The Design of Supercritical Fluid Processes: Availability and Shortcomings of Thermodynamic and Transport Data

Supercritical fluids (SCFs) show promise in the separation of high molecular mass compounds as they are able to distinguish between the hydrocarbon backbone length and the type and position of the functional group present. In order to design such a separation process one needs to estimate the number of theoretical stages required and determine the column diameter and height, both of which require accurate models, which in turn needs to be validated against experimental data. The aim of this paper is to evaluate the availability of the phase behaviour as well as thermodynamic and transport data required for the design of a SCF separation process. Additionally, the extent to which these properties can be predicted will be investigated. For binary systems containing CO2 some phase behaviour data is available yet in many cases additional measurements may be required. For binary systems containing other SCFs and for ternary and multi-component systems data is considerably more scarce and additional measurements are required. Additionally, due to the complexity of the interaction present, accurate thermodynamic modelling relies on mathematically complex equations of state (EOSs) that, to date, are not widely implemented. The thermodynamic data (enthalpy, heat of mixing, density etc.) are also required. In many cases, the density is measured together with the phase equilibria yet the other thermodynamic properties have only been measured for a few compounds highlighting a need for additional measurements. In theory, thermodynamic properties can be estimated using an EOS, yet, with possible exception of the density, EOSs generaly provide an order of magnitude estimation rather than an accurate description of the thermodynamic properties. Transport property data (viscosity, interfacial tension, diffusivity etc.) are used to determine the column height and diameter. Transport properties data is quite scarce, especially for mixtures and estimation techniques are not well developed.