Tunable capacitive resistance switching with low-power synaptic bionic potential in (1–x)Bi0.88Nd0.12FeO3–xCaBi4Ti4O15 thin films

Abstract

A memristor with a low power consumption, non-volatility, and adaptive abilities complex has a promising prospect in neural network computing systems due to its unique nonlinearity, memory, and local activity. Here, the Au/(1-x)Bi_0.88Nd_0.12FeO₃-xCaBi₄Ti₄O₁₅ (BNFO-CBTO, x = 0.1, 0.2, 0.3, 0.4, 0.5) non-volatile memory devices with resistance-switching (RS) behaviors are fabricated by sol–gel method. The (1-x) BNFO-xCBTO samples exhibit a tunable capacitive resistive switching behavior by the CBTO phase, i.e., the higher the content of the CBTO phase, the more obvious the phenomenon of capacitive resistance-switching behavior. Moreover, the CBTO phase improves the cyclic fatigue characteristics of the (1–x)BNFO–xCBTO samples. The lowest operating current (~ 1nA-100nA) is observed in the 0.6BNFO-0.4CBTO sample. Further, the multiple resistance states, conductive mechanisms, and synaptic behaviors with conductance continuous modulation, paired-pulse facilitation (PPF) behaviors, and excitatory postsynaptic current (EPSC) are also simulated. The 0.6BNFO-0.4CBTO non-volatile memory device with tunable abnormal resistance switching, low-power synaptic bionic potential, and a series of synaptic-like behaviors can provide a new opportunity to apply the RS behavior in high-performance computing with low power consuming, brain-like neuromorphic mimicry, and next-generation information-storage devices.