Toward neuromorphic pain sensors: Full-Spectrum artificial nociception via analog-switching memristors with tunable volatility

Neuromorphic computing systems could greatly benefit from electronic devices that exhibit analog resistive switching and dynamic adaptation similar to those of biological neurons, particularly for implementing sensory functions such as nociception. Here, we present a graphene oxide–cobalt oxide (GO–CoO) memristor that exhibits analog resistive switching with intrinsic current decay. Successive current–voltage sweeps with varying cutoff voltages demonstrate multilevel, finely tunable resistance states over a broad range without any compliance current. Charge–flux analysis confirms that the device operates as a true memristor, and cumulative conductance buildup under sequential positive and negative biases indicates robust, polarity-independent switching. Moreover, tuning the voltage sweep rate modulates the inertia of charge carriers that governs the switching kinetics. The GO–CoO architecture establishes a conductive network wherein oxygen vacancy dynamics within CoO particles and conductive rGO formation collectively govern analog resistive switching, providing the essential tunability and stability required to emulate synaptic plasticity. Under optimized pulsed stimulation, the memristor faithfully emulates key nociceptive neural responses, including a threshold response, peripheral sensitization (manifested as hyperalgesia and allodynia), central sensitization (temporal summation and facilitation), suprathreshold response, and post-stimulus recovery. The device also exhibits both short-term and long-term synaptic plasticity. These results pave the way for simple, cost-effective memristors capable of emulating neural synaptic functions and pain perception.