Improved manometric setup for the accurate determination of supercritical carbon dioxide sorption.

An improved version of the manometric apparatus and its procedures for measuring excess sorption of supercritical carbon dioxide are presented in detail with a comprehensive error analysis. An improved manometric apparatus is necessary for accurate excess sorption measurements with supercritical carbon dioxide due to the difficulties associated with the high sensitivity of density for pressure and temperature changes. The accuracy of the apparatus is validated by a duplicate measurement and a comparison with literature data. Excess sorption and desorption of CO(2) on Filtrasorb 400 at 318.11 K up to 17 069 mole/m(3) (15.474 MPa) were selected for this validation. The measured excess sorption maximums are 7.79+/-0.04 mole/kg at 2253 mole/m(3) for the first sorption isotherm and 7.91+/-0.05 mole/kg at 2670 mole/m(3) for its subsequent desorption isotherm. The sorption and desorption peaks of the duplicate experiment are 7.92+/-0.04 mole/kg at 2303 mole/m(3) and 8.10+/-0.05 mole/kg at 2879 mole/m(3), respectively. Both data sets show desorption data being higher than the sorption data of the same data set. The maximum discrepancy between the desorption and sorption isotherms of one data set is 0.15 mole/kg. The discrepancy between the two excess sorption isotherms is 0.12 mole/kg or less. The a priori error of the excess sorption measurements is between 0.02 and 0.06 mole/kg. The error due to He contamination is between 0.01 and 0.05 mole/kg. The difference between the a priori uncertainty and the observed maximum discrepancies is considered to be acceptable. The sorption isotherms show identical qualitative behavior as data in the literature. The quantitative behavior is similar but the peak height and the linear decrease in excess sorption at high gas densities are 10% higher. A plot of the excess sorption versus the density can be used to obtain the sorbed phase density and the specific micropore volume. These sorbed phase densities are in excellent agreement with the data in the literature. Furthermore, the excess sorption data scaled to this specific micropore volume in this work and in the literature collapse on a single curve when plotted versus gas density.