDC nanosheet did not very significantly for the capping groups tested. This finding is important as it means that different modulators and solutions used during the experimental synthesis are not likely to have a serious effect on the gas mixture separation. Finally, since the "birth" of the MOF field about two decades ago researchers have been modelling MOFs as perfect defect-free crystals. However, this is almost impossible to achieve in the laboratory. There is a very limited amount of publications addressing defects in MOFs. In this work, the effect of missing linker defects in CuBDC nanosheet for the equimolar CO2/CH4 gas mixture separation was studied. It was found that 10% missing linker defects were capable of cancelling the pore blocking mechanism and allowed methane molecules to penetrate CuBDC nanosheet structure which was no longer carbon dioxide selective. The development of nearly defect-free experimental synthesis methods is advised for MONs in which the gas separation mechanism is based on pore blocking effect.