A Computational Protocol Combining DFT and Cheminformatics for Prediction of pH-Dependent Redox Potentials

Area of Science:

• Computational Chemistry
• Materials Science
• Electrochemistry

Background:

• Developing new materials for energy storage is crucial for efficient energy solutions.
• Organic electrolytes for redox flow batteries require accurate prediction of redox potentials.
• Predicting properties of unknown compounds necessitates reliable computational protocols.

Purpose of the Study:

• To present and validate a robust procedure for calculating redox potentials of organic molecules at any pH.
• To compare different computational methods for reaction free energies, solvation, and pH effects.
• To develop a more realistic Pourbaix diagram by accounting for pH-dependent protonation states.

Main Methods:

• Utilized widely available quantum chemistry and cheminformatics methods.
• Employed the B3LYP hybrid functional with the COSMO solvation method.
• Incorporated thermal contributions from BLYP gas-phase harmonic frequencies.
• Proposed an improved Alberty-Legendre transform for pH-dependent Pourbaix diagrams.

Main Results:

• Achieved good prediction of pH = 0 redox potentials at moderate computational cost.
• Identified B3LYP/COSMO with BLYP thermal contributions as a reliable method.
• Demonstrated the ability to predict pH-dependent redox potentials accurately.

Conclusions:

• The developed computational procedure offers a reliable and efficient protocol for predicting organic molecule redox potentials.
• This method facilitates the discovery of novel organic electrolytes for redox flow batteries.
• The improved Pourbaix diagram construction provides a more realistic representation of electrochemical behavior across pH ranges.