Robust Metal and Semiconductor Phase Transition Memristor Using Ag‐Intercalated Transition Metal Dichalcogenide

W Whan Kyun Kim (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) G Ga Young Cho (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) T Thi Thanh Huong Vu (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) J Jun Sun Son (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) Y Yong Ha Shin (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) H Hong Woon Yun (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) M Min Seok Kim Y Yong Seon Shin (Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea) H Heejun Yang W Woo Jong Yu

Abstract

Abstract Metal (1T/1T')‐semiconductor (2H) phase transition memristors (PTMEMs) based on intercalated alkali metal ions (Li + ) in transition metal dichalcogenides (TMDs) exhibit excellent electrical properties, including heterosynaptic plasticity. However, the low stability of Li + ions limits retention and on/off ratios of the PTMEMs. Here, a phase transition in MoTe 2 induced by intercalated Ag + ions is demonstrated for the first time, enabling robust memristor operation. The migration of Ag + ions, controlled by voltage biases, clearly realizes reversible 2H‐1T/1T’ phase transitions, as revealed by transmission electron microscopy, X‐ray photoelectron spectroscopy, and Raman mapping. The memristive mechanism of MoTe 2 shifts from doping (4–8 h) to phase transition (12 h) as Ag intercalation time increases, achieving a 200 000 on/off ratio at a 4 nm thickness. MoTe 2 exhibits the most evident phase transition due to its low transition barrier (0.84) compared to other TMDs (>1.22). Intercalated Ag + ions provide outstanding memristive performance over Li + ions, with a 100 times higher on/off ratio, 300 times better retention, and 8 times lower non‐linearity (β Ag = 0.5–0.6, β Li = 4.0). Ag + MoTe 2 PTMEM achieves 91.7% accuracy in MNIST recognition, surpassing the 81.7% accuracy of Li + MoS 2 PTMEM. These findings demonstrate that Ag + MoTe 2 PTMEM holds great potential for advanced memory‐based neural network applications.

Article Details

Volume / Issue Vol. 38, Issue 5
Published January 01, 2026
ISSN 0935-9648
Publisher Unknown Publisher

Journal Info

Advanced Materials

Unknown Publisher

ISSN: 0935-9648 Physical Sciences

Authors (10)

W

Whan Kyun Kim

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

G

Ga Young Cho

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

T

Thi Thanh Huong Vu

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

J

Jun Sun Son

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

Y

Yong Ha Shin

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

H

Hong Woon Yun

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

M

Min Seok Kim

Y

Yong Seon Shin

Department of Electrical and Computer Engineering Sungkyunkwan University Suwon 16419 Republic of Korea

H

Heejun Yang

W

Woo Jong Yu