Quantifying and Interpreting Solvation Power of Cyclic Carbonate by Chemical Calculation and Machine Learning

Area of Science:

• Electrochemistry
• Materials Science
• Computational Chemistry

Background:

• Ion dynamics in electrolytes are crucial for battery performance.
• Solvation power of cyclic solvents is key but difficult to quantify due to complex electronic and steric effects.

Purpose of the Study:

• To decouple electronic and steric influences on solvation power in cyclic carbonate electrolytes.
• To develop a quantitative framework for predicting solvation power based on molecular descriptors.

Main Methods:

• Defined descriptors for functional group nature (atom charges, electron localization function) and structural adaptability (substituent volume and distance).
• Employed mathematical fitting and machine learning (ML) to correlate descriptors with solvation power.
• Investigated underlying mechanisms through integrated computational and analytical approaches.

Main Results:

• Quantified the contributions of functional group properties and structural adaptability to solvation power.
• Demonstrated that both factors significantly and comparably influence solvation behavior.
• Challenged the traditional view that functional groups solely dominate solvation.

Conclusions:

• Established a chemical foundation for rationally selecting cyclic carbonate-based electrolytes.
• Provided a predictive model for optimizing electrolyte properties for advanced battery chemistries.
• Highlighted the importance of structural adaptability alongside electronic factors in solvent design.