Industrial & Engineering Chemistry Research, Vol.46, No.21, 6998-7007, 2007
Accurate global thermophysical characterization of hydrofluoroethers through a statistical associating fluid theory variable range approach, based on new experimental high-pressure volumetric and acoustic data
In this work, the ability of a recently proposed statistical associating fluid theory of variable range (SAFT VR) version to estimate thermophysical properties of fluorinated compounds is demonstrated, focusing specially on second-order derivative properties. These properties are fundamental in the simulation and design of the industrial processes where these families of compounds find application as working fluids, but only very recently has their estimation been accomplished with the desired degree of accuracy, as they have been traditionally considered as a severely stringent test to any thermodynamic model. With this aim, a complete thermophysical characterization of two hydrofluoroethers, included among the so-called third generation of chlorofluorocarbon alternatives, with low global warming potential and zero ozone depletion potential, are presented. To achieve this goal, measurements of compressed-liquid densities and the speed of sound were performed in high-pressure conditions and set as the basis for the determination of molecular parameters of a version of the statistical associating fluid theory for chain molecules with attractive potentials of variable range (SAFT VR) equation of state. This coupling between accurate experimental determination of high pressures, density, and the speed of sound and the calculations of characteristic molecular parameters in the framework of a physically sound molecular model allows a complete description of the thermophysical behavior of the pure fluids studied, providing a precise simultaneous estimation of phase equilibria and other first- and second-order derivative properties, ensuring the reliability of the proposed characterization procedure. The relevance of the accurate experimental data is to be emphasized, as without the right experimental input, the applied model, accurate as it eventually is shown to be, would never achieve its maximum performance, as it will be discussed in the following.