Nanoporous Cation Limiter-Induced Enhancement of Threshold Switching and Oscillatory Behavior in Ag-Based Diffusive Memristors

Abstract

The development of memristors for neuromorphic computing has gained attention due to their ability to mimic biological neurons. Among the various switching mechanisms, volatile electrochemical metallization (ECM)-threshold switching (TS) memristors show promise for artificial neural networks due to their simple structure and low operating voltages. However, ECM TS memristors often suffer from poor switching uniformity, limiting practical applications. In this work, we demonstrate a highly uniform switching Ag-based TS memristor with a nanoporous-Pt (np-Pt) cation limiter. The device achieves ultralow leakage current (<1 pA), high selectivity (>10⁷), and high endurance (>10⁶ cycles). The np-Pt cation limiter also enhances the device’s stability by reducing variability in the operating voltages (V_th and V_hold) and enabling operations at higher current compliance levels (∼10 μA). In addition, the Ag/np-Pt TS device exhibits self-oscillation behavior at low voltage (<1 V), with oscillation frequency increasing with the applied voltage. The insertion of the np-Pt cation limiter provides a simplistic technique of metal ions manipulation in ECM TS devices, enhancing their performance for artificial neural network applications.