Density of Deep Eutectic Solvents: The Path Forward Cheminformatics-Driven Reliable Predictions for Mixtures

Abstract

Deep eutectic solvents (DES) are often regarded as greener sustainable alternative solvents and are currently employed in many industrial applications on a large scale. Bearing in mind the industrial importance of DES—and because the vast majority of DES has yet to be synthesized—the development of cheminformatic models and tools efficiently profiling their density becomes essential. In this work, after rigorous validation, quantitative structure-property relationship (QSPR) models were proposed for use in estimating the density of a wide variety of DES. These models were based on a modelling dataset previously employed for constructing thermodynamic models for the same endpoint. The best QSPR models were robust and sound, performing well on an external validation set (set up with recently reported experimental density data of DES). Furthermore, the results revealed structural features that could play crucial roles in ruling DES density. Then, intelligent consensus prediction was employed to develop a consensus model with improved predictive accuracy. All models were derived using publicly available tools to facilitate easy reproducibility of the proposed methodology. Future work may involve setting up reliable, interpretable cheminformatic models for other thermodynamic properties of DES and guiding the design of these solvents for applications.

Predicting the Surface Tension of Deep Eutectic Solvents: A Step Forward in the Use of Greener Solvents

“…Naturally, these validation strategies are only applicable to binary mixtures and require some guidance to follow. Due to the complexity of the data matrices, any random MO- or CO-based division scheme may not yield the most predictive model since variables selection depends largely on the training set [ 28 ]. Even though the CO-based validation is considered to be the most robust strategy [ 36 ], it may give rise to underfitted models with poor statistical quality.…”

“…Even though the CO-based validation is considered to be the most robust strategy [ 36 ], it may give rise to underfitted models with poor statistical quality. At the same time, while the MO-based validation is less robust, this strategy definitely provides more meaningful solutions than any random data distributions or other validation division schemes such as the points-out one proposed by the same authors [ 28 , 36 , 37 , 38 ].…”

“…As referred to above, we have recently developed a Python-based tool named QSAR-Mx [ 28 ], specifically devised to address and automate some crucial steps involved in the QSPR modeling of binary mixtures. A detailed description of the functionalities of this tool can be found in its instruction manual (accessible from , last accessed on 28 April 2022).…”

“…The present work is encouraged by our very recent QSPR modeling efforts on the density of DESs [ 28 ], which yielded statistically more robust models than a thermodynamic model developed with the same dataset [ 29 ]. On one hand, such QSPR efforts demonstrated to offer more options and versatility for setting up predictive models as compared to thermodynamic modeling.…”

“…On the other hand, QSPR models demand several statistical conditions be satisfied just as inspected here as per the guidelines of the Organization for Economic Co-operation and Development (OECD) [ 30 ] to expand their overall applicability as well as statistical reliability [ 31 ]. Moreover, in order to address the requirement of robust validation strategies applicable for the binary mixtures, we have recently designed an open-source standalone Python-based tool “QSAR-Mx” (freely available to download at , last accessed on 28 April 2022) [ 28 ]. This work extensively utilizes such a tool to set up the predictive QSPR models for probing the surface tensions of DESs.…”