Pith. sign in

REVIEW 4 major objections 6 minor 56 references

Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaOx/HfO2-Based ReRAMs using Reactive Molecular Dynamics Simulations

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read In bilayer TaOx/HfO2 ReRAM, forming begins with voltage-driven cation migration, and Joule heating accelerates it by creating oxygen vacancies near the filament edge.

desk verdict First atomistic MD of forming in bilayer TaOx/HfO2 gives a plausible VCM/TCM reconciliation, but the frozen-electrode assumption may predetermine the oxygen immobility it claims to explain. read the letter →

arxiv 2505.24468 v2 pith:BMBRSE52 submitted 2025-05-30 cond-mat.mtrl-sci physics.app-phphysics.comp-ph

Simanta Lahkar , Valeria Bragaglia , Behnaz Bagheri , Donato Francesco Falcone , Matteo Galetta , Marilyne Sousa , Aida Todri-Sanial This is my paper
classification cond-mat.mtrl-sciphysics.app-phphysics.comp-ph
keywords ReRAMelectroformingTaOx/HfO2bilayerreactivemoleculardynamicschargetransferionicpotentialEChemDIDoxygenvacancyfilamentJouleheating
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper tries to establish a division of labor in the forming of bilayer TaOx/HfO2 ReRAM devices: applied voltage rearranges metal cations, while heat supplies the defects that grow the filament. Reactive molecular dynamics at 1.2 V show tantalum ions moving away from the anode most strongly, hafnium ions next, and oxygen ions barely at all; the result is a tantalum-depleted, oxygen-rich screening layer at the anode and a cluster of oxygen vacancies at the cathode where the filament nucleates. Raising the temperature to 900–1300 K accelerates filament growth not by speeding vertical ion drift but by thermally generating oxygen vacancies near the filament edge, where they are stabilized by the filament acting as a virtual cathode. A 0.6 V simulation nucleates no filament even at 1300 K, implying a voltage threshold for vacancy clustering. If correct, the picture reconciles the valence-change and thermochemical descriptions for bilayer devices and explains why hotter devices form at lower voltages.

What carries the argument

The argument runs on a reactive molecular-dynamics scheme in which the charge transfer ionic potential (CTIP)—a hybrid potential treating metallic and ionic bonding with dynamic atomic charges—is coupled to the electrochemical dynamics with implicit degrees of freedom (EChemDID) method, which propagates the electrode potential through conductive clusters in the dielectric. Under applied bias, electrode electronegativities are shifted, atomic charges are re-equilibrated, and a diffusion equation carries the local potential through connected metallic atoms so the growing filament behaves as a virtual cathode. Conductive metal atoms are identified by oxygen coordination of 5 or below, and the filament is defined by cluster analysis with a 3.9 Å cutoff. This machinery is what lets the paper separate field-driven vertical displacement from thermally activated lateral motion and track filament size at 300, 900, and 1300 K.

What would settle it

Track oxygen and metal atom positions during early-stage forming in a real TiN/TaOx/HfO2 device held below 1.2 V—using elemental mapping in a transmission electron microscope or isotopic oxygen tracers. If oxygen ions migrate substantially toward the anode before tantalum depletion appears, or if a filament nucleates at 0.6 V while the device is held at 1300 K, the paper's cation-first, no-oxygen-exchange picture would be contradicted.

Watch

Extended reading notes

Core claim

The central claim is that electroforming in a pristine bilayer TaOx/HfO2 stack under 1.2 V is initiated by electric-field-driven migration of metal cations—Ta more than Hf—away from the positive electrode, not by oxygen anion migration. This creates a Ta-depleted, oxygen-rich region at the anode that partially screens the bulk dielectric from the field, while oxygen vacancies accumulate at the cathode and cluster into the seed of the conductive filament. Joule heating then acts by increasing the generation rate of oxygen-vacancy defects in the layer near the filament tip; these defects aggregate onto the filament because the filament, as a virtual cathode, stabilizes them. The paper further claims a threshold voltage between 0.6 and 1.2 V below which no filament nucleates even at 1300 K, so temperature accelerates forming only once voltage has enabled clustering. Together these claims assign the electric field the role of cation rearrangement and nucleation, and temperature the role of vacancy generation and growth.

Load-bearing premise

The load-bearing premise is that the frozen electrodes, which cannot absorb or release oxygen, faithfully represent real TiN and conductive TaOx contacts during the early stage of forming; if those contacts exchange oxygen with the dielectric under a 1.2 V bias, the simulated suppression of oxygen motion could be an artifact.

Editorial extensions

If this is right

  • Limiting Joule heating—through a compliance current or a series resistor—should yield a smaller filament and lower post-forming conductance, because temperature's main role is defect generation rather than field-driven drift.
  • The anodic Ta-depleted, oxygen-rich layer shields the bulk from the field, which explains why forming in these devices requires a high applied voltage (experimentally above 4 V) rather than continuous anodic electroreduction.
  • Raising the device's ambient temperature lowers the required forming voltage; the authors report observing a linear correlation in similar bilayer devices.
  • Below the nucleation threshold, extra heat alone cannot start the filament; the applied voltage must first overcome vacancy–vacancy repulsion to make clusters stick.
  • The combined field-plus-thermal mechanism reconciles valence-change and thermochemical behaviors observed in TaOx and HfO2 ReRAMs by giving each model a different stage of forming.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The cation-first response implies a design lever the paper does not explore: forming voltage could be tuned by doping or interface modifications that change Ta/Hf mobility or cation charge, rather than by engineering oxygen transport.
  • Because the central picture depends on oxygen-inert electrodes, a natural test is to rerun the same stack with electrodes that can absorb oxygen; if oxygen migration becomes substantial, the no-exchange boundary condition is responsible for part of the result.
  • If the threshold behavior transfers to single-layer TaOx and HfO2 devices, it would unify their apparently conflicting VCM/TCM observations; the authors gesture at this, but the extrapolation is ours.
  • A compact circuit model could encode a temperature-activated vacancy-generation rate localized at the filament tip, which would make electroforming-voltage predictions depend on thermal resistance and ambient temperature rather than on vacancy drift alone.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript presents reactive molecular dynamics simulations of a Ta2O5/HfO2 bilayer ReRAM stack under applied bias, using an extended CTIP+EChemDID method implemented in LAMMPS. The model is built from STEM/EDS data of real devices, with amorphous Ta2O5 and HfO2 functional layers between frozen Ta2O5 electrodes. Under 1.2 V at 300 K for 500 ps, the authors observe downward drift of Ta and Hf cations, with small upward displacement of O, leading to a Ta-depleted O-rich anodic layer; conductive metal atoms cluster near the cathode and nucleate a filament. At 900 and 1300 K, filament size increases, attributed to thermally activated generation and agglomeration of oxygen vacancies near the filament edge rather than enhanced vertical field-driven migration. At 0.6 V no filament nucleates even at 1300 K, leading to the claim of a threshold voltage for vacancy clustering. The paper concludes that forming in these bilayer devices decouples electric-field-driven cation redistribution from thermally driven vacancy generation, reconciling VCM and TCM pictures.

Significance. The central qualitative mechanism is novel and plausible: it provides a concrete atomistic rationale for why Ta cations are mobile during forming despite their low diffusion constant, and it offers a way to reconcile conflicting VCM/TCM observations in TaOx/HfO2 devices. The paper's strengths include the experimentally informed stack geometry, the use of a literature-optimized variable-charge potential with validation against cohesive energies and radial distribution functions, the explicit EChemDID implementation and promised code release, and the falsifiable predictions (threshold voltage, temperature-dependent forming). If the results are confirmed by ensemble-replicated and boundary-condition-robust simulations, they would be a significant contribution to the atomistic understanding of electroforming in bilayer ReRAMs.

major comments (4)
  1. [Results, 'Device Layer Structure and Modeling Approach' (also Methodology, 'External Electrochemical Potential…] The frozen-electrode, no-oxygen-exchange boundary condition is load-bearing for the paper's central observation that oxygen ions respond only minimally under bias. The text asserts (p. 8) that TiN and conductive TaOx have low oxygen affinities and do not exchange O, but the cited ref. 11 is titled 'Exchange of Ions across the TiN/TaOx Interface during Electroformation' and documents TiN taking up oxygen during forming. Because both Ta2O5 electrode regions are frozen, oxygen anions in the model cannot leave the functional dielectric; the anodic O-rich zone and the minimal O displacement could therefore be enforced by the boundary condition rather than by intrinsic electrostatics. The authors should either relax this assumption (e.g., allow O exchange at the anode, or use an oxygen reservoir) or provide a quantitative justification, such as O-affinity data or a control simulation, before the VCM/TCM decoupling claim can be considered robust.
  2. [Results, 'Atomistic Response to 1.2 V...' and 'Atomistic response to 0.6 V...' (Figs. 4-10)] All composition profiles, displacement profiles, and charge distributions are derived from a single 500 ps MD trajectory per condition. Figure 7(d) reports error bars from temporal averaging over the final 100 ps of one run, not run-to-run variability. Molecular dynamics is stochastic, and the displacement hierarchy Ta > Hf > O, the anodic depletion profile, and the threshold behavior are quantitative claims that require at least several independent initializations (e.g., different random seeds or melt-quench replicas) with reported means and standard deviations. Without such replicates, the central quantitative conclusions are not statistically supported.
  3. [Results, 'Atomistic response to 0.6 V...' and 'Discussion and Conclusions'] The claimed threshold voltage for filament nucleation is inferred from exactly two bias conditions: 0.6 V (no filament) and 1.2 V (filament), each from one trajectory. A threshold cannot be located from two points, and no error bar is attached to the 'no nucleation' outcome. A proper threshold characterization would require scanning several voltages (for example, 0.7-1.1 V) at multiple temperatures and reporting nucleation probability or average filament size, ideally with ensemble statistics. This is load-bearing because the voltage-gating of vacancy clustering is one of the paper's main conclusions.
  4. [Results, 'Atomistic Response to 1.2 V...' (Figs. 6(c-d) and p. 15)] The mechanism attributing the cation-dominated displacement to larger per-atom charges on metal ions than on oxygen relies on the CTIP charge equilibration model with imposed charge bounds. The reported mean charges are not validated against independent electronic-structure data, and the charge bounds could artificially reduce the oxygen partial charges and thus the electrostatic force on O. The authors should compare the CTIP charges with DFT-derived Bader charges or perform a sensitivity test varying the charge bounds, since this explanation is central to the proposed field-driven cation migration.
minor comments (6)
  1. [Figure 9 caption] The caption says '(a) and (b) show the atomic snapshots' and then '(b) and (d) plot the profiles'; the correct references should be (a)/(c) for snapshots and (b)/(d) for profiles.
  2. [Methodology, Eq. (4)] The effective diffusivity k is never given a numerical value; for reproducibility, report k (and its units) or state how it was set relative to the chosen iteration count.
  3. [Results, p. 13] The phrase 'above the 50 Åz position' should read 'above z = 50 Å' or 'above the 50 Å z-position'.
  4. [Discussion and Conclusions, p. 24] The sentence 'can aide future all-atom reactive MD simulations' should be 'can aid future all-atom reactive MD simulations'.
  5. [Supporting Information section] The GitHub repository is mentioned but no URL is given; include a link or a DOI in the final version for reproducibility.
  6. [Figure 5(a)] The y-axis label 'Vertical displacement' should specify the sign convention (positive upward vs downward) so that the 'downward' direction claims can be interpreted unambiguously.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the mechanisms are produced by reactive MD simulation, not fitted to the conclusion or reduced by self-citation.

full rationale

The paper's central claims—Ta and Hf cations displacing more than O ions under 1.2 V, formation of a Ta-depleted/O-rich anodic zone, cathode-side vacancy clustering, voltage threshold between 0.6 and 1.2 V, and thermally enhanced filament growth—are read out from MD trajectories, not defined into existence. The charge equilibration and voltage propagation methods are cited from independent prior work (refs 37-39, 46), and the interatomic potential parameters are taken from the literature rather than fitted to the conclusions. Self-citations such as refs 15 and 41 supply STEM/EDS device geometry and prior electrical characterization, which are external empirical inputs, not results equivalent to the simulation outcome. The frozen-electrode and no-oxygen-exchange boundary condition is a modeling assumption that may affect physical validity, especially given the cited literature on TiN/TaOx ion exchange, but it does not make the observed ionic displacement hierarchy or threshold behavior true by construction: O ions are still free to move within the functional layers, and no equation in the paper equates the predicted response to the boundary condition. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from the authors' prior work, and the VCM/TCM reconciliation is presented as an interpretation of independent simulation results. Thus no circular step meeting the evidence standard is present.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central mechanistic story rests on the CTIP forcefield, the frozen electrode boundary condition, the conductive-atom criterion, and single short trajectories. None of these is validated against quantitative experimental measurements of ion displacement in this specific device, so the simulation outcome is best treated as a mechanistic hypothesis.

free parameters (3)
  • Conductive metal coordination cutoff = O coordination <= 5 (from ref. 10)
    Used to define which Ta and Hf atoms are metallic and to identify the filament. The criterion is taken from a HfO2 study and applied to Ta2O5, which has a different oxygen stoichiometry.
  • Cluster and conductive cutoff distance RC = 3.9 Angstrom (from ref. 10)
    Defines which metallic atoms are connected in the filament network and is not recalibrated for this bilayer stack.
  • Potential propagation diffusivity k and iteration count = not specified; 10 iterations per MD step
    Effective diffusivity from the EChemDID literature. The authors state the exact dynamics are not critical, but this parameter controls the local potential distribution on evolving metallic atoms.
assumptions (6)
  • domain assumption The CTIP forcefield parameters from Wu et al. accurately describe charge transfer, redox chemistry, and ionic mobility in amorphous Ta2O5 and HfO2 bilayer stacks under bias.
    The central ion displacement and charge results depend entirely on this potential, yet validation is limited to crystalline cohesive energies and radial distribution functions.
  • domain assumption The frozen electrode approximation is valid: TiN and conductive TaOx electrodes do not exchange oxygen with the dielectric during early forming.
    The model freezes electrode atoms and assumes low oxygen affinity. The observed minimal oxygen displacement could be a consequence of this boundary condition rather than a general material property.
  • domain assumption Melt-quench generated amorphous models represent the functional dielectric layers of the real device.
    Amorphous structures are matched to STEM/EDS layer thicknesses, but the real TaOy interlayer composition is approximated by stoichiometric Ta2O5.
  • domain assumption 500 ps MD trajectories are representative of the early forming mechanisms despite experimental forming occurring on microsecond timescales.
    The authors acknowledge the timescale gap and restrict claims to early stages, but extrapolation to the full forming process is an assumption.
  • domain assumption Oxygen coordination less than or equal to 5 identifies conductive cations in both Ta2O5 and HfO2.
    The criterion is taken from a HfO2 study and applied to Ta for simplicity. The paper notes it is stricter for Ta2O5 because of higher oxygen content, which may bias where filaments are detected.
  • ad hoc to paper A single MD trajectory per condition is representative, with no ensemble averaging.
    Key composition and displacement profiles are presented without error bars. Thermal fluctuations in a small 2.5 by 2.5 nanometer cell could be substantial.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaOx/HfO2-Based ReRAMs using Reactive Molecular Dynamics Simulations." pith.science (2026). https://pith.science/paper/BMBRSE52

@misc{pith2026250524468,
  author       = {Pith},
  title        = {Pith review of: Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaOx/HfO2-Based ReRAMs using Reactive Molecular Dynamics Simulations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BMBRSE52}},
  note         = {Machine review of arXiv:2505.24468}
}
read the original abstract

Resistive random access memories (ReRAMs) with a bilayer TaOx/HfO2 stack structure have shown unique multi-level resistive switching capabilities. However, the physical processes governing their behavior, and specifically the atomistic mechanisms of forming, remain poorly understood. In this work, we present a detailed analysis of the forming mechanism at the atomic level using molecular dynamics (MD) simulations. An extended charge equilibration scheme, based on a combination of the charge transfer ionic potential (CTIP) formalism and the electrochemical dynamics with implicit degrees of freedom (EChemDID) method, is employed to model the localized effects of applied voltage. Our simulations reveal that tantalum ions exhibit the highest displacement under applied voltage, followed by hafnium ions, while oxygen ions respond only minimally. This results in the formation of a tantalum-depleted, oxygen-rich zone near the positive top electrode (anode), and the clustering of oxygen vacancies near the negative bottom electrode (cathode), where the conductive filament nucleates. This ionic segregation partially shields the bulk dielectric from the applied electric field, hindering further migration of ions in the vertical direction. We find that a minimum threshold voltage is required to initiate vacancy clustering. Filament growth proceeds through a localized mechanism, driven by thermally activated generation of oxygen vacancy defects, which are stabilized near the edge of the nucleated filament at the cathode.

Figures

Figures reproduced from arXiv: 2505.24468 by the authors.

Figure 1
Figure 1. Cross-section images of the bilayer ReRAM. (a) STEM Bright Field image of the [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Structural characteristics of the relaxed Ta [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Structural characteristics of the bilayer stack model at the end of 500 ps of MD [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Change in distribution of different ionic species in the device stack over 500 ps MD [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Variation in the motion of different ions across the device stack model under a bias [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Distribution of ionic charges across the device stack model at the start (initial) [PITH_FULL_IMAGE:figures/full_fig_p014_6.png]
Figure 7
Figure 7. Figure 7: Effect of temperature on the growth of filament after 500 ps simulation under [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: The change in average displacements of different ions for varying temperatures over [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]
Figure 9
Figure 9. Figure 9: Effect of temperature on the vacancy distribution and composition change in the [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: The change in average displacements of different ions for different temperatures [PITH_FULL_IMAGE:figures/full_fig_p021_10.png]
Figure 11
Figure 11. Figure 11: Schematic illustrating the possible atomic movements underlying forming in [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

56 extracted references · 55 canonical work pages

  1. [1]

    Y.; Govoreanu, B.; Goux, L.; Degraeve, R.; Fantini, A.; Kar, G

    Chen, Y. Y.; Govoreanu, B.; Goux, L.; Degraeve, R.; Fantini, A.; Kar, G. S.; Wouters, D. J.; Groeseneken, G.; Kittl, J. A.; Jurczak, M.; others Balancing SET/RESET Pulse for > 1010 Endurance in HfO2 Hf 1T1R Bipolar RRAM. IEEE Transactions on Electron devices 2012, 59, 3243–3249

    work page 2012

  2. [2]

    B.; Lee, D.; Lee, S

    Lee, M.-J.; Lee, C. B.; Lee, D.; Lee, S. R.; Chang, M.; Hur, J. H.; Kim, Y.-B.; Kim, C.- J.; Seo, D. H.; Seo, S.; others A fast, high-endurance and scalable non-volatile memory device made from asymmetric Ta2O5- x/TaO2- x bilayer structures. Nature materials 2011, 10, 625–630. 29

    work page 2011

  3. [3]

    Govoreanu, B.; Kar, G. S.; Chen, Y.; Paraschiv, V.; Kubicek, S.; Fantini, A.; Radu, I.; Goux, L.; Clima, S.; Degraeve, R.; others 10× 10nm 2 Hf/HfO x crossbar resistive RAM with excellent performance, reliability and low-energy operation. 2011 International Elec- tron Devices Meeting. 2011; pp 31–6

    work page 2011

  4. [4]

    Nanoionic memristive phenomena in metal oxides: the valence change mechanism

    Dittmann, R.; Menzel, S.; Waser, R. Nanoionic memristive phenomena in metal oxides: the valence change mechanism. Advances in Physics 2021, 70, 155–349

    work page 2021

  5. [5]

    M.; Li, D.; Cullen, D

    Ma, Y.; Goodwill, J. M.; Li, D.; Cullen, D. A.; Poplawsky, J. D.; More, K. L.; Bain, J. A.; Skowronski, M. Stable Metallic Enrichment in Conductive Filaments in TaOx-Based Re- sistive Switches Arising from Competing Diffusive Fluxes.Advanced Electronic Materials 2019, 5, 1800954

    work page 2019

  6. [6]

    Transport properties of oxygen vacancy filaments in metal/crystalline or amorphous HfO 2/metal structures

    Cartoixa, X.; Rurali, R.; Sune, J. Transport properties of oxygen vacancy filaments in metal/crystalline or amorphous HfO 2/metal structures. Physical Review B—Condensed Matter and Materials Physics 2012, 86, 165445

    work page 2012

  7. [7]

    M.; Th¨ oner, B.; Sch¨ onhals, A.; Menzel, S.; Wuttig, M.; Waser, R.; Sobolev, N

    Ros´ ario, C. M.; Th¨ oner, B.; Sch¨ onhals, A.; Menzel, S.; Wuttig, M.; Waser, R.; Sobolev, N. A.; Wouters, D. J. Correlation between the transport mechanisms in conduc- tive filaments inside Ta2O5-based resistive switching devices and in substoichiometric TaOx thin films. Applied physics letters 2018, 112

    work page 2018

  8. [8]

    M.; Th¨ oner, B.; Sch¨ onhals, A.; Menzel, S.; Meledin, A.; Barradas, N

    Ros´ ario, C. M.; Th¨ oner, B.; Sch¨ onhals, A.; Menzel, S.; Meledin, A.; Barradas, N. P.; Alves, E.; Mayer, J.; Wuttig, M.; Waser, R.; others Metallic filamentary conduction in valence change-based resistive switching devices: the case of TaO x thin film with x 1. Nanoscale 2019, 11, 16978–16990

    work page 2019

Show all 56 references
  1. [9]

    A.; Cullen, D

    Ma, Y.; Li, D.; Herzing, A. A.; Cullen, D. A.; Sneed, B. T.; More, K. L.; Nuhfer, N.; Bain, J. A.; Skowronski, M. Formation of the conducting filament in TaO x-resistive switching devices by thermal-gradient-induced cation accumulation. ACS applied mate- rials & interfaces 201...

    2018

  2. [10]

    L.; Islam, M

    Urquiza, M. L.; Islam, M. M.; Van Duin, A. C.; Cartoix` a, X.; Strachan, A. Atomistic insights on the full operation cycle of a HfO2-based resistive random access memory cell from molecular dynamics. ACS nano 2021, 15, 12945–12954

    2021

  3. [11]

    A.; Goodwill, J

    Ma, Y.; Cullen, D. A.; Goodwill, J. M.; Xu, Q.; More, K. L.; Skowronski, M. Exchange of Ions across the TiN/TaO x Interface during Electroformation of TaO x-Based Resistive Switching Devices. ACS applied materials & interfaces 2020, 12, 27378–27385

    2020

  4. [12]

    B.; Huang, X.; Zhang, W.; Zhang, Q.; Deng, N.; Shi, L.; Wong, H.-S

    Yao, P.; Wu, H.; Gao, B.; Eryilmaz, S. B.; Huang, X.; Zhang, W.; Zhang, Q.; Deng, N.; Shi, L.; Wong, H.-S. P.; others Face classification using electronic synapses. Nature communications 2017, 8, 15199

    2017

  5. [13]

    Improved synaptic behavior under identical pulses using AlO x/HfO 2 bilayer RRAM array for neuromor- phic systems

    Woo, J.; Moon, K.; Song, J.; Lee, S.; Kwak, M.; Park, J.; Hwang, H. Improved synaptic behavior under identical pulses using AlO x/HfO 2 bilayer RRAM array for neuromor- phic systems. IEEE Electron Device Letters 2016, 37, 994–997

    2016

  6. [14]

    Exploiting the switching dynamics of HfO2-based ReRAM devices for reliable analog memristive behavior

    C¨ uppers, F.; Menzel, S.; Bengel, C.; Hardtdegen, A.; Von Witzleben, M.; B¨ ottger, U.; Waser, R.; Hoffmann-Eifert, S. Exploiting the switching dynamics of HfO2-based ReRAM devices for reliable analog memristive behavior. APL materials 2019, 7

    2019

  7. [15]

    Advanced electronic ma- terials 2022, 8, 2200448

    Stecconi, T.; Guido, R.; Berchialla, L.; La Porta, A.; Weiss, J.; Popoff, Y.; Halter, M.; Sousa, M.; Horst, F.; D´ avila, D.; others Filamentary TaOx/HfO2 ReRAM Devices for Neural Networks Training with Analog In-Memory Computing. Advanced electronic ma- terials 2022, 8, 2200448

    2022

  8. [16]

    J.; Carta, F.; Horst, F.; Falcone, D

    Stecconi, T.; Bragaglia, V.; Rasch, M. J.; Carta, F.; Horst, F.; Falcone, D. F.; Ten Kate, S. C.; Gong, N.; Ando, T.; Olziersky, A.; others Analog resistive switch- ing devices for training deep neural networks with the novel tiki-taka algorithm. Nano Letters 2024, 24, 866–872

    2024

  9. [17]

    B.; Zhang, W.; Wu, D.; Deiss, S.; 31 Raina, P.; Qian, H.; Gao, B.; others A compute-in-memory chip based on resistive random-access memory

    Wan, W.; Kubendran, R.; Schaefer, C.; Eryilmaz, S. B.; Zhang, W.; Wu, D.; Deiss, S.; 31 Raina, P.; Qian, H.; Gao, B.; others A compute-in-memory chip based on resistive random-access memory. Nature 2022, 608, 504–512

    2022

  10. [18]

    Gong, N.; Rasch, M. J.; Seo, S.-C.; Gasasira, A.; Solomon, P.; Bragaglia, V.; Con- siglio, S.; Higuchi, H.; Park, C.; Brew, K.; others Deep learning acceleration in 14nm CMOS compatible ReRAM array: device, material and algorithm co-optimization. 2022 International Electron De...

    2022

  11. [19]

    F.; La Porta, A.; Jubin, D.; Offrein, B

    Choi, W.; Van Bodegraven, T.; Verest, J.; Maher, O.; Falcone, D. F.; La Porta, A.; Jubin, D.; Offrein, B. J.; Karg, S.; Bragaglia, V.; others Hardware Implementation of Ring Oscillator Networks Coupled by BEOL Integrated ReRAM for Associative Memory Tasks. 2025 IEEE Internatio...

    2025

  12. [20]

    Oscillatory neural networks for edge ai computing

    Delacour, C.; Carapezzi, S.; Abernot, M.; Boschetto, G.; Azemard, N.; Salles, J.; Gil, T.; Todri-Sanial, A. Oscillatory neural networks for edge ai computing. 2021 IEEE computer society annual symposium on VLSI (ISVLSI). 2021; pp 326–331

    2021

  13. [21]

    Computing with oscillators from theoretical underpinnings to applications and demonstrators

    Todri-Sanial, A.; Delacour, C.; Abernot, M.; Sabo, F. Computing with oscillators from theoretical underpinnings to applications and demonstrators. Npj unconventional com- puting 2024, 1, 1–16

    2024

  14. [22]

    Falcone, D. F.; Clerico, V.; Choi, W.; Stecconi, T.; Horst, F.; Begon-Lours, L.; Galetta, M.; La Porta, A.; Garg, N.; Alibart, F.; others All-in-One Analog AI Accel- erator: On-Chip Training and Inference with Conductive-Metal-Oxide/HfOx ReRAM Devices. arXiv preprint arXiv:250...

    2025 arXiv

  15. [23]

    F.; Menzel, S.; La Porta, A.; Stecconi, T.; Choi, W.; Of- frein, B

    Galetta, M.; Falcone, D. F.; Menzel, S.; La Porta, A.; Stecconi, T.; Choi, W.; Of- frein, B. J.; Bragaglia, V. Compact Model of Conductive-Metal-Oxide/HfO x Analog Filamentary ReRAM Devices. 2024 IEEE European Solid-State Electronics Research Conference (ESSERC). 2024; pp 749–752. 32

    2024

  16. [24]

    F.; Menzel, S.; Stecconi, T.; Galetta, M.; La Porta, A.; Offrein, B

    Falcone, D. F.; Menzel, S.; Stecconi, T.; Galetta, M.; La Porta, A.; Offrein, B. J.; Bragaglia, V. Analytical modelling of the transport in analog filamentary conductive- metal-oxide/HfO x ReRAM devices. Nanoscale Horizons 2024, 9, 775–784

    2024

  17. [25]

    F.; Menzel, S.; Stecconi, T.; La Porta, A.; Carraria-Martinotti, L.; Of- frein, B

    Falcone, D. F.; Menzel, S.; Stecconi, T.; La Porta, A.; Carraria-Martinotti, L.; Of- frein, B. J.; Bragaglia, V. Physical modeling and design rules of analog Conductive Metal Oxide-HfO 2 ReRAM. 2023 IEEE International Memory Workshop (IMW). 2023; pp 1–4

    2023

  18. [26]

    P.; Vine, D.; Kilcoyne, A

    Kumar, S.; Wang, Z.; Huang, X.; Kumari, N.; Davila, N.; Strachan, J. P.; Vine, D.; Kilcoyne, A. D.; Nishi, Y.; Williams, R. S. Conduction channel formation and dissolution due to oxygen thermophoresis/diffusion in hafnium oxide memristors. ACS nano 2016, 10, 11205–11210

    2016

  19. [27]

    E.; Strachan, J

    Kumar, S.; Graves, C. E.; Strachan, J. P.; Grafals, E. M.; Kilcoyne, A. L. D.; Tyliszczak, T.; Weker, J. N.; Nishi, Y.; Williams, R. S. Direct Observation of Local- ized Radial Oxygen Migration in Functioning Tantalum Oxide Memristors. Advanced Materials 2016, 28, 2771–2771

    2016

  20. [28]

    Redox-Based Resistive Switching Memories-Nanoionic Mechanisms, Prospects, and Challenges.Advanced Materials (Deer- field Beach, Fla.) 2009, 21, 2632–2663

    Waser, R.; Dittmann, R.; Staikov, G.; Szot, K. Redox-Based Resistive Switching Memories-Nanoionic Mechanisms, Prospects, and Challenges.Advanced Materials (Deer- field Beach, Fla.) 2009, 21, 2632–2663

    2009

  21. [29]

    Understanding the role of the Ti metal electrode on the forming of HfO2-based RRAMs

    Padovani, A.; Larcher, L.; Padovani, P.; Cagli, C.; De Salvo, B. Understanding the role of the Ti metal electrode on the forming of HfO2-based RRAMs. 2012 4th IEEE International Memory Workshop. 2012; pp 1–4

    2012

  22. [30]

    S.; Schroeder, H.; Breuer, U.; Waser, R

    Jeong, D. S.; Schroeder, H.; Breuer, U.; Waser, R. Characteristic electroforming behavior in Pt/TiO2/Pt resistive switching cells depending on atmosphere. Journal of applied physics 2008, 104 . 33

    2008

  23. [31]

    M.; Ramer, G.; Li, D.; Hoskins, B

    Goodwill, J. M.; Ramer, G.; Li, D.; Hoskins, B. D.; Pavlidis, G.; McClelland, J. J.; Centrone, A.; Bain, J. A.; Skowronski, M. Spontaneous current constriction in threshold switching devices. Nature communications 2019, 10, 1628

    2019

  24. [32]

    Russo, U.; Ielmini, D.; Cagli, C.; Lacaita, A. L. Self-accelerated thermal dissolution model for reset programming in unipolar resistive-switching memory (RRAM) devices. IEEE Transactions on Electron Devices 2009, 56, 193–200

    2009

  25. [33]

    Thermochemical resistive switching: materials, mechanisms, and scaling projections

    Ielmini, D.; Bruchhaus, R.; Waser, R. Thermochemical resistive switching: materials, mechanisms, and scaling projections. Phase Transitions 2011, 84, 570–602

    2011

  26. [34]

    Direct observation of conducting filaments on resistive switching of NiO thin films

    Son, J.; Shin, Y.-H. Direct observation of conducting filaments on resistive switching of NiO thin films. Applied Physics Letters 2008, 92

    2008

  27. [35]

    M.; Sharma, A

    Goodwill, J. M.; Sharma, A. A.; Li, D.; Bain, J. A.; Skowronski, M. Electro-thermal model of threshold switching in TaO x-based devices.ACS applied materials & interfaces 2017, 9, 11704–11710

    2017

  28. [36]

    A comparative study on the diffusion behaviors of metal and oxygen ions in metal-oxide-based resistance switches via ab initio molecular dynam- ics simulations

    Xiao, B.; Yu, X.; Watanabe, S. A comparative study on the diffusion behaviors of metal and oxygen ions in metal-oxide-based resistance switches via ab initio molecular dynam- ics simulations. ACS Applied Electronic Materials 2019, 1, 585–594

    2019

  29. [37]

    Modified charge transfer–embedded atom method potential for metal/metal oxide systems

    Zhou, X.; Wadley, H.; Filhol, J.-S.; Neurock, M. Modified charge transfer–embedded atom method potential for metal/metal oxide systems. Physical Review B 2004, 69, 035402

    2004

  30. [38]

    W.; Wadley, H

    Zhou, X. W.; Wadley, H. N. A charge transfer ionic–embedded atom method potential for the O–Al–Ni–Co–Fesystem. Journal of Physics: Condensed Matter 2005, 17, 3619

    2005

  31. [39]

    Voltage equilibration for reactive atomistic simulations of electrochemical processes

    Onofrio, N.; Strachan, A. Voltage equilibration for reactive atomistic simulations of electrochemical processes. The Journal of chemical physics 2015, 143 . 34

    2015

  32. [40]

    Fast parallel algorithms for short-range molecular dynamics

    Plimpton, S. Fast parallel algorithms for short-range molecular dynamics. Journal of computational physics 1995, 117, 1–19

    1995

  33. [41]

    J.; Bragaglia, V

    Stecconi, T.; Popoff, Y.; Guido, R.; Falcone, D.; Halter, M.; Sousa, M.; Horst, F.; La Porta, A.; Offrein, B. J.; Bragaglia, V. Role of Conductive-Metal-Oxide to HfO x, In- terfacial Layer on the Switching Properties of Bilayer TaO x/HfO x ReRAM. ESSDERC 2022-IEEE 52nd Europea...

    2022

  34. [42]

    V.; Pedersen, C

    Li, Y.; Sanna, S.; Norrman, K.; Christensen, D. V.; Pedersen, C. S.; Lastra, J. M. G.; Traulsen, M. L.; Esposito, V.; Pryds, N. Tuning the stoichiometry and electrical prop- erties of tantalum oxide thin films. Applied Surface Science 2019, 470, 1071–1074

    2019

  35. [43]

    M.; Hoffmann-Eifert, S.; Waser, R.; Menzel, S.; Wouters, D

    Sch¨ onhals, A.; Rosario, C. M.; Hoffmann-Eifert, S.; Waser, R.; Menzel, S.; Wouters, D. J. Role of the electrode material on the RESET limitation in oxide ReRAM devices. Ad- vanced electronic materials 2018, 4, 1700243

    2018

  36. [44]

    A unified formulation of the constant temperature molecular dynamics methods

    Nos´ e, S. A unified formulation of the constant temperature molecular dynamics methods. The Journal of chemical physics 1984, 81, 511–519

    1984

  37. [45]

    Hoover, W. G. Canonical dynamics: Equilibrium phase-space distributions. Physical review A 1985, 31, 1695

    1985

  38. [46]

    Developing a variable charge potential for Hf/Nb/Ta/Ti/Zr/O system via machine learning global optimization

    Wu, Y.; Yu, W.; Shen, S. Developing a variable charge potential for Hf/Nb/Ta/Ti/Zr/O system via machine learning global optimization. Materials & Design 2023, 230, 111999

    2023

  39. [47]

    J.; Aziz, A.; Rose, G

    Zeumault, A.; Alam, S.; Wood, Z.; Weiss, R. J.; Aziz, A.; Rose, G. S. TCAD modeling of resistive-switching of HfO2 memristors: Efficient device-circuit co-design for neuro- morphic systems. Frontiers in Nanotechnology 2021, 3, 734121

    2021

  40. [48]

    K.; Yildiz, B.; Waser, R.; others Nanoscale cation motion in TaO x, HfO x and TiO x memristive systems

    Wedig, A.; Luebben, M.; Cho, D.-Y.; Moors, M.; Skaja, K.; Rana, V.; Hasegawa, T.; 35 Adepalli, K. K.; Yildiz, B.; Waser, R.; others Nanoscale cation motion in TaO x, HfO x and TiO x memristive systems. Nature nanotechnology 2016, 11, 67–74

    2016

  41. [49]

    W.; Cottom, J.; Larcher, L.; Shluger, A

    Strand, J. W.; Cottom, J.; Larcher, L.; Shluger, A. L. Effect of electric field on defect generation and migration in HfO 2. Physical Review B 2020, 102, 014106

    2020

  42. [50]

    K.; Goddard III, W

    Rappe, A. K.; Goddard III, W. A. Charge equilibration for molecular dynamics simula- tions. The Journal of Physical Chemistry 1991, 95, 3358–3363

    1991

  43. [51]

    A modified embedded-atom method in- teratomic potential for ionic systems: 2 nnmeam+ qeq

    Lee, E.; Lee, K.-R.; Baskes, M.; Lee, B.-J. A modified embedded-atom method in- teratomic potential for ionic systems: 2 nnmeam+ qeq. Physical Review B 2016, 93, 144110

    2016

  44. [52]

    Electrostatic potentials for metal-oxide surfaces and interfaces

    Streitz, F.; Mintmire, J. Electrostatic potentials for metal-oxide surfaces and interfaces. Physical Review B 1994, 50, 11996

    1994

  45. [53]

    G.; Gray, S

    Sasikumar, K.; Narayanan, B.; Cherukara, M.; Kinaci, A.; Sen, F. G.; Gray, S. K.; Chan, M. K.; Sankaranarayanan, S. K. Evolutionary optimization of a charge transfer ionic potential model for Ta/Ta-Oxide heterointerfaces. Chemistry of Materials 2017, 29, 3603–3614

    2017

  46. [54]

    Sasikumar, K.; Chan, H.; Narayanan, B.; Sankaranarayanan, S. K. Machine learning ap- plied to a variable charge atomistic model for Cu/Hf binary alloy oxide heterostructures. Chemistry of Materials 2019, 31, 3089–3102

    2019

  47. [55]

    J.; Ghosh, S

    Mortier, W. J.; Ghosh, S. K.; Shankar, S. Electronegativity-equalization method for the calculation of atomic charges in molecules. Journal of the American Chemical Society 1986, 108, 4315–4320

    1986

  48. [56]

    Atomic origin of ultrafast resistance switching in nanoscale electrometallization cells

    Onofrio, N.; Guzman, D.; Strachan, A. Atomic origin of ultrafast resistance switching in nanoscale electrometallization cells. Nature materials 2015, 14, 440–446. 36

    2015

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.