Aqueous polyalkylated imidazoles have gained interest as potential CO2 capture solvents due to their high oxidative stability and low vapor pressures compared to traditional amines. In this work, 21 aqueous solutions of polyalkylatedimidazoles were screened as absorbents for CO2 capture and four solvent candidates were further characterized by measuring the vapor-liquid equilibria and the heat of absorption of CO2. The pKa values of the imidazoles were measured and a positive correlation between the absorption capacity and pKa of polyalkylated imidazoles was found. Increasing the pKa of imidazoles to 9 by alkylation improved the CO2 absorption capacity significantly. Based on the equilibrium experiments, the cyclic capacities of the selected solvents varied from 0.8 to 2 mol CO2/kg solvent. Furthermore, the heat of absorption of CO2 of the studied imidazoles was lower compared to primary amines. In general, the tested polyalkylated imidazoles are more feasible for processes with partial pressures of CO2 above 50 kPa. Trimethylimidazole that forms bicarbonate precipitate might be applicable for post combustion CO2 capture as a high cyclic capacity is obtained even at CO2 partial pressures around 10 kPa. The collected data indicates that the absorption rate of imidazoles is comparable to tertiary amines and promoters would most likely be required. The present study gives new important knowledge of the absorption properties of polyalkylated imidazoles.

The melting point (T m ) of an ionic liquid (IL) is of crucial importance in many applications. The T m can vary considerably depending on the choice of the anion and cation. This study explores the use of various machine learning (ML) methods to predict the melting points (−96 • C -359 • C range) of structurally diverse 2212 ILs based on a combination of 1369 cations and 141 anions. Among the ML models applied to independent training and test sets, tree-based ensemble methods (Cubist, random forest and gradient boosted regression) were found to demonstrate slightly better performance over support vector machines and k-nearest neighbour approaches. In comparison, quantum chemistry based COSMOtherm predictions were generally found to have significant deviations with respect to the experimental values. However, classification models were more efficient in discriminating between ILs with T m > 100 • C and those below 100 • C.

A computational screening strategy applied to 8 million synthetically diverse ionic liquids, demonstrates its value for sustainable solvent design.

No abstract