# Polarity‐Controlled Volatile HfO2 Memristors with Bimodal Conductance for Neuromorphic Synapses and Reservoir Computing

## TL;DR

This paper introduces a new type of memristor that can switch between multiple conductance states based on voltage polarity, enabling efficient neuromorphic computing and achieving high accuracy in data classification.

## Contribution

The novelty lies in the memristor's bimodal conductance behavior, which doubles the number of distinguishable states per cell by reversing the read voltage polarity.

## Key findings

- The memristor supports 16 conductance states with 4-bit pulsed inputs and 16 more with reversed polarity.
- A reservoir computing system using this device achieved 98.81% accuracy on the MNIST dataset.
- The device's bimodal behavior enhances encoding diversity and real-time signal processing efficiency.

## Abstract

In this work, an HfO2‐based memristor exhibiting bimodal switching, wherein the device's conductance is modulated not only by the input stimulus but also by the polarity of the read voltage, is introduced. Uniquely, this device demonstrates reliable short‐term memory (STM)‐like behavior and supports 16 well‐separated conductance states through 4‐bit pulsed inputs. Remarkably, under the same input conditions, reversing the polarity of the read voltage results in 16 more different conductance states, thereby doubling the number of levels that can be distinguished per cell. Employing the proposed device, a reservoir computing (RC) system, which takes advantage of this rich representational capability, is created. The system achieves a high classification accuracy of 98.81% on the MNIST dataset. These results show how powerful memristor‐based architectures can be and how this device could be a compact and energy‐efficient platform for the next generation of neuromorphic computing.

This study demonstrates a TiN/HfO2/ITO memristor exhibiting field‐induced bimodal volatile switching behavior originating from ion bombardment during sputtering. The dipole‐driven, polarity‐dependent states produce multilevel output responses that expand the reservoir's output dimensionality and enhance encoding diversity. Such a mechanism enables efficient real‐time signal processing, offering a compact and energy‐efficient hardware platform for advanced neuromorphic and reservoir computing applications.