In silico Design, Virtual Screening and Synthesis of Novel Electrolytic Solvents

Area of Science:

• Electrochemistry
• Materials Science
• Computational Chemistry

Background:

• Next-generation lithium-ion batteries require improved electrolytes for enhanced performance and safety.
• Developing new solvents with optimal properties is crucial for battery advancement.
• Existing methods for solvent discovery are often time-consuming and resource-intensive.

Purpose of the Study:

• To build and validate Quantitative Structure Property Relationship (QSPR) models for predicting electrolyte properties.
• To guide the rational design and synthesis of novel solvents for lithium-ion batteries.
• To accelerate the discovery of high-performance battery materials.

Main Methods:

• Compiled a dataset of 103 electrolytes with properties like oxidation potential, ionic conductivity, melting, and boiling points.
• Developed consensus QSPR models using Support Vector Machine (SVM) algorithms and ISIDA fragment descriptors.
• Implemented models in ISIDA/Predictor software for virtual screening of chemical libraries.

Main Results:

• Successfully built and validated QSPR models for key electrolyte properties.
• Screened a virtual library of 9965 esters and sulfones.
• Identified and synthesized promising candidate solvents based on model predictions.

Conclusions:

• QSPR modeling is an effective strategy for accelerating the design of advanced battery electrolytes.
• The developed models provide a valuable tool for predicting solvent properties and guiding experimental synthesis.
• This approach facilitates the discovery of novel materials for next-generation lithium-ion batteries.