Ion Intercalation Enabled Tunable Frequency Response in Lithium Niobite Memristors

have emerged as a viable component for developing neuromorphic hardware platforms, which can compete with biological systems in density, accuracy, and energy efficiency. Among contemporary memristive systems, intercalation-driven lithium niobite (LiNbO2) memristors have the advantage of inherent flux-driven large analog conductance tunability (delta R/R > 89), a wide resistance range (similar to 10-1E7 omega) selected via geometry selection, and low-power (similar to 100-150 mV) neuromorphic functionality resembling synaptic weight updates in the synaptic mem-brane. Emerging neural systems require nonstatic temporal responses to implement temporally diverse architectures, such as recurrent neural networks (RNNs), yet these tem-porally diverse memristors are rare. There is a gap in understanding of the frequency response of nonvolatile memristors, which is a fundamental characteristic in mem-ristive systems and ultimately dictates the speed of adaptive learning in deployed neural networks and the memory win-dows available for designers in RNNs. Using both large and small signal characterization methods, these memristors demonstrate up to a decade of span in tunable frequency -dependent response, statically controlled via device geom-etry and scaling design rules, and dynamically tuned via channel lithium concentration. Thus, the tunable frequency dependence of resistance modulation in LiNbO2 memristors is evaluated, enabling future neural network design rules to be identified for scalable multifunctional memristive sys-tems for neuromorphic computing applications.

Automated summary

Memristors based on intercalation-driven lithium niobite (LiNbO2) have emerged as a promising option for developing neuromorphic hardware platforms. These memristors offer advantages such as large analog conductance tunability, wide resistance range, and low-power neuromorphic functionality resembling synaptic weight updates. However, there is limited understanding of their frequency response, which is a crucial characteristic for adaptive learning in neural networks. This study evaluates the tunable frequency dependence of resistance modulation in LiNbO2 memristors, enabling the identification of design rules for future scalable multifunctional memristive systems in neuromorphic computing applications.