Equa­tion of State (EOS) on the basis of hy­brid data­sets. A com­bined ap­proach for the de­vel­op­ment of ref­er­ence equa­tions and ex­act mo­lecu­lar mod­els

A new approach to the development of empirical equations for pure component systems on the basis of hybrid datasets is elaborated, based on both experimental and simulation data. The optimization of such empirical equations of state requires a large set of reliable thermodynamic data. Albeit such data were acquired in the past with tremendous effort, their availability even for commonly used fluids, such as methanol or formic acid, is very limited. Recent advances in molecular modelling and simulation have demonstrated that the predictive power is satisfactory for thermodynamic data. Molecular models (geometry, electrostatics) are generated on the basis of quantum chemical ab-initio calculations and an additional small set of experimental data. Data generated by simulation cannot be more accurate than the experimental data used to parameterize the molecular interaction potentials (LJ type), nonetheless, it is possible to generate large sets of reliable data. In order to reach the relatively small uncertainties of contemporary experimental data using molecular simulations highly accurate interaction potentials have to be formulated. The leading idea of this project is to develop a new generation of empirical reference equations, based partly on data generated by simulation and to exploit methods used to formulate empirical reference equations in order to refine molecular interaction potentials.