An Organic–Inorganic Hybrid Memristor with Trap-Assisted Hopping for Synaptic Plasticity and Artificial Nociceptor

Abstract

Neuromorphic hardware demands artificial synaptic devices that mimic biological synapses. Achieving multifunctional emulation of synaptic plasticity and nociceptive behavior in a single, highly stable, low-power device remains a major challenge. In this report, zinc oxide nanoparticles (ZnONP) are synthesized via a surfactant-assisted sol–gel method. These are incorporated into an organic small-molecule matrix of 6,13-bis(triisopropylsilylethynyl)pentacene (TP) to fabricate a solution-processed organic–inorganic hybrid memristor. The as-fabricated hybrid memristor mimicked artificial synaptic behavior at a low operating voltage (<±1 V) while demonstrating high stability with retention exceeding 105 s in ambient atmosphere. The TP/ZnONP-based memristor has illustrated both short-term and long-term synaptic plasticity, facilitating the exemplification of learning and memory functions of biological synapses. The memristor has been found to exhibit pain-perception-related nociceptive behavior, including protective responses to excessive stimulation and self-recovery, closely resembling that of biological nociceptors. The device exhibits neurapraxia-like responses with a recovery time of only 40 s. The experimentally obtained synaptic characteristics are further employed in a device-aware fully connected artificial neural network, yielding a handwritten digit recognition accuracy of 97.9%. Various investigations on the dynamics of charge transport have revealed trap-assisted hopping through transient localization, which is induced by shallow and deep traps. The integration of artificial nociception and pattern recognition in a highly stable low-power synaptic device demonstrates the potential of the hybrid memristor for multifunctional neuromorphic hardware.