Key Factors of Uniform Polarization Reversal Barrier in Wurtzite Materials Utilizing Machine Learning Methods

Area of Science:

• Materials Science
• Solid State Physics
• Computational Materials Science

Background:

• Scandium-doped aluminum nitride (wurtzite structure) shows ferroelectric promise but faces challenges due to high coercive voltage.
• Understanding polarization reversal is crucial for reducing coercive voltage and enabling applications.
• Current knowledge lacks a unified set of factors influencing polarization reversal in these materials.

Purpose of the Study:

• To develop machine-learning models for predicting the uniform polarization reversal barrier (Eua) in wurtzite materials.
• To identify key intrinsic factors governing polarization reversal processes.
• To provide insights for reducing coercive voltage in ferroelectric materials.

Main Methods:

• Machine learning regression models were employed.
• Data sets included 41 binary and 113 simple ternary wurtzite materials.
• Features included elemental, crystal, mechanical, and electronic properties; Eua calculated via first-principles methods.

Main Results:

• Average cation-ion potential identified as the primary intrinsic factor influencing Eua.
• Secondary impacts observed from relative cation-anion height ratio, cell parameter ratio, and average cation Mendeleev number.
• Models successfully predicted Eua across diverse wurtzite material systems.

Conclusions:

• This study systematically evaluates factors affecting uniform polarization reversal barrier (Eua) in wurtzite materials.
• Identifies key predictors for Eua, advancing understanding beyond single material systems.
• Offers a foundation for designing ferroelectric materials with tailored coercive voltages.