Adaptive Shannon entropy in ferroelectric thin-film transistors for functional neuromorphic applications
The increasing demand for adaptive and energy-efficient systems highlights the limitations of conventional devices, which are often constrained by rigid and deterministic behaviors. In contrast, nature, with its inherent randomness and variability, presents an opportunity to harness this intrinsic d...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-08-01
|
| Series: | Materials Today Advances |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2590049825000438 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The increasing demand for adaptive and energy-efficient systems highlights the limitations of conventional devices, which are often constrained by rigid and deterministic behaviors. In contrast, nature, with its inherent randomness and variability, presents an opportunity to harness this intrinsic disorder to advance functional devices beyond traditional paradigms. However, the effective utilization of intrinsic randomness in physical devices for adaptive, multilevel data processing and beyond has yet to be fully demonstrated. By leveraging Shannon entropy as a quantifiable measure of randomness, we demonstrate how ferroelectric In2O3/HfO2 thin-film transistors (TFTs) can exhibit adaptive, multilevel dynamic behavior, essential for applications in secure systems and neuromorphic computing. The observed multilevel dynamics are attributed to the reorientation of ferroelectric polarization and charge trapping effects, as confirmed by local probe force microscopy. This entropy-driven mechanism enables dynamic functionalities, including secure noise-based authentication, random data encoding/obfuscation, and nociceptor-like responses such as hyperalgesia and allodynia, offering low-energy operation (∼2 nJ read) operation. Our findings establish Shannon entropy as a foundational metric, linking physical randomness with ferroelectric properties to create robust, energy-efficient, and adaptable devices for next-generation secure and neuromorphic architectures. |
|---|---|
| ISSN: | 2590-0498 |