Excess density as a descriptor for electrolyte solvent design.

Area of Science:

• Electrochemistry
• Materials Science
• Computational Chemistry

Background:

• Electrolytes are crucial for battery performance, mediating cathode-anode interactions.
• Multi-component solvent mixtures are used to tailor electrolyte properties like viscosity and boiling point.
• Linear mixing approximations often fail, showing significant deviations in real electrolyte mixtures.

Purpose of the Study:

• Investigate excess density in common Li-ion battery solvents and salt-containing electrolytes.
• Develop a predictive model for electrolyte property deviations based on molecular structure.
• Validate simulation findings with experimental data.

Main Methods:

• Molecular dynamics simulations of various solvent mixtures (carbonates, ethers, nitriles) and electrolytes.
• Fitting excess density data to Redlich-Kister polynomials.
• Experimental validation using an automated test stand (Clio).
• Quantifying molecular structure similarity using smooth overlap of atomic position (SOAP) fingerprints.

Main Results:

• Excess density varies smoothly with mole percent and is influenced by solvent type.
• Mixtures of similar solvents exhibit lower magnitude excess properties than dissimilar mixtures.
• Simulation trends are validated by experimental results.
• A structure-based descriptor for excess density was developed.

Conclusions:

• Excess quantities provide insights into molecular interactions within electrolyte mixtures.
• Structural dissimilarity of components can predict deviations from ideal electrolyte behavior.
• This approach enables the rational design of high-performance battery electrolytes.