Predicting Ion Diffusion from the Shape of Potential Energy Landscapes.

Area of Science:

• Materials Science
• Computational Chemistry
• Solid-State Physics

Background:

• Predicting ion diffusion in crystalline solids is crucial for developing advanced energy storage materials.
• Current methods like molecular dynamics are computationally expensive, limiting large-scale screening.
• Understanding diffusion mechanisms requires accurate calculation of diffusion coefficients.

Purpose of the Study:

• To present an efficient computational method for determining diffusion coefficients in multiparticle systems.
• To validate the method by applying it to lithium-ion diffusion in inorganic crystalline solids.
• To establish a faster and transferable workflow for screening solid-state ion conductors.

Main Methods:

• The method computes diffusion coefficients directly from the geometry and topology of the potential energy field.
• It analyzes the potential energy landscape of migrating particles in crystalline structures.
• The approach was tested on lithium-ion diffusion in various inorganic crystalline solids.

Main Results:

• The method predicts lithium-ion diffusion coefficients within one order of magnitude of molecular dynamics simulations.
• The new approach is several orders of magnitude faster than traditional molecular dynamics.
• The workflow demonstrated high speed and transferability for screening applications.

Conclusions:

• The developed method offers an efficient and accurate way to compute diffusion coefficients.
• Its speed and transferability make it ideal for screening solid-state ion conductor candidates.
• This platform promises to advance diffusion prediction capabilities, even at the density functional theory level.