Molecular simulations in zeolites for industrial applications

Resumen

Zeolites are known for their suitability as molecular sieves. Actually, they are already firmly settled in industry to selectively separate gases or capture them. In this thesis, the processes selected share the objective of contributing to make more competitive or enhance alternative processes to those based on carbon technology. These processes have been approached from a computational point of view. Molecular simulations have been carried out to gain molecular insight on gas mixtures separation processes at the same time as obtaining valuable information to suggest operational schemes to perform these separations. Additionally, some specific underlying molecular phenomena have been tackled to study supplementary aspects, not centered on particular industrial processes but relevant to understand the behavior of adsorbed molecules in separation procedures. Therefore, two blocks can be differentiated in this thesis: Gas mixture separations for industrial processes Chapter 2 An operating procedure is proposed to separate the components of a tail gas expelled from a Fischer-Tropsch process. The quinary gas mixture is composed of light gases, i.e. CO2 , CO, CH4 , N2 , and H2 in a typical, described composition. The Fischer-Tropsch process is a key step in a global gas-to-liquid process to synthesize hydrocarbons. The separation procedure has a twofold goal: upstream reintroduction of the captured CH4 and CO and trapping of CO2 to avoid its emission. To that end, the performance in the described separation is assessed for four zeolites of high industrial impact (DDR, FAU, MFI, and MOR), with main focus on the effect of location and amount of aluminum atoms in the zeolite lattices. The decision on the final scheme is based on the combination of computed adsorption isotherms, selectivity and diffusion coefficients. In addition, applicability and accuracy of IAST is studied in these adsorbate–adsobent systems. Chapter 3 A large multi-scale computational study is performed to achieve a selective separation of a mixture of CO2 , CO, and O2 gases. The separation takes place in the context of the nonthermal plasma-assisted CO2 dissociation process, involved in a process scheme aiming at the production of fuels with neutral carbon footprint. The proposed route requires an additional step to obtain pure CO from the mixture and avoid a high CO2 recombination. A widespread screening on 174 zeolites, evaluating selectivity at low coverage and successive adsorption simulations at molecular level for specific structures, combined with IAST, provides a scheme to perform the separation under mild operation conditions. Then PSA simulations are carried out to find the optimal parameters to achieve the desired separation at a pilot-plant scale. Chapter 4 Deuterium and tritium from hydrogen separations are studied over a wide range of pressures and low and cryogenic temperatures. Due to the nature of the adsorbates and operation conditions, quantum corrections are considered. Models for the deuterium and tritium molecules are proposed, derived from the hydrogen model previously reported. Experimental adsorption isotherms for H2 and D2 respectively check and validate the models for such molecules in pure silica zeolites. Then, a study on the adsorption selectivity at infinite dilution is done on 210 pure silica zeolites, and subsequent diffusion and adsorption simulations are performed over a range of pressures and temperatures for the most promising zeolites. Three zeolites, BCT, AVL, and MVY, are identified as the best candidates to perform a separation of a 1:1 D2/H2 mixture. One of them, BCT topology, is found to show, at low temperature, the highest adsorption selectivity reported to the best of our knowledge. The same structure is also found to obtain an extremely high selectivity for a 1:1 T2/H2 mixture. Molecular insights on additional aspects for molecular separation Chapter 5 A study on the effect of cations on diffusion of CO2 and CH4 molecules in MFI zeolite is carried out. The industrial relevance of both molecules and of the zeolite framework is beyond all discussion. Theoretical aluminum distributions are generated considering the 12 T crystallographic positions of framework MFI. Monovalent and divalent counterbalancing cations are considered to neutralize the negative charge introduced in the system by the aluminum atoms. Probability density of cations and energy profiles for the adsorbates, both depending on aluminum distributions, are evaluated together to produce a prediction on the behavior of adsorbates and cations, which is found to be consistent with subsequent diffusion simulations. All the results shed light on why zeolites with the same chemical composition have different dynamical behaviors. Chapter 6 Understanding the role of water in LTA zeolite is key due to the use of this topology in water removal and dehydration processes. The wide range of aluminum substitutions available experimentally for this topology, from the pure siliceous lattice up to the theoretical maximum of Si:Al=1, means the degree of hydrophobicity or hydrophilicity of the zeolite can be tailored. In the theoretical study, besides considering different Si:Al ratios, two lattices are examined for each one, keeping the crystallographic positions fixed after substitutions and allowing the lattice to relax. Adsorption isotherms in fixed and energy-minimized lattices and a thorough analysis of the location of water molecules reveals that: adsorption sites are determined by the hydrophilicity of the lattice. The more hydrophilic, the bigger the lattice, which reinforces the ability of the structure to adsorb in the narrow pores. The pressure is found to affect strongly the preference for large or narrow pores. A case is identified, in which at increasing pressures or loading narrow pores are first populated, then emptied as the large pores fill, and then finally populated again. Chapter7 Conclusions