Ferroelectric synaptic devices based on CMOS-compatible HfAlO for neuromorphic and reservoir computing applications

Area of Science:

• Materials Science
• Solid State Physics
• Nanotechnology

Background:

• Ferroelectric tunnel junctions (FTJs) are researched for low-power, CMOS-compatible memristor applications.
• Hafnium oxide-based materials offer advantages for ferroelectricity induction.

Purpose of the Study:

• To propose and experimentally demonstrate a hafnium aluminum oxide (HAO) based FTJ as a synaptic device.
• To optimize HAO FTJ performance by analyzing thickness and cell area effects on ferroelectric properties.
• To evaluate the potential of HAO FTJs for neuromorphic computing applications.

Main Methods:

• Fabrication and electrical characterization of HAO-based FTJs.
• Experimental evaluation of ferroelectric properties, including remanent polarization (2Pr) and tunneling electroresistance (TER).
• Demonstration of synaptic behaviors like paired-pulse facilitation/depression and spike-timing-dependent plasticity.

Main Results:

• A HAO-based FTJ was successfully fabricated and demonstrated as a synaptic device.
• Optimized devices achieved a large remanent polarization (2Pr) exceeding 43 μC cm⁻².
• Synaptic plasticity phenomena and reservoir computing capabilities were experimentally verified.

Conclusions:

• Hafnium aluminum oxide is a viable material for high-performance ferroelectric tunnel junctions.
• HAO-based FTJs show significant potential for artificial synaptic devices and neuromorphic computing.
• The study verifies the synaptic properties of FTJs for artificial synapse implementation.