Fast and robust multilevel optoelectronic memory based on van der Waals heterostructure

Optoelectronic memory (OEM) has attracted tremendous attention for its great potential to boost the storage capacity of memory chips and break through the von Neumann bottleneck in the post-Moore era. Two-dimensional (2D) van der Waals (vdW) heterostructures, formed by artificially stacking differen...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied physics letters 2024-02, Vol.124 (9)
Hauptverfasser: Liu, Tao, Wang, Yue, Cao, Yi, Tan, Feixia, Li, Honghong, Wang, Tinghao, Xiang, Du
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Optoelectronic memory (OEM) has attracted tremendous attention for its great potential to boost the storage capacity of memory chips and break through the von Neumann bottleneck in the post-Moore era. Two-dimensional (2D) van der Waals (vdW) heterostructures, formed by artificially stacking different 2D layered materials, offer tremendous possibilities in OEMs due to their extraordinary capability to integrate and process optical/electrical signals. However, the realization of 2D vdW OEMs with high writing speed and robust memory performance has long been challenging. Here, we report a 2D vdW OEM consisting of tungsten diselenide (WSe2) and hexagonal boron nitride, which functions based on the fast charge transfer dynamics at a 2D interface. The OEM demonstrates high writing speed reaching up to 50 μs, approximately one order of magnitude faster than those of other 2D OEMs. Moreover, the outstanding robustness of such OEM is demonstrated by long retention time exceeding 14 days, together with a broad temperature endurance window from 100 to 420 K. Additionally, through continuously switching laser pulse on the OEM, we achieve 17 distinct current levels (over 4-bit storage) with random access. Our findings envision 2D vdW heterostructure-based OEM as a potential platform to overcome the “memory wall” in the conventional von Neumann configuration and to promote a promising paradigm for big data storage.
ISSN:0003-6951
1077-3118
DOI:10.1063/5.0188750