Spatially resolved deep-ultraviolet (UV) Raman spectroscopy was applied to solutions of CO2 and H2O or D2O subject to a temperature gradient in a thermally regulated high-pressure concentric-tube Raman cell in an attempt to measure a Soret effect in the vicinity of the critical point of CO2. Although Raman spectra of solutions of CO2 dissolved in D2O, at 10 MPa and temperatures near the critical point of CO2, had adequate signal-to-noise and spatial resolution to observe a Soret effect with a Soret coefficient with magnitude |ST| > 0.03, no evidence for an effect of this size was obtained for applied temperature gradients up to 19 °C. In contrast, the concentration of CO2 dissolved in H2O was shown to vary significantly across the temperature gradient when excess CO2 was present, but the results could be explained simply by the variation in CO2 solubility over the temperature range and not by kinetic factors. For mixtures of D2O dissolved in scCO2 at 10 MPa and temperatures close to the critical point of CO2, the Raman peaks for D2O were too weak to measure with confidence even at the limit of D2O solubility.
Charles F. Windisch, Gary D Maupin, and B. Peter McGrail, "Ultraviolet (UV) Raman Spectroscopy Study of the Soret Effect in High-Pressure CO2-Water Solutions," Appl. Spectrosc. 66, 731-739 (2012)