High Perpendicular Anisotropy in Mo‐Inserted Mg Composite Free Layer for Nonvolatile Magnetoresistive Random Access Memory in 4K‐400K Universal Temperature Applications

Read full text
Authors: Ming‐Chun Hong, Chen‐Yu Yang, Hsin‐Han Lee, Guan‐Long Chen, Ching Shih, Yu-Chen Hsin, Chiao‐Yun Lo, Sin-You Huang, Cheng‐Yi Shih, Shan-Yi Yang, I‐Jung Wang, Yao‐Jen Chang
+10 more Shih-Ching Chiu, Yi-Hui Su, C.-Y. Wang, Kuan‐Ming Chen, H. L. Tsai, Jeng‐Hua Wei, Shyh-Shyuan Sheu, Wei‐Chung Lo, Shih‐Chieh Chang, Tuo‐Hung Hou
Cited by: 0
References: 44

Abstract

ABSTRACT In recent years, demand for nonvolatile memory (NVM) has surged across diverse applications. Spin‐transfer‐torque magnetoresistive random access memory (STT‐MRAM) is a leading scalable solution for embedded NVM, mass‐producible below 28 nm, offering superior performance over other NVM technologies. However, achieving low switching voltage (V SW ) and high thermal stability (Δ) across temperatures from 4 to 400K remains challenging due to MRAM's temperature sensitivity. This study introduces a universal temperature‐friendly nonvolatile MRAM (UTF‐NVMRAM) based on an STT‐MRAM framework, designed for applications spanning 4 to 400K. By optimizing the MgO/MgO x capping layer and incorporating Mo into a CoFeB composite‐free layer (CFL) with an Mg spacer, this design minimizes temperature sensitivity in V SW and Δ while ensuring potential compatibility with back‐end‐of‐line annealing temperatures. It maintains a large write margin without compromising the magnetoresistance ratio or switching current. The device achieves low V SW sensitivity (V SW /T ≈ 0.68 mV/K), low Δ sensitivity (Δ/T ≈ 0.1/K) with retention exceeding 10 years, a write error rate below 1 ppm, an operating speed of 10 ns, and endurance surpassing 10 11 cycles across the full temperature range. These attributes position UTF‐NVMRAM as the most promising NVM solution based on commercialized STT‐MRAM technology for wide‐temperature applications.

Fields of Study & Subjects

Magnetic properties of thin films Advanced Memory and Neural Computing Magnetic Properties of Alloys

Publication details

Journal / source
Small
Volume
22
Issue
14
Pages
e09788–e09788
Publication date
17 January 2026
Language
EN