Pulsed Low-Frequency Noise Characterization on Al/ZnO/Al ReRAM to Investigate the Role of Deep-Level Oxygen Vacancy State in HRS Leakage

A low-frequency noise (LFN)-based experimental method has been developed within the scope of the present work, which was utilized for extracting space-energy distributions of trap states within the reactively sputtered ZnO film in an Al/ZnO/Al resistive random access memory (ReRAM) cell. The main strategy of the experimental scheme was to tailor applied voltage pulses during LFN characterization, so that they could selectively impact electron capture and emission from the trap states, strictly speaking donor like oxygen vacancy states, which reside over a continuum of space and energy within the ZnO film. Experimental noise current power spectral density (PSD) was interpreted by first principle analytic noise current PSD model, which was subsequently employed for profiling the oxygen vacancy states. We found that ~2-eV deep oxygen vacancy state within the ZnO bandgap via coupling with shallow oxygen vacancy state (0.85 eV) contributed high-resistance state (HRS) leakage, which had been explicated by phonon-assisted tunneling transport (PATT) of electrons, and the spatial density of traps clustered around the same energy state as profiled by the proposed pulsed LFN characterization was employed during PATT model fitting further to check the consistency of the same leakage current with measurement temperatures.