Heat capacity estimation of complex materials for energy technologies.

Area of Science:

• Materials Science
• Computational Physics
• Thermodynamics

Background:

• Heat capacity is crucial for engineering design, influencing free energy and thermal transport.
• Traditional models like Debye and Dulong-Petit lack accuracy due to simplified assumptions.
• Advanced considerations such as anharmonicity and electronic carriers are needed for improved heat capacity predictions.

Purpose of the Study:

• To develop a physics-based model for accurate heat capacity estimation.
• To integrate vibrational, dilation, and electronic contributions into a unified model.
• To compare different methods for determining phonon density of states.

Main Methods:

• Developed the vibrational + dilation + electronic (VDE) model.
• Incorporated user-provided phonon density of states (from analytical, machine-learned, and first-principles methods).
• Included a phonon pressure-based dilation term and an electronic component.

Main Results:

• Machine-learned phonon density of states demonstrated significant advantages.
• The VDE model achieved heat capacity estimates within 5% of experimental values for 38 materials (200-600 K).
• Accurate modeling of temperature-dependent heat capacity and phase transitions (e.g., Cu2Se) was demonstrated.

Conclusions:

• The VDE model offers a significant improvement over traditional methods for heat capacity prediction.
• Machine learning enhances the accuracy of phonon density of states calculations.
• The model's accuracy enables sophisticated material analysis and supports engineering design.