Efficient calculations of impurity diffusivity in metals by linearized multi-band embedded atom method potentials.

Area of Science:

• Materials Science
• Computational Materials Science
• Chemical Engineering

Background:

• Impurity diffusivity in metals is critical for material performance and integrity.
• Traditional molecular dynamics (MD) simulations using empirical potentials (e.g., EAMs) lack accuracy, especially for isotope effects.
• Machine learning (ML) potentials offer high accuracy but are computationally expensive.

Purpose of the Study:

• To develop an accurate and computationally efficient method for simulating impurity diffusivity in metals.
• To introduce and validate the linearized multi-band EAM (LMB-EAM) for impurity diffusion calculations.
• To provide guidance on constructing accurate potential models for solid metals.

Main Methods:

• Developed an extended embedded atom method (EAM) potential, termed linearized multi-band EAM (LMB-EAM).
• Constructed LMB-EAMs using a force-matching method with first-principles calculation data and regularization.
• Validated LMB-EAM performance for hydrogen (H) diffusion in body-centered cubic tungsten (bcc-W) and oxygen (O) diffusion in liquid sodium (Na).

Main Results:

• LMB-EAMs were constructed using minimal training data.
• LMB-EAMs demonstrated superior performance compared to traditional empirical potentials.
• Calculated diffusion coefficients, including isotope effects, with reasonable accuracy, outperforming empirical potentials and approaching ML potential accuracy.

Conclusions:

• LMB-EAM provides a balance between accuracy and computational efficiency for impurity diffusivity simulations.
• The method accurately captures isotope effects in impurity diffusion.
• Findings guide the development of improved potential models for materials simulations.