Related Experiment Video
Updated: Jun 12, 2025

09:57
Imaging Membrane Potential with Two Types of Genetically Encoded Fluorescent Voltage Sensors
Published on: February 4, 2016
10.7K
Photoresponsive Two-Dimensional Magnetic Junctions for Reconfigurable In-Memory Sensing
Wenxuan Zhu1, Jiacheng Sun2, Yuan Cheng2,3
1Key Laboratory of Advanced Materials (MOE), School of Materials Science and Engineering, Tsinghua University, Beijing 100084,China.
ACS Nano
|September 17, 2024
Summary
Researchers developed a novel all-two-dimensional magnetic junction for optical sensing. This breakthrough enables magnetic memory devices to perform in-memory computing for artificial intelligence, enhancing machine vision systems.
Area of Science:
- Materials Science
- Condensed Matter Physics
- Nanotechnology
Background:
- Magnetic tunneling junctions (MTJs) are key to magnetic random access memory (MRAM), offering potential for integrated memory and computing.
- Traditional MTJs lack light sensitivity, hindering their use in optical sensing for AI applications like machine vision.
- All-two-dimensional (2D) materials offer unique electronic and magnetic properties for novel device applications.
Purpose of the Study:
- To integrate optical sensing capabilities into magnetic memory devices.
- To develop an all-2D magnetic junction with tunable magnetoresistance controlled by light.
- To construct a machine vision system utilizing in-memory sensing and computing.
Main Methods:
- Fabrication of an all-2D magnetic junction using Fe3GaTe2/WSe2/Fe3GaTe2 heterostructures.
- Characterization of the junction's magnetic and optoelectronic properties at room temperature.
- Demonstration of optical control over junction states and magnetoresistance.
- Implementation of a machine vision system based on the developed device.
Main Results:
- The Fe3GaTe2/WSe2/Fe3GaTe2 junction exhibits intrinsic band gap and prominent photoresponse from WSe2.
- The junction's on-off states and magnetoresistance are optically tunable by light intensity.
- A high-performance machine vision system for image recognition was successfully constructed.
- The device demonstrates the advantages of 2D magneto-electronic devices.
Conclusions:
- The developed all-2D magnetic junction successfully merges magnetic memory with optical sensing.
- Optical tunability of magnetoresistance enables efficient in-memory sensing and computing for AI.
- This work expands the application scope of magnetic memory devices in artificial intelligence and machine vision.

