Ab initio calculation of intermolecular potentials, prediction of second virial coefficients for dimers H2-H2, H2-O2, F2-F2 and H2-F2, and Monte Carlo simulations of the vapor-liquid equilibria for hydrogen and fluorine

The pure elements hydrogen, fluorine and oxygen and the mixtures hydrogen-oxygen and hydrogen-fluorine are used in several industrial applications nowadays. Hydrogen might become a renewable energy carries in fuel cell technologies [122, 16]. Hydrogen is considered a fuel which can replace all the major fuels [46, 95]. Consequently, the estimation of thermodynamic data for the mentioned systems over a wide range of temperature and pressure is a need for future practical applications. This thesis presents the results of the calculations of four new ab initio intermolecular pair potentials, the second virial coefficients with first-order quantum corrections of the dimers H2-H2, H2-O2, F2-F2 and H2-F2, and thermodynamic properties of phase equilibria for the pure fluids hydrogen and fluorine derived from the Gibbs ensemble Monte-Carlo simulation techniques. The new intermolecular interaction potentials of the dimers H2-H2, H2-O2, F2-F2 and H2-F2 were developed from quantum mechanics, using coupled-cluster theory CCSD(T) and correlation-consistent basis sets aug-cc-pVmZ (m = 2, 3, 4); the results were extrapolated to the complete basis set limit (denoted aug-cc-pV23Z). The constructed potential energy surface of the dimer H2-H2 turned out to be in good agreement with that proposed by Diep [25]. The interaction energies were corrected for the basis set superposition error (BSSE) with the counterpoise scheme. For comparison also Møller-Plesset perturbation theory (at levels 2 to 4) as well as the basis sets 6-31G and 6-311G were investigated, but the results proved inferior. Molecular properties were calculated for assessing the accuracy level of each theoretical method and the basis set, respectively. The quantum mechanical results were used to establish four new analytical pair potential functions. The adjustable parameters of these functions were determined by global least square fits to the ab