Pith. sign in

REVIEW 4 major objections 5 minor 28 references

Electronic properties and transport in metal/2D material/metal vertical junctions

T0 review · 4 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read In Ni-contacted vertical junctions, a monolayer of an insulating or semiconducting 2D material becomes metallic and conducts.

desk verdict Useful comparative DFT/NEGF dataset on Ni/2D/Ni junctions, but the missing spin treatment of Ni is a real gap; qualitative message holds, quantitative rankings do not. read the letter →

arxiv 2502.03318 v1 pith:PNGWLO2Q submitted 2025-02-05 cond-mat.mes-hall

Gaëlle Bigeard , Zineb Kerrami , François Triozon , Alessandro Cresti This is my paper · ORCID
classification cond-mat.mes-hall
keywords verticaltransportmetal-2Dinterfaceorbitalhybridizationdensityfunctionaltheorynon-equilibriumGreen'sfunctionhBNMoSe2atomristor
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

This paper uses density functional theory and non-equilibrium Green's function transport simulations to establish that in nickel-contacted vertical junctions, the electronic structure of the 2D layer is set by the contact, not by the isolated material. A single monolayer of insulating hBN or semiconducting 2H-MoSe2 becomes metallic through orbital hybridization with the nickel electrodes, with new states appearing inside the gap, so the monolayer conducts. In trilayers, the outer layers are metallized and act as passivating contacts while the middle layer keeps its intrinsic band structure and becomes the tunneling barrier, which explains why conductivity drops sharply with thickness. The result matters for atomristors and vertical beyond-CMOS devices because it says the high-resistance state is controlled by the inner-layer barrier and by interface hybridization rather than by the intrinsic gap of the 2D material alone.

What carries the argument

The mechanism is orbital hybridization at the metal/2D interface: nickel d states hybridize with the p states of carbon, nitrogen, boron, or selenium, producing gap states and metallizing the first 2D layer, which then acts as a passivating contact. The counteracting element is the intrinsic middle layer in multilayers, which keeps its isolated band gap and imposes a tunneling barrier that controls the high-resistance state. The authors quantify both through layer-resolved densities of states, planar-averaged Hartree potentials, and conductance normalized by device area.

What would settle it

Measure the zero-bias conductance of a clean Ni|hBN|Ni junction with one, two, and three hBN layers and compare with the simulated values: the monolayer should be conductive and the drop from monolayer to trilayer should be roughly three orders of magnitude; alternatively, repeat the DFT with vdW-corrected functionals and check whether the interfacial distance and barrier heights change enough to reverse the conductivity ranking.

Watch

Extended reading notes

Core claim

The central claim is that out-of-plane transport through metal/2D material/metal junctions is governed by the metal-induced hybridization at the interfaces and by the intrinsic electronic structure of the inner layers, not by the band structure of the isolated monolayer. For monolayers of hBN and 2H-MoSe2, the paper finds that contact with Ni(111) creates hybridized states inside the gap, making the layer metallic and conductive; for graphene the interface coupling further lowers the tunneling barrier and gives the highest conductivity, while hBN has the highest barrier and the lowest conductivity. For trilayers, the layers in direct contact with the electrodes behave like the monolayer case while the middle layer remains essentially intrinsic, so the conductivity falls by orders of magnitude as thickness increases from one to four layers. This is why the paper concludes that a single hBN monolayer is conductive inside non-passivated metallic contacts despite being an insulator when isolated.

Load-bearing premise

The computed interfacial distances, hybridization strengths, and barrier heights come from DFT with the PBE functional and no van der Waals corrections, so if those distances or gaps are off, the conductivity ordering could change.

Editorial extensions

If this is right

  • A monolayer of an insulating or semiconducting 2D material between non-passivated metal contacts is not a tunneling barrier; it conducts.
  • The high-resistance state of vertical devices is set by the number of 2D layers, through the intrinsic middle layers, so thickness can tune ON/OFF behavior.
  • Stacking orientation matters most when the supercell is small (graphene, hBN): atomic alignment can change interaction energy and interface distance by up to about 0.8 Å.
  • For semiconducting materials, conductivity should go to zero with increasing thickness, while for metallic 2D materials it saturates to a finite value set by interlayer coupling.
  • The simulated trilayer conductivities exceed experimental values by orders of magnitude, implying that real devices have additional resistances such as multilayer regions, residues, or oxide.

Reading between the lines

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

  • If the hybridization picture is right, the contact metal choice is as important as the 2D material: metals with lower reactivity or weaker d-p coupling should preserve the intrinsic gap better and restore a true tunnel barrier at monolayer thickness.
  • The strong stacking sensitivity of graphene and hBN suggests that rotational alignment between the two electrodes could be a knob to engineer barrier height, testable by twisting the contacts relative to each other.
  • The discrepancy with experiments points to a testable hypothesis: intentional contamination or oxidation at the interface should increase the effective barrier and recover the experimentally low conductance, which could be checked by controlled exposure studies.
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 / 5 minor

Summary. This manuscript reports first-principles DFT/NEGF simulations of vertical Ni(111)/2D-material/Ni(111) junctions for graphene, hBN, 1T- and 2H-MoSe2, for monolayer and multilayer (2-4 layers) thicknesses, considering two stacking modes. The central claim is that the 2D layers in direct contact with Ni electrodes become metallic through orbital hybridization with Ni d states, so an insulating monolayer of hBN or semiconducting monolayer of 2H-MoSe2 conducts, while in trilayers the middle layer retains its intrinsic electronic structure and acts as the tunneling barrier, causing a sharp decrease of conductivity with layer number. The authors also provide geometric parameters (distances, interaction energies), electronic structure (DoS, Hartree potentials), and computed 2D conductivities.

Significance. If the central claim is robust, the work has practical implications for understanding the high-resistance state of atomristors and for designing vertical 2D-material devices, since it implies that interface hybridization, not the isolated material's band gap, controls out-of-plane transport. The paper is transparent about its approximations: it explicitly lists the pseudopotentials, basis sets, K-grids, and the exclusion of vdW and SOC, and it acknowledges that the computed three-layer conductivities exceed experimental values by orders of magnitude. It also provides a systematic comparison across four materials and several thicknesses, which is valuable as a qualitative benchmark.

major comments (4)
  1. [Section 2.1 (Methods)] The manuscript never states whether spin polarization is included in the DFT and NEGF calculations. Ni(111) is strongly ferromagnetic, with a magnetic moment of approximately 0.6 μB per atom, and the exchange-split Ni d bands are precisely the states that hybridize with the C, B, N, and Se orbitals to create the interfacial gap states (Figs. 2–5) and set the barrier heights (Fig. 4 and Table 1). If the simulations are spin-unpolarized, the central claim that monolayer hBN and 2H-MoSe2 are metallized by hybridization, and the resulting conductivity rankings in Tables 2 and 3, would rest on a nonmagnetic Ni electronic structure. Please state the spin treatment explicitly and provide at least one spin-polarized test (e.g., for the Ni|hBN|Ni junction) showing that the gap states and barrier heights are qualitatively unchanged.
  2. [Section 2.1 and Table 1] The exclusion of van der Waals corrections is justified by the assertion that strong coupling dominates, but no supporting calculation is provided. The electrode–2D material distances in Table 1 are very short (e.g., 2.04–2.08 Å for the strongly coupled stacking of graphene and hBN), and the tunneling barriers in Fig. 4 and the conductivities in Table 2 depend sensitively on these distances. A vdW-corrected calculation for at least one system (e.g., Ni|Gr|Ni or Ni|hBN|Ni) is needed to establish that the hybridization-driven metallization and the barrier heights are robust against the choice of exchange-correlation functional; otherwise the agreement with the qualitative picture may be coincidental.
  3. [Section 2.1 and Tables 2–3] The paper reports convergence tolerances for the self-consistent DFT loop (10 mÅ, 5×10−3 e/Å3, 100 meV) but gives no convergence tests for the transport quantities themselves, such as the K-grid for TRANSIESTA/TBtrans, the number of bias points used for the spline interpolation, or the basis-set size. The monolayer conductivities in Table 2 span only a factor of 2.5, so without error estimates it is unclear whether the ordering Gr > 1T-MoSe2 > 2H-MoSe2 > hBN is numerically significant. Please provide at least one convergence check (e.g., doubling the TRANSIESTA K-grid and/or basis size for one device) and report the resulting change in conductivity.
  4. [Section 3.1 and Fig. 2] The text in Section 3.1 states that the DoS of the Ni atom plane closest to the 2D material is also provided, with coinciding peaks proving hybridization, but Fig. 2 shows only the total DoS of the 2D material (and the free-standing reference). The Ni-projected DoS does not appear anywhere in the paper. Either add the Ni-plane DoS to Fig. 2 (or Fig. 3) or remove the sentence, because the claim of hybridization-driven metallization is currently supported only by the appearance of gap states in the 2D material, not by direct evidence of Ni–2D orbital overlap.
minor comments (5)
  1. [Table 3] The rows for 1T-MoSe2 are missing for 2L and 4L; please state in the text whether these systems were not calculated, and why.
  2. [Fig. 5] The statement that the middle layer 'preserves its intrinsic electronic structure' would be better supported by overlaying the free-standing DoS in the same panel; currently the comparison is only implicit.
  3. [References] The SISL Python library is used for post-processing but is not cited; please add the appropriate reference (e.g., N. Papior, sisl, Zenodo, 2020).
  4. [Section 3.2] The comparison with experiment is only qualitative ('still much higher than experimentally observed [9]'). Please report the experimental conductivity values from ref. [9] so the reader can gauge the discrepancy quantitatively.
  5. [Abstract] The phrase 'two phases of molybdenum diselenide' could be made clearer by naming 1T and 2H explicitly in the abstract itself.

Circularity Check

0 steps flagged · score 0.0 of 10

Direct DFT/NEGF simulation with no fitted target; no circular step identified.

full rationale

The paper's derivation chain is self-contained and non-circular. Structural relaxation, density of states, potential barriers, and conductivities are all obtained from first-principles DFT (SIESTA/PBE) and NEGF (TranSIESTA/TBtrans), with simulation parameters fully specified in the Appendix. The basis set is benchmarked against plane-wave Abinit band structures within a stated energy window, which is an external reference rather than a fitted target. No parameter appearing in the conductivity calculation is adjusted to reproduce the reported conductivities or the central 'interface metallization' claim. The stacking modes are selected using interaction energies reported in Table 1, not by the resulting conductance, so the structural input is independent of the transport output. The monolayer metallization and the trilayer middle-layer behavior are read off from projected DoS and Hartree-potential profiles, and the conductivity rankings are then interpreted in terms of these computed quantities; this is an interpretive correlation, not a circular definition. The only self-citation is reference [7] in the introduction, used as general context for beyond-CMOS interest, and it is not load-bearing for any result. The paper explicitly acknowledges that computed trilayer conductivities are orders of magnitude above experiment and attributes this to neglected experimental factors; that is an honest limitation, not an input-to-output reduction. Concerns about PBE accuracy, omitted vdW corrections, or unspecified spin polarization would be correctness or robustness risks, but they do not make any step circular under the defined criteria.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The calculation is a direct DFT/NEGF simulation of known materials; no quantity is fitted to the reported conductivities. The unverified inputs are methodological: PBE exchange-correlation, exclusion of vdW and SOC, the PAO basis convergence protocol, and the strain tolerance. These are reasonable but untested assumptions, and they affect quantitative reliability more than circularity.

assumptions (5)
  • domain assumption The PBE functional without vdW corrections yields reliable interfacial distances and hybridization for Ni/2D contacts.
    Section 2.1 states vdW corrections are excluded because strong coupling dominates; this assumption directly affects the interfacial distances and barriers in Tables 1 and 2 and Fig. 4.
  • domain assumption The PAO basis optimized to reproduce plane-wave band structure within +-100 meV in a +-3 eV window is converged for transport quantities.
    Section 2.2 describes basis optimization only against band structure, not against conductance; no basis-size convergence check for the NEGF conductivities is shown.
  • domain assumption Artificial strain up to 1.4% with band deformation within +-100 meV (Ni) and +-50 meV (2D) does not qualitatively alter the transport trends.
    Section 2.3 sets maximum allowed strain; strain values differ across devices (Table A3), so cross-material comparisons assume this strain does not change the ranking.
  • domain assumption Spin-orbit coupling does not affect vertical transport in these systems.
    Section 2.1 excludes SOC for computational cost, asserting it does not change conclusions; this is relevant for MoSe2 valence bands.
  • domain assumption PBE's band gap underestimation does not change the qualitative picture of metal-induced gap states and tunneling barriers.
    PBE is used throughout; for hBN and 2H-MoSe2 the computed gap and gap states are functional-dependent, though qualitative hybridization effects are likely robust.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Electronic properties and transport in metal/2D material/metal vertical junctions." pith.science (2026). https://pith.science/paper/PNGWLO2Q

@misc{pith2026250203318,
  author       = {Pith},
  title        = {Pith review of: Electronic properties and transport in metal/2D material/metal vertical junctions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PNGWLO2Q}},
  note         = {Machine review of arXiv:2502.03318}
}
read the original abstract

We simulate the electronic and transport properties of metal/two-dimensional material/metal vertical heterostructures, with a focus on graphene, hexagonal boron nitride and two phases of molybdenum diselenide. Using density functional theory and non-equilibrium Green's function, we assess how stacking configurations and material thickness impact important properties, such as density of states, potential barriers and conductivity. For monolayers, strong orbital hybridization with the metallic electrodes significantly alters the electronic characteristics, with the formation of states within the gap of the semiconducting 2D materials. Trilayers reveal the critical role of interlayer coupling, where the middle layer retains its intrinsic properties, thus influencing the overall conductivity. Our findings highlight the potential for customized multilayer designs to optimize electronic device performance based on two-dimensional materials.

Figures

Figures reproduced from arXiv: 2502.03318 by the authors.

Figure 1
Figure 1. Lateral and top views of the structural configurations of (a) Ni [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Total DoS per unit cell for monolayer (a) graphene, (b) hBN, (c) 2H-MoSe2 and (d) 1T-MoSe2 versus energy (eV). Fermi level at E = 0 is indicated by a vertical dotted line. Comparison between the 2D material relaxed in Ni electrodes with CB (or BA for MoSe2) stacking (red dashed line), AA stacking (blue dashed line), and relaxed free-standing 2D material (black solid line). to the 2D material is also provided, with c… view at source ↗
Figure 3
Figure 3. Projected DoS per atom on the indicated valence orbitals for monolayer (a) graphene, (b) hBN, (c) 2H-MoSe2 and (d) 1T-MoSe2 versus energy and for different indicated stacking modes. Fermi level at E = 0 is indicated by a vertical dotted line. metallic as a result of the strongly hybridized states at the interface with the contacts. From [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Planar average of the Hartree potential (left) and charge density (right) for [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: DoS on the layers in contact with the top/bottom electrodes and the middle [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Planar average of the Hartree potential (left) and charge density (right) for [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    IEEE International Roadmap for Devices and Systems 2023 Beyond CMOS and Emerging Materials Integration

    work page 2023

  2. [2]

    Avci U E, Morris D H and Young I A 2015 IEEE Journal of the Electron Devices Society3 88–95

    work page 2015

  3. [3]

    Robinson J A 2018 APL Materials 6 058202

    work page 2018

  4. [4]

    Briggs N, Subramanian S, Lin Z, Li X, Zhang X, Zhang K, Xiao K, Geohegan D, Wallace R, Chen L Q, Terrones M, Ebrahimi A, Das S, Redwing J, Hinkle C, Momeni K, van Duin A, Crespi V, Kar S and Robinson J A 2019 2D Materials 6 022001

    work page 2019

  5. [5]

    Dragoman M and Dragoman D 2021 Atomic-Scale Electronics Beyond CMOS (Springer International Publishing) ISBN 9783030605636

    work page 2021

  6. [6]

    Hao Z, Yan Y, Shi Y and Li Y 2022 The Journal of Physical Chemistry Letters13 1914–1924

    work page 2022

  7. [7]

    Cresti A 2023 Beyond-CMOS: State of the Art and Trends(Wiley) ISBN 9781394228713

    work page 2023

  8. [8]

    Lemme M C, Akinwande D, Huyghebaert C and Stampfer C 2022 Nature Communications 13 1392

    work page 2022

Show all 28 references
  1. [9]

    Ge R, Wu X, Kim M, Shi J, Sonde S, Tao L, Zhang Y, Lee J C and Akinwande D 2017 Nano Letters 18 434–441

    2017

  2. [10]

    Kim M, Ducournau G, Skrzypczak S, Yang S J, Szriftgiser P, Wainstein N, Stern K, Happy H, Yalon E, Pallecchi E and Akinwande D 2022 Nature Electronics5 367–373

    2022

  3. [11]

    Hus S M, Ge R, Chen P A, Liang L, Donnelly G E, Ko W, Huang F, Chiang M H, Li A P and Akinwande D 2020 Nature Nanotechnology16 58–62

    2020

  4. [12]

    Shah A A, Kumar J, Dar A B and Shrivastava M 2023 IEEE Transactions on Electron Devices 70 6622–6629

    2023

  5. [13]

    Shah A A, Dar A B and Shrivastava M 2024 npj 2D Materials and Applications8 80

    2024

  6. [14]

    Mitra S and Mahapatra S 2024 npj 2D Materials and Applications8 26

    2024

  7. [15]

    Sangwan V K, Jariwala D, Kim I S, Chen K S, Marks T J, Lauhon L J and Hersam M C 2015 Nature Nanotech 10 403–406

    2015

  8. [16]

    Chen H, Kang Y, Pu D, Tian M, Wan N, Xu Y, Yu B, Jie W and Zhao Y 2023 Nanoscale 15 4309–4316

    2023

  9. [17]

    Sangwan V K, Lee H S, Bergeron H, Balla I, Beck M E, Chen K S and Hersam M C 2018 Nature 554 500–504

    2018

  10. [18]

    Wang M, Cai S, Pan C, Wang C, Lian X, Zhuo Y, Xu K, Cao T, Pan X, Wang B, Liang S J, Yang J J, Wang P and Miao F 2018 Nature Electronics1 130–136

    2018

  11. [19]

    Jeong H Y, Kim J Y, Kim J W, Hwang J O, Kim J E, Lee J Y, Yoon T H, Cho B J, Kim S O, Ruoff R S and Choi S Y 2010 Nano Letters 10 4381–4386

    2010

  12. [20]

    Krishnaprasad A, Dev D, Han S S, Shen Y, Chung H S, Bae T S, Yoo C, Jung Y, Lanza M and Roy T 2022 ACS Nano 16 2866–2876

    2022

  13. [21]

    Zhang F, Zhang H, Krylyuk S, Milligan C A, Zhu Y, Zemlyanov D Y, Bendersky L A, Burton 16 B P, Davydov A V and Appenzeller J 2019 Nature Matererials 18 55–61

    2019

  14. [22]

    Garc ´ ıa A, Papior N, Akhtar A, Artacho E, Blum V, Bosoni E, Brandimarte P, Brandbyge M, Cerd´ a J I, Corsetti F, Cuadrado R, Dikan V, Ferrer J, Gale J, Garc ´ ıa-Fern´ andez P, Garc ´ ıa-Su´ arez V M, Garc ´ ıa S, Huhs G, Illera S, Koryt´ ar R, Koval P, Lebedeva I, Lin L, L´...

    2020

  15. [23]

    Papior N, Lorente N, Frederiksen T, Garc ´ ıa A and Brandbyge M 2017 Computer Physics Communications 212 8–24

    2017

  16. [24]

    Perdew J P, Burke K and Ernzerhof M 1996 Physical Review Letters77 3865–3868

    1996

  17. [25]

    Hamann D R 2013 Physical Review B88 085117

    2013

  18. [26]

    van Setten M J, Giantomassi M, Bousquet E, Verstraete M J, Hamann D R, Gonze X and Rignanese G M 2018 Computer Physics Communications226 39–54

    2018

  19. [27]

    Papior N, Brandbyge M and Calogero G 2018 Journal of Physics: Condensed Matter30

    2018

  20. [28]

    Gonze X, Amadon B, Antonius G, Arnardi F, Baguet L, Beuken J M, Bieder J, Bottin F, Bouchet J, Bousquet E, Brouwer N, Bruneval F, Brunin G, Cavignac T, Charraud J B, Chen W, Cˆ ot´ e M, Cottenier S, Denier J, Geneste G, Ghosez P, Giantomassi M, Gillet Y, Gingras O, Hamann D R,...

    2020

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.