IntroductionRecently, the chromatographic technique has been a focus of attention as an effective tool to measure various thermodynamic properties of fluids, 1 such as the diffusion coefficient, vapor pressure, partition coefficient, solubility and adsorption isotherm. As well as the properties under ambient conditions, the chromatographic technique is useful for the measurements in supercritical fluids.The authors have applied the technique of supercritical fluid chromatography, referred to as the chromatographic impulse response (CIR) method, to simultaneous measurements of the binary diffusion coefficient, D12, and the retention factor, k, in supercritical fluids.
2-5This technique is a type of transientresponse method in which a solute species is pulse-injected to a polymer-coated diffusion column, and the two parameters are determined by fitting the calculated response curve to the measured one. The Taylor dispersion method, which is also a type of transient-response method, has been commonly employed to measure D12 values in the gaseous, supercritical and liquid phases with an uncoated diffusion column. In particular, most D12 data in supercritical fluids were measured by the Taylor dispersion method. However, this method is not adequate for low-volatility and/or polar compounds.
2The authors measured the D12 and k values for a variety of fatty acids and their derivatives, i.e. fatty acids having carbon numbers of 18 to triglycerides having molecular weights, M, of over 1000, in supercritical carbon dioxide by the CIR method. [2][3][4][5] However, accurate data for compounds having relatively short chain lengths are scarcely available in the literature. Thus, in this study the CIR method was employed to measure the binary diffusion coefficients and retention factors for myristoleic acid and its methyl ester in supercritical carbon dioxide, and the partial molar volumes were estimated from the retention factor and CO2 density. The validities of the predictive correlations for the D12 values were also tested.
TheoryThe theory for the CIR technique was described previously in detail, 2,4,5 and is briefly mentioned. When a tracer is pulseinjected into a fully developed laminar flow in a coated cylindrical column, the cross-sectional average concentration, Ca(t), at the outlet of the column is given aswhereand m is the amount of the tracer species, R the radius, t the time, L the distance from the inlet of the diffusion column, ua the average fluid velocity, k the retention factor, and D12 the infinite dilution binary diffusion coefficient of the tracer species, respectively. When ua is measured, D12 and k can be simultaneously determined by minimizing the root-mean-square (rms) error, ε, defined as follows for the measured Cexp(t) and calculated Ca(t) curves: