.png){kind=icon}
.png){kind=icon}
Return
Programmable Logic and Reservoir Computing Based on Hydrogenation-Engineered Spin-Orbit Torque
.png){kind=icon}
PRE
AI
DOI：10.1002/adfm.202519753.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To achieve efficient manipulation of magnetism for next-generation magnetic memory and spintronic devices, integration of distinct electrical control manners within a single device represents a critical pursuit in spintronics. Here, an effective approach combining voltage-controlled hydrogenation and current-induced spin-orbit torque (SOT) is proposed to manipulate the ferromagnetism in an all-solid-state magneto-ionic device with GdOx/Pd/Pd65Co35 heterostructure. Hydrogenation drives a reversible ferromagnetic-to-paramagnetic transition, while the SOT enables switching of magnetization without the external magnetic field. Furthermore, SOT-driven magnetization switching can be finely tuned by the hydrogenation, leading to regulation of switching amplitude, reversal of switching polarity, and reduction of critical switching current. Through magnetization sign and magnitude engineering via coordinated voltage/current inputs, multiple programmable spin logic gates are demonstrated within a single device. Additionally, the semi-nonvolatile feature of hydrogenation at room temperature results in a nonlinear evolution of magnetization with time and is utilized to construct a simulated physical reservoir computing system, which achieves exceptional performance in temporal tasks, including high accuracy (98.4%) in speech recognition and low error (root mean square error <0.1) in chaotic prediction. This work offers a promising route toward the integration of memory, logic, and neuromorphic computing within a unified spintronic device.
Keywords:
hydrogenation
logic
magnetoelectric coupling
neuromorphic computing
spin-orbit torque
Journal
IF:
19
Papers:
3.4W
Citations:
32.1W