ChemRxiv
These are preliminary reports that have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information. For more information, please see our FAQs.
1/1
2 files

Developing Machine Learning Models for Ionic Conductivity of Imidazolium-Based Ionic Liquids

preprint
submitted on 21.04.2021, 22:08 and posted on 22.04.2021, 12:56 by Pratik Dhakal, Jindal Shah
In this work, we have developed machine learning models based on support vector machine (SVM) and artificial neural network (ANN) to correlate ionic conductivity of imidazolium-based ionic liquids. The data, collected from the NIST ILThermo Database, spans six orders of magnitude and ranges from 275-475 K. Both models were found to exhibit very good performance. The ANN-model was then used to predict ionic conductivity for all the possible combinations of cations and anions contained in the original dataset, which led to the identification of an ionic liquid with 30% higher ionic conductivity than the highest conductivity reported in the database at 298 K. The model was further employed to predict ionic conductivity of binary ionic liquid mixtures. A large number of ionic liquid mixtures were found to possess non-ideal behavior in that an intermediate mole fraction for such ionic liquid mixtures resulted in either a maximum or minimum in the ionic conductivity.

Funding

National Science Foundation CBET-1706798

History

Email Address of Submitting Author

jindal.shah@okstate.edu

Institution

Oklahoma State University

Country

United State of America

ORCID For Submitting Author

0000-0002-3838-6266

Declaration of Conflict of Interest

A provisional patent application has been filed.

Exports

ChemRxiv

Exports