3T-VASP: fast ab-initio electrochemical reactor via multi-scale gradient energy minimization

Area of Science:

• Computational Chemistry
• Materials Science
• Electrochemistry

Background:

• Ab-initio methods like density functional theory (DFT) are crucial for atomistic studies but are computationally intensive for discovering rare electrochemical reaction byproducts.
• The scalability of DFT is limited by the high number of steps required for complex reaction pathway exploration.

Purpose of the Study:

• To develop a more efficient in-silico method for generating numerous elementary electrochemical reaction byproducts.
• To reduce the computational cost associated with ab-initio simulations for byproduct discovery.

Main Methods:

• Implementation of a multi-scale approach using the tiered tensor transform (3T) method.
• Utilizing a small number of ab-initio energy minimization steps for byproduct generation.
• Demonstration on organic molecule passivation on perovskite solar cells and complex liquid electrolytes for lithium-ion batteries.

Main Results:

• Successfully generated many elementary electrochemical reaction byproducts in-silico with significantly fewer DFT steps (50-100) compared to traditional ab-initio molecular dynamics (>10,000 steps).
• Validated the method's applicability on diverse systems, including perovskite solar cells and lithium-ion battery electrolytes, with numerous byproducts confirmed by prior experimental studies.
• The method requires no machine learning training data and is directly applicable to new chemistries.

Conclusions:

• The multi-scale 3T method offers a computationally efficient alternative for ab-initio investigation of elementary chemical reaction byproducts.
• This approach enhances the scalability of DFT for discovering rare byproducts in various electrochemical systems.
• The method is suitable for applications where temperature dependence is not a primary concern.