(103c) Prediction of Transport Properties of Hydrogen Bonding Liquids by Molecular Simulation

The success of process design in chemical engineering depends on the availability and accuracy of thermodynamic properties. Particularly the prediction of transport properties of liquids is of great interest since the experimental data base is usually insufficient and classical prediction methods are frequently unsatisfactory. In recent years, molecular simulation has become a promising tool to accurately predict transport properties of fluids. In preceding work of our group [1,2], it was shown that transport properties of simple fluids can be accurately predicted with molecular models that were adjusted to vapour-liquid equilibrium data only. In the present work, that approach is extended to highly polar and hydrogen bonding fluids: ammonia [3], monomethylamine [4], dimethylamine [4], methanol [5], ethanol [6], and water [7]. The pure fluids are simulated for a wide range of thermodynamic conditions. Furthermore, some transport properties of binary mixtures containing both alcohols and water are also predicted in this work. The transport properties self- and Maxwell-Stefan diffusion coefficients as well as the shear viscosity are determined by equilibrium molecular dynamics and the Green-Kubo formalism. The thermal conductivity is obtained by reverse boundary driven non-equilibrium molecular dynamics [8]. The molecular models of the studied fluids are rigid, non-polarizable and composed of a set of Lennard-Jones sites with superimposed point charges. With the exception of water, all models were developed by our group exclusively on the basis of quantum-chemical information and experimental vapor-liquid equilibrium data. In case of water, the TIP4P_2005 molecular model from the literature [7] was used. Throughout this work, it is shown that the predicted transport properties from molecular simulation agree well with the experimental data as far as it is available. E.g., the transport properties are predicted with average deviations to experimental values ranging from 5 to 17%.

References: [1] G. A. Fernandez, J. Vrabec and H. Hasse, Int. J. Thermophys. 2005, 16, 1389. [2] G. A. Fernandez, J. Vrabec and H. Hasse, Mol. Sim. 2005, 31, 787. [3] B. Eckl, J. Vrabec, and H. Hasse, Mol. Phys. 2008, 106, 1039. [4] T. Schnabel, J. Vrabec, and H. Hasse, Fluid Phase Equilib. 2008, 263, 144. [5] T. Schnabel, J. Vrabec, and H. Hasse, Fluid Phase Equilib. 2005, 233, 134. [6] T. Schnabel, A. Srivastava, J. Vrabec, and H. Hasse, J. Phys. Chem. B 2007, 111, 9871. [7] J.L.F Abascal and C. Vega, J. Chem. Phys. 2005, 123, 234505. [8] F. Müller-Plathe, J. Chem. Phys. 106, 1997, 6082.