Monolithic fabrication of 1S1R cross-point device using single GexTe1-x material by controlling composition between memory and threshold switching
Abstract Three-dimensional (3D) integrated one-selector one-resistor (1S1R) cross-point arrays are emerging as a leading next-generation memory technology. Achieving both nonvolatile memory switching and volatile threshold switching within a single material system offers tremendous potential for simplifying fabrication. This study demonstrates that the memory and threshold switching behavior of germanium-telluride (Ge x Te 1-x ) devices can be precisely tuned by controlling their atomic ratio. In the Te-deficient Ge x Te 1-x system, the device exhibits nonvolatile memory switching with fast operating speed. Conversely, the Te-rich system functions as an excellent bidirectional selector for high-density memory applications, exhibiting high selectivity (> 10 4 ) and a short delay time (< 55 ns). Material analyses using X-ray photoelectron spectroscopy and Raman spectroscopy reveal that these distinct electrical behaviors are driven by composition-dependent changes in the atomic structure. Furthermore, the fully integrated 1S1R stack demonstrates robust SET/RESET switching characteristics, maintaining a stable memory window of 0.2 V and an on/off current ratio exceeding two orders of magnitude (~ 1.12 × 10 2 ). Notably, this integrated configuration provides superior operational reliability and enhanced endurance (~ 8 × 10 6 cycles) compared to standalone phase-change memory devices. By elucidating these underlying mechanisms, this work reduces the fabrication complexity of stackable devices and opens up wide-ranging applications for unified memory and selector systems.
Authors
- 이지혜
- Sangheon Park (ORCID: https://orcid.org/0000-0001-8563-7681)
- Jong‐Souk Yeo (ORCID: https://orcid.org/0000-0001-6452-2404)
- Young-Min Kim
- Siwon Park
- Sangyeop Kim
Institutions
- Yonsei University (KR)
Publication Details
- Journal
- Han-guk hyeonmigyeong hakoeji/Applied microscopy
- Published
- 2026-10-14
- DOI
- https://doi.org/10.1186/s42649-026-00145-7
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- National Research Foundation
- Ministry of Education
- Ministry of Science and ICT, South Korea
- Samsung