Ferroelectric Schottky Barrier MOSFET as Analog Synapses for Neuromorphic Computing

Abstract

In this paper, artificial synapses based on ferroelectric Schottky barrier MOSFETs (FE-SBFETs) are presented. The FE-SBFETs are fabricated with Si doped Hf02 ferroelectric layers scaling down to a gate length of 40 nm and using single crystalline NiSi <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> contacts on siliconon-insulator (SOI) substrates. The ferroelectric polarization switching dynamics gradually modulate the Schottky barriers, thus programming the device conductance by applying stimulus on the gate to imitate the short- and long-term plasticity of biological synapse, including excitatory/inhibitory postsynaptic current (EPSC/IPSC), paired-pulse facilitation (PPF) and long-term potentiation (LTP) and long-term depression (LTD) behaviors. Based on a multilayer perceptron artificial neural networks, a high recognition accuracy (83.6%) is achieved for handwritten digits. These findings demonstrate FE-SBFET has high potential as an ideal synaptic component for the future intelligent neuromorphic network.