pith. sign in

arxiv: 2605.22943 · v1 · pith:XDBYCCC4new · submitted 2026-05-21 · ❄️ cond-mat.mtrl-sci

A new method to probe conducting filaments in MoS₂-based memristors

Pierre Trousset , Lucie Le Van-Jodin , Bruno Reig , Clotilde Ligaud , Thomas Jalabert , Hanako Okuno , Le Van-Hoan , Paul Brunet
show 2 more authors
St\'ephane Cadot Matthieu Jamet
This is my paper

Pith reviewed 2026-05-25 05:47 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci
keywords MoS2memristorsconducting filamentsmechanical exfoliationresistive switchingKelvin Probe Force MicroscopyTransmission Electron Microscopymetallic migration
0
0 comments X

The pith

Metallic atoms migrate from the top electrode into the MoS2 layer to form conducting filaments.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a mechanical exfoliation technique that peels away the top metal electrode from MoS2 memristors by taking advantage of weak van der Waals forces. This leaves the MoS2 surface intact for direct examination in its initial, low-resistance, and high-resistance states through Kelvin Probe Force Microscopy, Raman mapping, and cross-sectional TEM. The measurements show that filaments arise when metal atoms move from the electrode into the MoS2. Different electrode metals produce different switching behaviors because their adsorption and diffusion energies differ. The approach therefore supplies direct evidence for the filament mechanism in these devices.

Core claim

The mechanical exfoliation technique selectively removes the top metallic electrode while preserving the MoS2 layer and any filaments in their operating states. Kelvin Probe Force Microscopy and Raman spectroscopy mapping of the exposed surface, together with cross-sectional TEM, establish that the conducting filament consists of metallic atoms that have migrated from the top electrode into the MoS2. The choice between gold and nickel electrodes alters the switching because the two metals differ in adsorption and diffusion energies with respect to the MoS2 surface.

What carries the argument

The mechanical exfoliation technique that removes only the top electrode via weak van der Waals forces, enabling surface and cross-sectional characterization of the same device in multiple states.

If this is right

  • Filament formation is driven by electrode-metal atom migration rather than vacancy movement or other processes.
  • Switching voltages and stability can be tuned by selecting electrode metals with appropriate adsorption and diffusion energies.
  • The exfoliation method allows repeated, non-destructive surface characterization of the same memristor across resistance states.
  • Direct multi-scale imaging of filaments becomes possible without sectioning every device.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same peel-and-image sequence could be tested on other layered materials to check whether electrode migration is a general filament route.
  • Mapping adsorption energies for additional electrode metals would give a practical selection rule for desired switching thresholds.
  • If the method preserves filament conductivity, it could support in-operando studies that track filament growth and dissolution in real time.

Load-bearing premise

The mechanical exfoliation technique selectively removes only the top electrode while leaving the MoS2 surface and any formed filaments unaltered in their conductive state.

What would settle it

Observation of identical KPFM potential maps or Raman spectra in the ON and OFF states after exfoliation, or absence of electrode metal atoms inside the MoS2 layer in TEM cross-sections of the ON state, would falsify the migration mechanism.

read the original abstract

Two-dimensional (2D) transition metal dichalcogenides (TMDs), such as molybdenum disulfide (MoS$_2$), are emerging as promising materials for next-generation electronic devices. They have proved to be serious candidates for integration with memristors in non-volatile memory and radio frequency (RF) applications. However, the physical mechanisms behind their resistive switching, particularly the formation and resorption of conducting filaments, remain unclear. In this study, we present a novel mechanical exfoliation technique that selectively removes the top metallic electrode from MoS$_2$-based memristors by exploiting the weak van der Waals interaction between MoS$_2$ and the top electrode. This method enables direct and multi-scale characterization of the MoS$_2$ surface in different states (initial, ON and OFF) using Kelvin Probe Force Microscopy (KPFM) and Raman spectroscopy mapping. To complete this study, cross-sectional Transmission Electron Microscopy (TEM) was also performed in different conductive states. Our results reveal that the conducting filament is formed by metallic atom migration from the top electrode into the MoS$_2$ layer. Additionally, we demonstrate that the choice of metallic electrodes (gold vs. nickel) significantly impacts the switching behavior due to differences in adsorption and diffusion energies. This work not only clarifies the filament formation mechanism and introduces a reproducible approach for in-operando characterization but also represents a real progress in the understanding and optimization of 2D material-based memristors.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces a mechanical exfoliation technique that exploits van der Waals interactions to selectively remove the top metallic electrode from MoS₂-based memristors. This enables direct multi-scale characterization of the MoS₂ surface in initial, ON, and OFF states via KPFM, Raman spectroscopy mapping, and cross-sectional TEM. The central claims are that conducting filaments form by metallic-atom migration from the top electrode into the MoS₂ layer and that the choice of top electrode (Au versus Ni) significantly alters switching behavior owing to differences in adsorption and diffusion energies.

Significance. If the exfoliation method is shown to be non-perturbative and the filament-composition interpretation is substantiated, the work supplies a practical route for in-operando surface and cross-sectional analysis of resistive-switching mechanisms in 2D TMD memristors. The multi-technique approach (KPFM + Raman + TEM) is a constructive contribution that could aid device optimization for memory and RF applications.

major comments (2)
  1. [Methods (exfoliation technique)] Methods section describing the mechanical exfoliation: the claim that filaments consist of migrated metallic atoms (and that Au/Ni differences arise from adsorption/diffusion energies) is interpreted from post-exfoliation KPFM, Raman, and TEM maps. No pre-/post-exfoliation I-V curves on the same devices or control experiments comparing filament visibility with versus without peel are reported. This verification is load-bearing; without it the surface features cannot be unambiguously linked to the original switching states.
  2. [Results (electrode comparison)] Results on electrode comparison: the statement that Au versus Ni electrodes produce distinct switching behavior 'due to differences in adsorption and diffusion energies' is presented without accompanying calculations, simulations, or direct experimental quantification of those energies. If the energies are taken from external literature, the link to the observed I-V differences must be made explicit; otherwise the causal attribution remains unsupported.
minor comments (2)
  1. All KPFM and Raman maps should include quantitative scale bars, color bars with units, and explicit labeling of initial/ON/OFF regions; error bars or statistics on filament size or potential contrast are needed to support the migration claim.
  2. The number of devices examined and any reproducibility metrics (e.g., yield of filament observation across devices) should be stated clearly.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thorough review and valuable comments. We address the major comments point-by-point below and will revise the manuscript to incorporate the requested verifications and clarifications.

read point-by-point responses

  1. Referee: [Methods (exfoliation technique)] Methods section describing the mechanical exfoliation: the claim that filaments consist of migrated metallic atoms (and that Au/Ni differences arise from adsorption/diffusion energies) is interpreted from post-exfoliation KPFM, Raman, and TEM maps. No pre-/post-exfoliation I-V curves on the same devices or control experiments comparing filament visibility with versus without peel are reported. This verification is load-bearing; without it the surface features cannot be unambiguously linked to the original switching states.

    Authors: We agree that direct verification of the exfoliation process is essential to link the post-exfoliation observations unambiguously to the switching states. In the revised manuscript we will add I-V curves acquired on the same devices before and after exfoliation, together with control experiments on devices that were not peeled, to demonstrate that the filament features are not introduced by the mechanical removal step. These data will be included in the Methods and Results sections. revision: yes

  2. Referee: [Results (electrode comparison)] Results on electrode comparison: the statement that Au versus Ni electrodes produce distinct switching behavior 'due to differences in adsorption and diffusion energies' is presented without accompanying calculations, simulations, or direct experimental quantification of those energies. If the energies are taken from external literature, the link to the observed I-V differences must be made explicit; otherwise the causal attribution remains unsupported.

    Authors: The attribution relies on literature values for adsorption and diffusion energies of Au and Ni on MoS2. In the revision we will cite the specific references, explicitly connect the reported energy differences to the measured I-V characteristics (e.g., lower set voltage for Ni), and add a short discussion paragraph making this causal link quantitative where possible. No new calculations will be performed, but the connection to external data will be stated clearly. revision: yes

Circularity Check

0 steps flagged

No circularity: purely experimental observations without derivations or fits

full rationale

The manuscript describes an experimental technique (mechanical exfoliation) followed by KPFM, Raman mapping, and cross-sectional TEM to observe surface features in initial/ON/OFF states of MoS2 memristors with Au or Ni electrodes. No equations, fitted parameters, predictions, or self-citations appear in the load-bearing claims. The filament-migration conclusion and electrode-material comparison are presented as direct interpretations of the post-exfoliation data; they do not reduce to any input by construction, renaming, or self-referential fitting. The work is therefore self-contained against external benchmarks and receives the default non-circularity score.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work rests on the domain assumption that van der Waals forces between MoS2 and the top metal are weak enough for clean delamination without surface damage or filament alteration; no free parameters or new entities are introduced.

axioms (1)
  • domain assumption Weak van der Waals interaction between MoS2 and top electrode permits selective mechanical removal without disturbing the MoS2 or filament
    Invoked in the abstract as the enabling principle of the new method.

pith-pipeline@v0.9.0 · 5842 in / 1239 out tokens · 29909 ms · 2026-05-25T05:47:07.499763+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

37 extracted references · 37 canonical work pages · 1 internal anchor

  1. [1]

    A. V. Kolobov et J. Tominaga, Two-Dimensional Transition-Metal Dichalcogenides, vol. 239. in Springer Series in Materials Science, vol. 239. Cham: Springer International Publishing, 2016. doi: 10.1007/978-3-319-31450-1

    work page doi:10.1007/978-3-319-31450-1 2016

  2. [2]

    K. S. Novoselov et al., « Electric Field Effect in Atomically Thin Carbon Films », Science, vol. 306, no 5696, Art. no 5696, oct. 2004, doi: 10.1126/science.1102896

    work page internal anchor Pith review doi:10.1126/science.1102896 2004

  3. [3]

    Cadot et al., « A novel 2-step ALD route to ultra-thin MoS2 films on SiO2 through a surface organometallic intermediate », Nanoscale, vol

    S. Cadot et al., « A novel 2-step ALD route to ultra-thin MoS2 films on SiO2 through a surface organometallic intermediate », Nanoscale, vol. 9, no 2, p. 538‑546, 2017, doi: 10.1039/c6nr06021h

    work page doi:10.1039/c6nr06021h 2017

  4. [4]

    G. R. Bhimanapati et al., « Recent Advances in Two-Dimensional Materials beyond Graphene », ACS Nano, vol. 9, no 12, p. 11509‑11539, 2015, doi: 10.1021/acsnano.5b05556

    work page doi:10.1021/acsnano.5b05556 2015

  5. [5]

    M. T. Dau et al., « Beyond van der Waals Interaction: The Case of MoSe2 Epitaxially Grown on Few-Layer Graphene », ACS Nano, vol. 12, no 3, p. 2319‑2331, mars 2018, doi: 10.1021/acsnano.7b07446

    work page doi:10.1021/acsnano.7b07446 2018

  6. [7]

    J. Wang, J. Suo, Z. Song, W. J. Li, et Z. Wang, « Nanomaterial-based flexible sensors for metaverse and virtual reality applications », Int. J. Extrem. Manuf., vol. 5, no 3, p. 032013, juill. 2023, doi: 10.1088/2631-7990/acded1

    work page doi:10.1088/2631-7990/acded1 2023

  7. [8]

    Radisavljevic , author A

    B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti, et A. Kis, « Single-layer MoS2 transistors », Nature Nanotech, vol. 6, no 3, Art. no 3, mars 2011, doi: 10.1038/nnano.2010.279

    work page doi:10.1038/nnano.2010.279 2011

  8. [9]

    M. A. Fanton et al., « Characterization of Graphene Films and Transistors Grown on Sapphire by Metal-Free Chemical Vapor Deposition », ACS Nano, vol. 5, no 10, Art. no 10, oct. 2011, doi: 10.1021/nn202643t

    work page doi:10.1021/nn202643t 2011

  9. [10]

    Kim et al., « High performance field-effect transistors based on partially suspended 2D materials via block copolymer lithography », Polymers, vol

    S. Kim et al., « High performance field-effect transistors based on partially suspended 2D materials via block copolymer lithography », Polymers, vol. 13, no 4, Art. no 4, 2021, doi: 10.3390/polym13040566

    work page doi:10.3390/polym13040566 2021

  10. [11]

    Tong et al., « Contact optimisation strategy for wafer-scale field-effect transistors based on two-dimensional semiconductors », Journal of Materials Science & Technology, vol

    L. Tong et al., « Contact optimisation strategy for wafer-scale field-effect transistors based on two-dimensional semiconductors », Journal of Materials Science & Technology, vol. 133, p. 230‑237, janv. 2023, doi: 10.1016/j.jmst.2022.05.055

    work page doi:10.1016/j.jmst.2022.05.055 2023

  11. [12]

    Ma et al., « A universal etching-free transfer of MoS2 films for applications in photodetectors », Nano Res., vol

    D. Ma et al., « A universal etching-free transfer of MoS2 films for applications in photodetectors », Nano Res., vol. 8, no 11, Art. no 11, nov. 2015, doi: 10.1007/s12274-015-0866- z

    work page doi:10.1007/s12274-015-0866- 2015

  12. [13]

    Xia et al., « CVD synthesis of large-area, highly crystalline MoSe2 atomic layers on diverse substrates and application to photodetectors », Nanoscale, vol

    J. Xia et al., « CVD synthesis of large-area, highly crystalline MoSe2 atomic layers on diverse substrates and application to photodetectors », Nanoscale, vol. 6, no 15, p. 8949‑8955, juill. 2014, doi: 10.1039/C4NR02311K

    work page doi:10.1039/c4nr02311k 2014

  13. [14]

    Wu et al., « Phase-controlled van der Waals growth of wafer-scale 2D MoTe2 layers for integrated high-sensitivity broadband infrared photodetection », Light Sci Appl, vol

    D. Wu et al., « Phase-controlled van der Waals growth of wafer-scale 2D MoTe2 layers for integrated high-sensitivity broadband infrared photodetection », Light Sci Appl, vol. 12, no 1, Art. no 1, janv. 2023, doi: 10.1038/s41377-022-01047-5

    work page doi:10.1038/s41377-022-01047-5 2023

  14. [15]

    Anichini, W

    C. Anichini, W. Czepa, D. Pakulski, A. Aliprandi, A. Ciesielski, et P. Samorì, « Chemical sensing with 2D materials », Chem. Soc. Rev., vol. 47, no 13, Art. no 13, juill. 2018, doi: 10.1039/C8CS00417J

    work page doi:10.1039/c8cs00417j 2018

  15. [16]

    Z. Meng, R. M. Stolz, L. Mendecki, et K. A. Mirica, « Electrically-transduced chemical sensors based on two-dimensional nanomaterials », Chemical Reviews, vol. 119, no 1, Art. no 1, 2019, doi: 10.1021/acs.chemrev.8b00311

    work page doi:10.1021/acs.chemrev.8b00311 2019

  16. [17]

    Kim et al., « Zero-static power radio-frequency switches based on MoS2 atomristors », Nat Commun, vol

    M. Kim et al., « Zero-static power radio-frequency switches based on MoS2 atomristors », Nat Commun, vol. 9, no 1, p. 2524, déc. 2018, doi: 10.1038/s41467-018-04934-x

    work page doi:10.1038/s41467-018-04934-x 2018

  17. [18]

    Kim et al., « Monolayer molybdenum disulfide switches for 6G communication systems », Nat Electron, vol

    M. Kim et al., « Monolayer molybdenum disulfide switches for 6G communication systems », Nat Electron, vol. 5, no 6, Art. no 6, juin 2022, doi: 10.1038/s41928-022-00766-2

    work page doi:10.1038/s41928-022-00766-2 2022

  18. [19]

    Kim et al., « Emerging memory electronics for non-volatile radiofrequency switching technologies », Nat Rev Electr Eng, vol

    D. Kim et al., « Emerging memory electronics for non-volatile radiofrequency switching technologies », Nat Rev Electr Eng, vol. 1, no 1, Art. no 1, janv. 2024, doi: 10.1038/s44287-023- 00001-w

    work page doi:10.1038/s44287-023- 2024

  19. [20]

    U. Das, S. Bhattacharjee, P. K. Sarkar, et A. Roy, « A multi-level bipolar memristive device based on visible light sensing MoS2 thin film », Mater. Res. Express, vol. 6, no 7, Art. no 7, avr. 2019, doi: 10.1088/2053-1591/ab154d

    work page doi:10.1088/2053-1591/ab154d 2019

  20. [21]

    Bhattacharjee et al., « Insights into Multilevel Resistive Switching in Monolayer MoS2 », ACS Appl

    S. Bhattacharjee et al., « Insights into Multilevel Resistive Switching in Monolayer MoS2 », ACS Appl. Mater. Interfaces, vol. 12, no 5, Art. no 5, févr. 2020, doi: 10.1021/acsami.9b15677

    work page doi:10.1021/acsami.9b15677 2020

  21. [22]

    Chen et al., « The plasticity of synaptic memristor based on 2D-MoS2 thin film prepared in large-scale by a PLD-assisted CVD method », Materials Today Communications, vol

    J. Chen et al., « The plasticity of synaptic memristor based on 2D-MoS2 thin film prepared in large-scale by a PLD-assisted CVD method », Materials Today Communications, vol. 35, p. 105511, juin 2023, doi: 10.1016/j.mtcomm.2023.105511

    work page doi:10.1016/j.mtcomm.2023.105511 2023

  22. [23]

    Wang et al., « All-atomristor logic gates », Nano Res., oct

    S. Wang et al., « All-atomristor logic gates », Nano Res., oct. 2022, doi: 10.1007/s12274-022- 5042-7

    work page doi:10.1007/s12274-022- 2022

  23. [24]

    Ligaud et al., « Development and optimization of large-scale integration of 2D material in memristors », 2D Mater., vol

    C. Ligaud et al., « Development and optimization of large-scale integration of 2D material in memristors », 2D Mater., vol. 11, no 4, Art. no 4, juill. 2024, doi: 10.1088/2053-1583/ad5bd6

    work page doi:10.1088/2053-1583/ad5bd6 2024

  24. [25]

    Wu et al., « Thinnest Nonvolatile Memory Based on Monolayer h-BN », Advanced Materials, vol

    X. Wu et al., « Thinnest Nonvolatile Memory Based on Monolayer h-BN », Advanced Materials, vol. 31, no 15, Art. no 15, 2019, doi: 10.1002/adma.201806790

    work page doi:10.1002/adma.201806790 2019

  25. [26]

    Ge et al., « Atomristor: Nonvolatile Resistance Switching in Atomic Sheets of Transition Metal Dichalcogenides », Nano Lett., vol

    R. Ge et al., « Atomristor: Nonvolatile Resistance Switching in Atomic Sheets of Transition Metal Dichalcogenides », Nano Lett., vol. 18, no 1, p. 434‑441, janv. 2018, doi: 10.1021/acs.nanolett.7b04342

    work page doi:10.1021/acs.nanolett.7b04342 2018

  26. [27]

    Mitra et S

    S. Mitra et S. Mahapatra, « Atomistic description of conductive bridge formation in two- dimensional material based memristor », npj 2D Mater Appl, vol. 8, no 1, Art. no 1, mars 2024, doi: 10.1038/s41699-024-00465-w

    work page doi:10.1038/s41699-024-00465-w 2024

  27. [28]

    Figueroa

    X.-D. Li, B.-Q. Wang, N.-K. Chen, et X.-B. Li, « Resistive switching mechanism of MoS2 based atomristor », Nanotechnology, vol. 34, no 20, Art. no 20, févr. 2023, doi: 10.1088/1361- 6528/acb69d

    work page doi:10.1088/1361- 2023

  28. [29]

    Zhang, S

    Y. Zhang, S. Sang, Y. Ge, X. Chai, et Y. Liu, « Unraveling the defect-induced resistance mechanism of MoS2 memristors at the atomic scale », Applied Surface Science, vol. 715, p. 164614, janv. 2026, doi: 10.1016/j.apsusc.2025.164614

    work page doi:10.1016/j.apsusc.2025.164614 2026

  29. [30]

    M. D. Ganeriwala et al., « Effect of grain boundaries on metal atom migration and electronic transport in 2D TMD-based resistive switches », Nanoscale, mai 2025, doi: 10.1039/D4NR05321D

    work page doi:10.1039/d4nr05321d 2025

  30. [31]

    Tang et al., « Wafer-scale solution-processed 2D material analog resistive memory array for memory-based computing », Nat Commun, vol

    B. Tang et al., « Wafer-scale solution-processed 2D material analog resistive memory array for memory-based computing », Nat Commun, vol. 13, no 1, Art. no 1, juin 2022, doi: 10.1038/s41467-022-30519-w

    work page doi:10.1038/s41467-022-30519-w 2022

  31. [32]

    Lanza, F

    M. Lanza, F. Hui, C. Wen, et A. C. Ferrari, « Resistive Switching Crossbar Arrays Based on Layered Materials », Advanced Materials, vol. 35, no 9, Art. no 9, 2023, doi: 10.1002/adma.202205402

    work page doi:10.1002/adma.202205402 2023

  32. [33]

    S. M. Hus et al., « Observation of single-defect memristor in an MoS2 atomic sheet », Nat. Nanotechnol., vol. 16, no 1, p. 58‑62, janv. 2021, doi: 10.1038/s41565-020-00789-w

    work page doi:10.1038/s41565-020-00789-w 2021

  33. [34]

    J. W. Suk et al., « Transfer of CVD-Grown Monolayer Graphene onto Arbitrary Substrates », ACS Nano, vol. 5, no 9, Art. no 9, sept. 2011, doi: 10.1021/nn201207c

    work page doi:10.1021/nn201207c 2011

  34. [35]

    Moon et al., « Layer-engineered large-area exfoliation of graphene », Science Advances, vol

    J.-Y. Moon et al., « Layer-engineered large-area exfoliation of graphene », Science Advances, vol. 6, no 44, p. eabc6601, oct. 2020, doi: 10.1126/sciadv.abc6601

    work page doi:10.1126/sciadv.abc6601 2020

  35. [36]

    Rao et al., « Spectroscopic evaluation of charge-transfer doping and strain in graphene/${\mathrm{MoS}}_{2}$ heterostructures », Phys

    R. Rao et al., « Spectroscopic evaluation of charge-transfer doping and strain in graphene/${\mathrm{MoS}}_{2}$ heterostructures », Phys. Rev. B, vol. 99, no 19, Art. no 19, mai 2019, doi: 10.1103/PhysRevB.99.195401

    work page doi:10.1103/physrevb.99.195401 2019

  36. [37]

    S. E. Panasci et al., « Strain, Doping, and Electronic Transport of Large Area Monolayer MoS2 Exfoliated on Gold and Transferred to an Insulating Substrate », ACS Appl. Mater. Interfaces, vol. 13, no 26, p. 31248‑31259, juill. 2021, doi: 10.1021/acsami.1c05185

    work page doi:10.1021/acsami.1c05185 2021

  37. [38]

    W. A. Saidi, « Trends in the Adsorption and Growth Morphology of Metals on the MoS2(001) Surface », Crystal Growth & Design, vol. 15, no 7, Art. no 7, juill. 2015, doi: 10.1021/acs.cgd.5b00269

    work page doi:10.1021/acs.cgd.5b00269 2015