Controlled multilevel switching and artificial synapse characteristics in transparent HfAlO-alloy based memristor with embedded TaN nanoparticles

Abstract

Atomic layer deposition technique has been used to prepare tantalum nitride nanoparticles (TaN-NPs) and sandwiched between Al-doped HfO2 layers to achieve ITO/HfAlO/TaN-NP/HfAlO/ITO RRAM device. Transmission electron microscopy along with energy dispersive spectroscopy confirms the presence of TaN-NPs. X-ray photoelectron spectroscopy suggests that part of TaN converted to tantalum oxynitride (TaOxNy) which plays an important role in stable cycle-to-cycle resistive switching. Charge trapping and oxygen vacancy creation were found to be modified after the inclusion of TaN-NPs inside RRAM structure. Also, HfAlO/TaOxNy interface due to the presence TaN-NPs improves the device-to-device switching reliability by reducing the probability of random rupture/formation of conductive filaments (CFs). DC endurance of more than 10(3) cycles and memory data retention up to 10(4) s was achieved with an insignificant variation of different resistance states. Multilevel conductance was attained by controlling RESET voltage with stable data retention in multiple states. The volatile threshold switching was monitored after controlling the CF forming at 200 nA current compliance with high selectivity of similar to 10(3). Synaptic learning behavior has been demonstrated by spike-rate-dependent plasticity (SRDP). Reliable potentiation and depression processes were observed after the application of suitable negative and positive pulses which shows the capability of the TaN-NPs based RRAM device for transparent synaptic devices. (C) 2021 Published by Elsevier Ltd on behalf of Chinese Society for Metals.