No abstract

A predictive model is developed by using a group contribution method for the estimation of the infinite dilution activity coefficient (IDAC) of methanol in ionic liquids (ILs). A simple van’t Hoff model consists of three parameters is used to calculate the value of IDAC using a multiple linear regression (MLR) method and optimized by a generalized reduced gradient (GRG) nonlinear algorithm in order to obtain a similar value from both experimental and predicted data points. An absolute average relative deviation (AARD) is used to quantify the percentage of deviation between predicted and experimental values. The predicted model developed from the training set shows an error of 11.43%. The predicted model is then validated using experimental data points from several kinds of literature. Thus, an overall AARD value of 11.97% is obtained from the combination of the training and validation set. Imidazolium, pyridinium, pyrrolidinium, ammonium, and phosphonium are the common cations studied in this work as they are often used and commercialized especially in synthesizing ILs for several other applications. The predicted IDAC values are compared with the experimental values based on the anion, cation headgroup, and alkyl chain length. The observation also confirms that hydrogen-bonding basicity of anion plays a crucial role in their interaction with methanol. With respect to the cation headgroup and alkyl chain length variation, there is no significant difference in IDAC values. These results are demonstrated here to help understand the interactions between IL and methanol. The effect of various structural features of ILs on IDAC values can be observed, which aids in the development of various steps for the design of most suitable ILs with improved interaction with methanol.

No abstract

No abstract