Conductance quantization control and neuromorphic properties in Pt-nanoparticle incorporated HfAlOx alloy memristor

Abstract

In this work, platinum nanoparticles have been utilized to achieve better control of conductance quantization for high-density memory storage nonvolatile memory applications. Here, atomic layer deposited Pt-nanoparticles are sandwiched between HfAlOx switching layers. An Au/Ti/HfAlOx/Pt-NP/HfAlOx/ITO resistive random-access memory (RRAM) device exhibits bipolar resistive switching SET/RESET properties at a very low external electric field with memory window >10 and an endurance of >103 cycles. With a very slow voltage sweep rate (0.002 V/step) during current–voltage characteristics under both SET and RESET conditions, a controlled stepwise increase/decrease in distinct conductance quantization behavior with integer and half-integer multiples was observed. This phenomenon predicts atomic contact formation and rupture of oxygen vacancies between conductive filaments and Pt-NPs. Control of post-synaptic conductance properties with modulation of pre-spike width, number, and frequency showed the robustness of the RRAM device studied here. Gradual, controlled change in conductance obtained under dc and pulse conditions in the experiments is very promising for next-generation multi-level high-density storage RRAM devices to develop artificial electric synapses for neuromorphic applications.