(772a) New Experimental Data and Physico-Chemical Model for the Solubility of Carbon Dioxide in Aqueous Solutions of Monoethanolamine

The most widely used solvent for Post Combustion Carbon Capture (PCC) is a 30 mass % aqueous solution of monoethanolamine (MEA). That solvent is generally accepted as a reference to which new PCC solvents are to be compared. [1, 2] Quite a large amount of experimental data is available for the vapor-liquid equilibrium (and related physico-chemical properties) such of such mixtures and one might assume that there is little need for new data. However, a closer look reveals that literature data for the solubility of carbon dioxide in a 30 mass % aqueous solution of MEA scatter widely. Even when obvious outliers are removed, the remaining uncertainty is so large that results of process simulations can be widely tuned simply by selecting certain data sets to which models are fitted.

Therefore in the present work, a comprehensive study on the solubility of carbon dioxide in aqueous solutions of MEA was carried out using two different experimental techniques: at low partial pressures of carbon dioxide (from 5 to 80 kPa) a headspace gas chromatographic technique (HS-GC) was used whereas at high pressures (from 0.4 to 8.5 MPa) a synthetic method was applied. Both apparatuses have been used in our laboratory for more than twenty years in many studies and are known to yield very reliable and accurate data. Aqueous MEA solutions with 15 and 30 mass % MEA were investigated at 313 K, 353 K and 393 K. The investigated pressure range corresponds to a molar ratio of carbon dioxide to MEA in the liquid from 0.1 to 1.3.

A physico-chemical model was developed to describe the new experimental data. Such modeling is challenging as it has to take simultaneously into account the complex chemical reactions and the strong physical non-ideality of the studied reactive electrolyte solutions. Five reactions are included: the autoprotolysis of water (I), the formation of bicarbonate (II) and carbonate (III), the protonation of MEA (IV) and the formation of carbamate (V). The extended Pitzer model is used to describe the liquid phase non-ideality. The equilibrium constants of reactions (I) – (V) were taken from the literature. The results are checked using NMR spectroscopic literature data on the speciation. [3, 4, 5]

Only a small number of parameters for interactions between the dominant species, are fitted. The model gives a good description of the entire experimental gas solubility data set as well as a good agreement with literature data for the liquid phase speciation (from NMR investigations). 

Besides presenting the new experimental and modeling results, we will also briefly review of the previous approaches both regarding the experiments and the modeling and we will discuss the important impact of the physico-chemical data on the process simulation.

[1]       Mangalapally, H. P.; Hasse, H., Chem. Eng. Res. Des. 2011, 89, 1216-1228.

[2]       Wang,M.; Lawal, A.; Stephenson, P.; Sidders, J.; Ramshaw, C.,Chem. Eng. Res. Des. 2011, 89, 1609-1624.

[3]       Hilliard, M. D. A, Ph.D. thesis, University of Texas at Austin, 2008.

[4]       Böttinger, W.; Maiwald, M.; Hasse, H., Fluid Phase Equilib. 2008, 263, 131-143.

[5]       Jakobsen, J. P.; Krane, J.; Svendsen, H. F., Ind. Eng. Chem. Res. 2005, 44, 9894-9903.