.png){kind=icon}
.png){kind=icon}
Return
Spin-Charge Conversion-Based Artificial Synaptic Device for Neuromorphic Computing
.png){kind=icon}
PRE
AI
DOI：10.1021/acsaelm.4c02048.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
A synaptic function is demonstrated using spin-charge conversion in a Rashba system. In an asymmetric quantum well channel, fast-moving charges induce a Rashba effective magnetic field, which separates spin-up and spin-down potentials. The ferromagnet detects these spin-dependent potentials, corresponding to the spin information on the channel. The multiple ferromagnetic electrodes, each with different switching fields, probe their respective spin potentials, and the output terminal reads the superposition of the detected potentials, thereby realizing multiple voltage states. These multiple states are systematically modulated and changed to any desired state directly, enabling both the potentiation and depression of synaptic behavior. In this memristive function device, both charge-to-spin and spin-to-charge conversions are demonstrated in a single device, consistent with the reciprocal relation. Additionally, neuromorphic pattern recognition is clearly demonstrated by controlling the V max/V min ratio and offset resistance.
Keywords:
spin-chargeconversion
Rashba spin-orbitcoupling
multistate memory
artificial synapses
neuromorphic computing
Journal
IF:
4.7
Papers:
5.0K
Citations:
1.4W