Machine-Learning-Designed BCZT-SBT Heterointerface Unlocks Fatigue-Resistant Energy Storage.

Area of Science:

• Materials Science
• Energy Storage
• Dielectric Materials

Background:

• Dielectric capacitors offer fast energy storage but suffer from fatigue at high electric fields.
• Developing fatigue-free dielectrics is crucial for advanced energy storage applications.

Purpose of the Study:

• To design and synthesize a fatigue-free lead-free dielectric material for enhanced energy storage.
• To investigate the atomic-scale mechanisms behind fatigue resistance in novel dielectric compositions.

Main Methods:

• Utilized machine learning (ML) to guide the composition of a (1-x)Ba_0.85Ca_0.15Zr_0.1Ti_0.9O₃-(x)SrBi₂Ta₂O₉ solid solution.
• Synthesized the material and performed atomic-scale analyses to understand interfacial properties and phase coexistence.
• Characterized energy storage performance, including energy density, efficiency, and charge-discharge cycle stability.

Main Results:

• A (1-x)BCZT-xSBT solid solution with x = 0.10 exhibited a unique epitaxial interfacial layer due to phase coexistence (perovskite and tungsten bronze).
• This hetero-barrier effectively suppressed carrier migration and promoted polarization homogenization, enhancing voltage endurance.
• Achieved a recoverable energy density of 9.94 J cm^-3 with 92.1% efficiency and demonstrated stable performance over 10⁹ cycles without degradation.

Conclusions:

• The developed lead-free dielectric exhibits remarkable fatigue resistance attributed to an elevated Schottky barrier and interfacial engineering.
• This work presents a ML-guided strategy for designing high-performance, fatigue-free dielectric capacitors for reliable energy storage.