Non‐Destructive, Reference‐Free Quantitative Analysis of TaOx Memristive Devices Using Soft X‐Ray Radiation
ABSTRACT Valence change memristive devices based on tantalum‐oxides (TaO x ) exhibit excellent switching performance due to thermodynamics of the oxide phases, resulting in a robust switching mechanism, based on predominant movement of oxygen ions. The accurate physical details of how the conducting filaments form at a quantitative level remain only partially understood, largely because traditional characterization techniques only provide indirect, device‐level insights into the complex nanoscale switching dynamics. In this work, we present a quantitative, reference free, non‐destructive approach to investigating TaO x ‐based memristive devices using nano‐X‐ray fluorescence analysis with monochromatic synchrotron radiation in the soft X‐ray regime. Several spots with different origins and compositions were observed, indicating the switching dynamics are more complex than supposed. Our approach enables direct spatially resolved probing of elemental distributions within the device, including those in the buried layers critical for the resistive switching.
Authors
- Christian Stadelhoff
- André Wählisch (ORCID: https://orcid.org/0000-0002-6126-1306)
- Andreas Haidl
- Ilia Valov (ORCID: https://orcid.org/0000-0002-0728-7214)
- Rainer Unterumsberger (ORCID: https://orcid.org/0000-0001-8076-1288)
- Jan Weser
- Leona Bauer (ORCID: https://orcid.org/0000-0002-7409-3495)
- Marc Dummin
- Xin Zheng (ORCID: https://orcid.org/0009-0006-8834-6672)
- Burkhard Beckhoff
Institutions
- Physikalisch-Technische Bundesanstalt (DE)
- Koblenz University of Applied Sciences (DE)
- Forschungszentrum Jülich (DE)
- Ernst Ruska Centre (DE)
- Institute of Electrochemistry and Energy Systems (BG)
- Helios Hospital Berlin-Buch (DE)
- Technische Universität Berlin (DE)
Publication Details
- Journal
- Advanced Electronic Materials
- Published
- 2026-04-13
- DOI
- https://doi.org/10.1002/aelm.202500784
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Horizon 2020