High‐Density Energy Storage in Si‐Integrated Dielectric Multilayers Through Interfacial Polarization and ML‐Guided Structural Optimization
Year: 2026
Type: article
Abstract: ABSTRACT Epitaxial dielectric thin films have attracted considerable attention for better energy storage performance than bulks and polymers, yet their reported recoverable energy density ( W rec ) commonly excludes the substrate, which limits its relevance to practical device‐level energy storage. Here, we develop a Si‐integrated BZT//S‐BZT multilayer capacitor to address this issue and simultaneously enhance polarization and breakdown strength ( E b ). By directly integrating the oxide multilayer on Si‐based substrates, the contribution of the substrate to the effective device volume is substantially reduced, enabling a realistic energy density ( W rec‐real ) of 105.51 J/cm 3 , exceeding that of the overwhelming majority of comparable systems. The BZT and S‐BZT sublayers exhibit small lattice mismatch but pronounced dielectric contrast, which induces strong interfacial polarization ( P inter ) without introducing interfacial structural perturbation. Landau–Ginzburg–Devonshire theory and phase‐field simulations consistently verify the interface‐induced polarization enhancement and agree well with the experimental observations. To further improve E b , machine learning was employed to establish the relationship between the multilayer structural parameters and E b and to identify the optimal starting‐layer thickness. Experimental results closely reproduce the predicted trend, leading to an optimized Big‐Mac‐ n architecture with a high W rec of 131.21 J/cm 3 and an E b of 6.5 MV/cm. The multilayer also exhibits excellent fatigue, frequency, and temperature stability, together with ultrafast discharge capability. This work establishes a data‐driven interface‐engineering strategy for reliable, Si‐integrated dielectric energy‐storage capacitors.
Source: Advanced Energy Materials
DOI: 10.1002/aenm.71783
Authors: Jun Yang, Hongyu Yang, Zixiong Sun
Language: English
Subfield: Materials Chemistry
Field: Materials Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W7224762744
