Two-dimensional (2D) MXene nanomaterials have shown great promise for electronic devices, attributed to their metal-resembling conductivity and abundant surface functional groups. However, the utilization of intrinsic property of MXene in memristors remains challenging due to its free electron conducting behavior rather than semiconducting property. Here, a N-fused perylenediimide organic semiconductor (CBIN) with conjugated skeleton and heteroatoms (O, S, N) is designed to successfully actuate the surface modification of MXene. The organic CBIN-decorated MXene demonstrates remarkable bipolar memristive properties, such as low threshold voltages of approximate ±1.4 V, exalted retention time exceeding 10⁴ s, and outstanding environmental stability even after exposure to ultraviolet and x-ray irradiations. Furthermore, the CBIN-MXene hybrid memristive device can mimic synaptic plasticity and holds potential for information encoding as quick response codes and image recognition processing. This study provides efficient guidelines for implementing MXene-based memristors by organic semiconductor modulation and opens up possibilities of extending their functionalities into information encryption and neuromorphic computing applications.

Graphene oxide (GO)-based memristors offer the promise of low cost, eco-friendliness, and mechanical flexibility, making them attractive candidates for outstanding flexible electronic devices. However, their resistive transitions often display abrupt change rather than bidirectional progressive tuning, which largely limits their applications for biological synapse emulation and neuromorphic computing. Here, a memristor with a novel layered structure of GO/pyridinium/GO is presented with tunable bidirectional feature. The inserted organic pyridinium intercalation succeeds in serving as a satisfactory buffer layer to intrinsically control the formation of conductive filaments during device operation, leading to progressive conductance regulation. Thus, the essential synaptic behaviors including analog memory characteristics, excitatory postsynaptic current, paired pulse facilitation, prepulse inhibition, spike-timing-dependent plasticity, and spike-rate-dependent plasticity are replicated. The emulation of brain-like “learning-forgetting-relearning” process is also implemented. Additionally, the instant responses of the memristor can be stimulated by low operational voltages and short pulse widths. This study paves one way for GO-based memristors to actuate appealing features such as bidirectional tuning and fast speed switching that are desirable for the development of bio-inspired neuromorphic systems.