Hafnium-Based Ferroelectric Post-Moore Electronics: Device Physics, Integration Architectures, and Neuromorphic...

Area of Science:

• Materials Science
• Electrical Engineering
• Computer Science

Background:

• Perovskite ferroelectrics face limitations, driving research into alternatives.
• Hafnium-based ferroelectrics (Hf-FEs) offer CMOS compatibility and scalability.
• Hf-FEs are crucial for advancing non-volatile memory and neuromorphic computing.

Purpose of the Study:

• To review material systems, device structures, and mechanisms of Hf-FEs.
• To discuss evaluation parameters for efficient neuromorphic computing with Hf-FEs.
• To explore the progress and future prospects of Hf-FEs in advanced applications.

Main Methods:

• Systematic review of Hf-FE material systems and device physics.
• Analysis of performance metrics for neuromorphic applications, including synaptic plasticity.
• Review of array integration and hardware advancements.

Main Results:

• Hf-FEs overcome perovskite limitations, enabling applications in image processing and in-memory logic.
• Key parameters for neuromorphic efficiency, storage, and plasticity are identified.
• Progress in Hf-FE arrays and hardware integration is systematically presented.

Conclusions:

• Hf-FEs show significant potential for high-performance neuromorphic computing.
• Further research is needed to address challenges in advanced applications.
• This work provides guidance for developing next-generation Hf-FE based devices.