Electro-Mechano-Optical Stimulation Sensitivity Enabling Hf_0.5Zr_0.5O₂/ZnO Nanowires Heterosynapse for Adaptive Conditioning Decision-Making and Motion Direction Detection

TY - JOUR

T1 - Electro-Mechano-Optical Stimulation Sensitivity Enabling Hf0.5Zr0.5O2/ZnO Nanowires Heterosynapse for Adaptive Conditioning Decision-Making and Motion Direction Detection

AU - Han, Qianqian

AU - Zhang, Mingzhen

AU - Zhang, Yaju

AU - Li, Ying

AU - Wang, Jiantao

AU - Liu, Bei

AU - Zhang, Yuanzheng

AU - Ge, Chen

AU - Shen, Guozhen

AU - Zheng, Haiwu

N1 - Publisher Copyright: © 2026 Wiley-VCH GmbH.

PY - 2026/7/22

Y1 - 2026/7/22

N2 - Bio-inspired neuromorphic intelligence hardware is noted for integrated-mode perception and adaptive learning-memory-processing capabilities, becoming a pursued spotlight for high-level interaction with the surroundings. However, sensing and encoding multi-stimuli inputs into dynamic bio-realistic behavior within a single synapse device—oriented for high-order adaptive activities—remain a stiff challenge. Herein, artificial heterosynapse constructed from Hf0.5Zr0.5O2/ZnO nanowires heterojunction allows for multi-mode electro-mechano-optical stimulation to be encoded into adaptive synaptic plasticity benefiting from the coupling of ferro-piezo-phototronic effect. The intrinsic ferroelectric and piezoelectric polarization serve as additional modulators for enlarging weight update change ratio and extending memory level. These comprehensive performance metrics of the heterosynapse enable multi-scenario applications such as visual imaging and memory, cognitive improvement. Notably, a more intelligent “rabbit” thoroughly emulates associative learning features namely acquisition, extinction, generalization, and differentiation, supporting adaptive conditioning decision-making. The heterosynapse-based multi-mode reservoir computing system with 8 distinguishable reservoir states and optimized weight update parameters ensures high-precision recognition accuracy when performing motion direction detection even under high-level noise emphasizing singular robustness against interference. This work provides a step forward in integrating cross-mode perception and adaptive neuromorphic behavior into biomimetic hardware, underpinning artificial heterosynapse well-suited to effectively tackle multi-tasks.

AB - Bio-inspired neuromorphic intelligence hardware is noted for integrated-mode perception and adaptive learning-memory-processing capabilities, becoming a pursued spotlight for high-level interaction with the surroundings. However, sensing and encoding multi-stimuli inputs into dynamic bio-realistic behavior within a single synapse device—oriented for high-order adaptive activities—remain a stiff challenge. Herein, artificial heterosynapse constructed from Hf0.5Zr0.5O2/ZnO nanowires heterojunction allows for multi-mode electro-mechano-optical stimulation to be encoded into adaptive synaptic plasticity benefiting from the coupling of ferro-piezo-phototronic effect. The intrinsic ferroelectric and piezoelectric polarization serve as additional modulators for enlarging weight update change ratio and extending memory level. These comprehensive performance metrics of the heterosynapse enable multi-scenario applications such as visual imaging and memory, cognitive improvement. Notably, a more intelligent “rabbit” thoroughly emulates associative learning features namely acquisition, extinction, generalization, and differentiation, supporting adaptive conditioning decision-making. The heterosynapse-based multi-mode reservoir computing system with 8 distinguishable reservoir states and optimized weight update parameters ensures high-precision recognition accuracy when performing motion direction detection even under high-level noise emphasizing singular robustness against interference. This work provides a step forward in integrating cross-mode perception and adaptive neuromorphic behavior into biomimetic hardware, underpinning artificial heterosynapse well-suited to effectively tackle multi-tasks.

KW - HfZrO/ZnO nanowires

KW - adaptive synaptic plasticity

KW - associative learning

KW - ferro-piezo-phototronic effect

KW - reservoir computing system

UR - https://www.scopus.com/pages/publications/105032903025

U2 - 10.1002/lpor.202502977

DO - 10.1002/lpor.202502977

M3 - Article

AN - SCOPUS:105032903025

SN - 1863-8880

VL - 20

JO - Laser and Photonics Reviews

JF - Laser and Photonics Reviews

IS - 14

M1 - e02977

ER -