COSMO-RS Aided Design of New, Natural Deep Eutectic Solvents for CO2 Capture

CO₂ capture from power plants is regarded as an effective midterm solution to reduce anthropogenic greenhouse gas emissions. In the case of an integrated gasification combined cycle (IGCC) power plant with precombustion CO₂ capture, the product out of the water–gas shift (WGS) reactors is sent for CO₂ separation. Physical absorption using hydrophobic absorbents with low volatility, low viscosity, and high CO₂ solubility is a promising technique.

In a recent paper, hydrophobic deep eutectic solvents (DESs) have been designed as effective physical solvents for CO₂ capture by using the conductor-like screening model for real solvents (COSMO-RS). Researchers from the TU Eindhoven proved that the ADF-COSMO-RS calculations could reliably predict the CO₂ solubilities in various pure compounds and binary solvents. Through this computational screening, 38 naturally hydrophobic compounds with a high CO₂ affinity have been selected to prepare new DESs. The possibilities for any two components to form a DES are evaluated using their predicted individual activity coefficients through COSMO-RS calculations. The CO₂ solubility in the theoretically feasible DESs is predicted, and the promising DESs like vanillin-4-oxoisophorone (1:3) and methyl salicylate-4-oxoisophorone (1:1) are screened out. The new DESs are experimentally identified to be very promising for precombustion capture, showing comparable CO₂ solubility and viscosity to the well-known, conventionally used Selexol solvent.

The researchers also showed the quantitative structure–property relationship (QSPR) models could accurately predict the density of the non-ionic DESs, with average prediction error less than 10%. Moreover, the viscosity of DESs can also be predicted with QSPR models using COSMO-RS based descriptors.

These findings demonstrate that ADF-COSMO-RS is a strong tool to reduce the time and experimental resources to develop new solvents for gas separation applications.