SinoTechIntel Academic Portal
Official PDF TranslationJournal of Inorganic Materials (无机材料学报)

Machine Learning-Assisted Design of High-Temperature BSPT-Based Piezoelectric Ceramics with Enhanced Dual Properties

Authors: ZUO Zhiping; GUO Chun; ZHOU Zhiyong

DOI: 10.15541/jim20260017Status: Verified Translated Edition
reCAPTCHA Bot Shield Active

Preparing Secure Academic Download

Verifying human reader & generating high-resolution document...

Verifying Document Integrity15s remaining
← Back to Article
Protected by Google reCAPTCHA v3.PrivacyTerms

Key Findings in This Report

• • BSPTGW10 (x=0.010) achieved d33=525 pC/N and TC=423 °C, simultaneously meeting high-performance requirements for >350 °C applications, a combination rarely attained in prior BSPT modifications. • • Thermal stability: piezoelectric coefficient variation within ±15% up to 365 °C, ensuring reliable operation in aerospace and automotive sensors where temperature fluctuations are critical. • • Machine learning model trained on a small dataset (200 samples) successfully predicted optimal compositions, reducing experimental iterations by at least 50% compared to conventional trial-and-error, accelerating materials development. • • Ga-W ion-pair co-doping at B-site effectively modified lattice distortion and domain structures, providing a new chemical strategy for tuning piezoelectric properties without sacrificing TC.