상세 보기
Defect-coupled Co3O4@In2O3 heterostructures on plasma-activated polyimide with interpretable machine learning for breath acetone sensing
- Hilal, Muhammad;
- Ali, Yasir;
- Ullah, Zahid;
- Cai, Zhicheng;
- Kim, Hyojung;
- ... Hwang, Yongha;
- 외 3명
WEB OF SCIENCE
0SCOPUS
0초록
Defect-regulated interfacial electrostatics critically determine the stability and reliability of ceramic gas sensors under realistic operating environments. Here, a flexible three-dimensional Co3O4@In2O3 heterostructure integrated on O2-plasma-activated (PI) is systematically engineered to couple oxygen-vacancy modulation with chemically stabilized oxide-polymer interfaces. Dual-site Co2+<-> In3+ substitution and p-n junction formation regulate carrier density and vacancy population, enabling controlled depletion-layer modulation and stable surface oxygen activation. Plasma activation introduces polar functional groups (-OH, C=O, -COOH) that form chemically bonded interfaces through hydrogen-bonding and dipolar interactions (-COO-center dot center dot center dot M, M-OH center dot center dot center dot O=C), reducing interfacial energy and enhancing mechanical and environmental durability. The hydrophilic interfacial layer suppresses humidity-induced performance degradation, enabling stable sensing at 95% relative humidity. Benefiting from the low thermal conductivity of PI, the device achieves stable operation at 120 degrees C, significantly lower than Si-supported counterparts (>= 150 degrees C), indicating improved thermal utilization efficiency. The optimized Co3O4@In2O3/PI sensor demonstrates reliable acetone detection down to 50 ppb with rapid and reversible response characteristics and stable operation over extended mechanical deformation cycles. Time-domain resistance features coupled with machine-learning classification achieve a Macro-F1 score of 65.0% for distinguishing low (<= 25 ppm) and high (>= 30 ppm) exposure regimes. These results establish defect-interface co-engineering as a viable route toward mechanically robust, humidity-tolerant, and intelligent ceramic sensing platforms. Code is available at: https://github.com/Zahid672/CI-3-PI
키워드
- 제목
- Defect-coupled Co3O4@In2O3 heterostructures on plasma-activated polyimide with interpretable machine learning for breath acetone sensing
- 저자
- Hilal, Muhammad; Ali, Yasir; Ullah, Zahid; Cai, Zhicheng; Kim, Hyojung; Alnaser, Ibrahim A.; Abdo, Hany S.; Lee, Seonghyeon; Hwang, Yongha
- 발행일
- 2026-06
- 유형
- Article
- 권
- 236