Development of a temperature dependent 2D-QSPR model for viscosity of diverse functional ionic liquids

Ionic liquids (ILs) have significant uses in academic research and industrial process development. The tunable physicochemical properties like low vapor pressure, exceptional solvating properties, low inflammability etc. make them eco-friendly and green solvents. For industrial or laboratory purposes, it is very difficult to operate with high viscous ILs. Therefore, in this work, our attention is focused on modeling of viscosity of ILs using quantitative structure-property relationship (QSPR), in order to provide essential information to chemists enabling design of lower viscous ILs for addressing operational problems. In the present study, we have developed a QSPR model using a pool of 291 ILs using easily computable two-dimensional descriptors. Different strategies were applied to select appropriate variables for development of the final partial least squares (PLS) regression model. The model was validated extensively using different validation metrics to check the robustness and predictivity of the model for regulatory acceptance and enhancing confidence in QSPR predictions. The results showed that the model was statistically robust. Based on the insights obtained from the PLS model as mapped by the loading plot, scores plot, regression coefficient plot etc., we have concluded that intrinsic properties of cations, number of path count of order three at the anionic part, temperature, cation containing branched alkyl chain with large numbers of heteroatoms, shape of anions (spherical or rod), presence or absence of CH2RX fragment, presence or absence of quaternary nitrogen at the cationic part, aliphatic keto group at the anionic part, a CS bond at topological distance of 4 at the cationic part and presence or absence of a pyrrolidine ring at the cationic part of ILs play crucial roles to regulate the viscosity of ILs.