The interest in green and sustainable solvents has been dramatically increasing in recent years because of the growing awareness of the impact of classical organic solvents on environmental pollution and human health.

The case of sustainable solvents is of great interest both academically and industrially. With research communities becoming more aware of the negative impacts of conventional organic solvents, a range of greener and more sustainable solvents have been developed to counter the harmful drawbacks associated with conventional solvents. Among these, eutectic solvents (ESs) attracted considerable attention for their "green" properties and have proven their usefulness as environmentally benign alternatives to classical solvents. Among the various desirable characteristics of ESs, pH is a key property with significant implications for the design and control of industrial-scale applications. However, selecting an ES with the required pH for a particular application is a challenging task, especially with extensive experimentally determined data being time consuming and expensive. Therefore, in this work, the pH of various ESs have been predicted via novel quantitative structure−property relationships (QSPR) models using two machine learning algorithms, a multiple linear regression (MLR) and an artificial neural network (ANN), with a set of molecular descriptors generated by COSMO-RS. A total of 648 experimental points for 41 chemically unique ESs prepared from 9 HBAs and 21 HBDs at different temperatures were utilized for sufficient data set representation. On the basis of the statistical analysis of the models, it can be concluded that both approaches can be utilized as powerful predictive tools in estimating the pH of new ESs with the ANN model having better predictive capabilities and the MLR model being more interpretable. These models inspire and stimulate the development of robust models to predict the properties of designer solvents from the drawn molecular structures, which will save time and resources.

This work presents the development of molecular-based mathematical models for the prediction of electrical conductivity of deep eutectic solvents (DESs). Two new quantitative structure–property relationship (QSPR) models based on conductor-like screening model for real solvent (COSMO-RS) molecular charge density distributions (S σ-profiles) were developed using the data obtained from the literature. The data comprise 236 experimental electrical conductivity measurements for 21 ammonium- and phosphonium-based DESs, covering a wide range of temperatures and molar ratios. First, the hydrogen-bond acceptors (HBAs) and hydrogen-bond donors (HBDs) of each DES were successfully modeled using COSMO-RS. Then, the calculated S σ-profiles were used as molecular descriptors. The relation between the conductivity and the descriptors in both models has been expressed via multiple linear regression. The first model accounted for the structure of the HBA, the HBD, the molar ratio, and temperature, whereas the second model additionally incorporated the interactions between the molecular descriptors. The results showed that by accounting for the interactions, the regression coefficient (R 2) of the predictive model can be increased from 0.801 to 0.985. Additionally, the scope and reliability of the models were further assessed using the applicability domain analysis. The findings showed that QSPR models based on S σ-profiles as molecular descriptors are excellent at describing the properties of DESs. Accordingly, the obtained model in this work can be used as a useful guideline in selecting DESs with the desired electrical conductivity for industrial applications.