麻豆淫院

A research team from the Ningbo Institute of Materials Technology and Engineering (NIMTE) of the Chinese Academy of Sciences (CAS) has proposed a hybrid transfer and epitaxy strategy, enabling the heterogeneous integration of single-crystal oxide spin Hall materials on silicon substrates for high-performance oxide-based spintronic devices.

The study is published in .

Spintronic devices are gaining attention as a key direction for next-generation information technologies due to their low power consumption, non-volatility, and ultra-fast operating capabilities.

Single-crystal oxide spin Hall materials are known for their exceptional charge-spin conversion efficiency, making them promising candidates for low-power spintronic devices, particularly spin-orbit torque (SOT) devices. However, integrating these materials with silicon substrates poses significant challenges.

To address these challenges, the researchers developed an innovative method that combines transfer technology with epitaxial deposition, successfully integrating oxide spin Hall materials onto silicon substrates. Using this approach, they were able to create single-crystal SrRuO₃ (SRO) films on silicon substrates and prepare corresponding SOT devices.

The SRO film demonstrated a high spin Hall conductivity of 6.1脳10⁴ 魔/2e S路m^-1, enabling magnetization switching with a low critical current density of 1.3脳10¹⁰ 础路尘^-2 in the SOT devices.

Furthermore, multi-state magnetization switching characteristics were observed, allowing these devices to simulate biological synaptic and neuronal functions. An artificial neural network utilizing these devices achieved an impressive accuracy rate of 88% in image recognition tasks.

This study creates a new integration method for silicon-based oxide electronics, showcasing the broad applicability of the hybrid transfer and epitaxy strategy across various oxide material systems.

It advances the development of high-performance silicon-based spintronic devices and holds great potential for low-power electronics and neuromorphic computing applications.