Approche multi-échelles pour l’estimation des propriètés des liquides ioniques

Abstract

An important properties of Ionic Liquids are Thermal ( Melting point (Tm), glass transition temperature(Tg), Decomposition temperature(Td), Freezing point (Tf)), Thermophysical (Conductivity (K), heat Capacity(Cp), Surface tension (σ) and heat fusion (ΔHm), Environmental (Toxicity (EC50) and partition coefficient octanol/water (logp), electrochemical (Cathode limit, Anode limit and Electrochemical window) and Photochromic (ET(30)), There are excellent candidates for a number of solvent extraction applications. In this study, the proposed method was applied using the group-interaction contribution (GIC) approach previously developed by Pardillo and Gonzalez-Rubio (Pardillo and González-Rubio,1998) as well as Marrero and Pardillo (Marrero and Pardillo,1999). This concept was adopted in this study to avoid the complexity of the property estimation which would be engendered by the incorporation of a multi-order approach. the first order is associated with interaction contributions from simple groups (e.g.: CH3- and -CH2-) and interactions between a cation and anion (e.g.: interactions of single groups >N<+ and Cl-) and the second order uses binary interaction contributions between bonding groups (e.g.: -CN & -N-- & -CN), In addition, correction factor as a third degree. It was developed using a large database (more 2400 data points) associated with a wide variety of 15 cations (Imidazolium[IM], Guanidinium [Guan], Triazolium[Taz], Sulfonium[S]...) and 40 anion ((Hexafluorophosphate [PF6]), (Tetrafluoroborate [BF4]),…). The results show that, this method represents an excellent alternative to previous approaches for the estimation of the Td, Tg, Tm, Tf, logp, EC50, K, ET(30), Elec.w of diverse ionic liquids with an average absolute relative deviation (AARD%) of 4,2 ; 3,06; 4,09; 5,04; 8,60; 5,43; 4,77; 1,12 and 5,07 respectively and a correlation coefficient ( R2) of 0,866; 0,96; 0,849; 0,933; 0,96; 0,908; 0,999; 0,94 and 0,93 respectively . More importantly, the new model accounts for isomers distinguishing, which represents an advantage over the conventional group contribution methods reported in the literature.