Pith. sign in

REVIEW 4 major objections 5 minor 62 references

Biaxial strain alone stabilizes a buckled silicon kagome sheet while narrowing its near-Fermi flat band and raising its crystalline-to-amorphous transition temperature.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-14 03:09 UTC pith:PWMRXXVZ

load-bearing objection Solid multi-scale proposal of an elemental buckled Si kagome with real DFT/AIMD strain control of a near-Fermi band; the ~600 K classical map is the softest piece, not the whole story. the 4 major comments →

arxiv 2607.11790 v1 pith:PWMRXXVZ submitted 2026-07-13 cond-mat.mtrl-sci cond-mat.mes-hall

Strain-controlled crystalline--amorphous transition and flat-band tuning in buckled silicon kagome

Chenhaoyue Wang , Amartya S. Banerjee This is my paper
classification cond-mat.mtrl-sci cond-mat.mes-hall
keywords silicon kagome latticeflat bandsbiaxial straincrystalline-amorphous transitiontwo-dimensional materialsbuckled siliconmolecular dynamicskagome-derived bands
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper proposes an unfunctionalized silicon monolayer built from linked Si3 triangles and dodecagonal pores—a buckled silicon kagome lattice (SiKL). The planar parent hosts a dispersionless band near the Fermi level but is unstable to out-of-plane soft modes; following those modes yields two nearly degenerate buckled forms, of which the high-buckling structure keeps a partially flat kagome-derived band. Biaxial tension is shown to do three jobs at once: harden the softest phonon, keep large ab initio cells ordered at 315 K while the unstrained sheet disorders, and cut the reported bandwidth from 0.86 to 0.47 eV at 10% strain with the density-of-states peak moving toward the Fermi level. Longer classical molecular-dynamics runs of large sheets further map a strain-controlled crystalline–amorphous transition with a local-bonding crossover near 2% strain that switches gradual two-stage disordering into an abrupt collapse and lifts the finite-time transition temperature to about 600 K at 10% strain. A sympathetic reader cares because the design stays elemental silicon and uses one mechanical knob to couple thermal metastability, bond rearrangement, and near-Fermi flat-band tuning without passivation or hybrid-lattice redesign, offering a candidate platform for strain-controlled correlation physics and a possible epitaxial route via tensile templates such as Ag(111).

Core claim

An unfunctionalized six-atom buckled silicon kagome lattice retains a kagome-derived partially flat band near the Fermi level, and biaxial tensile strain alone narrows that band (0.86 to 0.47 eV at 10% strain), hardens the dominant soft phonon, and converts a thermally labile sheet into a finite-temperature metastable ordered network whose classical crystalline–amorphous transition temperature reaches approximately 600 K at 10% strain, with a local-bonding crossover near 2% that changes the disordering pathway from gradual two-stage collapse to abrupt first-order-like failure.

What carries the argument

Buckled silicon kagome lattice (SiKL): a six-atom monolayer of bond-linked Si3 triangles and dodecagonal pores whose out-of-plane buckling lets biaxial tension simultaneously harden soft modes, reorganize intra- versus inter-triangle bond populations, and reduce the kagome-derived near-Fermi bandwidth.

Load-bearing premise

The long-time crystalline-to-amorphous map and the roughly 600 K claim rest on an empirical silicon potential never trained on this high-energy linked-triangle allotrope, so the barrier ranking may not match first-principles free energies.

What would settle it

Long first-principles free-energy sampling or large-cell AIMD of strained SiKL that either disorders well below the classical transition temperatures or shows no local-bonding crossover near 2% strain would overturn the claimed strain-controlled pathway and temperature scale.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Strain alone can serve as the control parameter for both structural persistence and near-Fermi bandwidth in elemental SiKL, without hydrogenation or hybrid-lattice redesign.
  • Epitaxial mismatch on a template such as Ag(111) may impose comparable tension, preserve crystalline order above room temperature, and further narrow the extracted SiKL band.
  • The strain-tuned reduction of bandwidth and rise of near-Fermi density of states make spin-polarized and many-body searches for magnetism, superconductivity, or topological phases natural next calculations at appropriate filling.
  • The disordering pathway itself is designable: below about 2% strain collapse is gradual and two-stage; above it the transition becomes abrupt and first-order-like within the simulated trajectories.

Where Pith is reading between the lines

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

  • If the classical barrier ranking is qualitatively correct, modest tensile templates could kinetically trap SiKL long enough for ambient spectroscopic searches of the narrow band.
  • The same soft-mode buckling plus tensile-strain recipe may transfer to the heavier planar group-14 linked-triangle analogues the paper already relaxes.
  • For growth, adsorption-site registry that preferentially nucleates linked triangles over silicene or compact clusters will matter as much as pure lattice mismatch.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The manuscript proposes an unfunctionalized buckled silicon kagome lattice (SiKL) obtained by following soft zone-centre phonons of a planar linked-triangle parent. Using DFT/DFPT, AIMD, and classical MD, it argues that biaxial tension simultaneously narrows a kagome-derived near-Fermi band (0.86 o0.47 eV at 10% strain), hardens the dominant soft phonon, and converts a labile sheet into a finite-temperature metastable ordered network. Classical 50 ns trajectories of 36 imes36 sheets are used to map a strain-controlled crystalline–amorphous transition with a local-bonding crossover near 2% that changes the disordering pathway, with TCA reaching ~600 K at 10% strain; an exploratory Ag(111) model is offered as a possible epitaxial strain template.

Significance. If the coupled structural–electronic picture holds, SiKL would be a rare elemental 2D platform in which a single mechanical variable tunes near-Fermi bandwidth, soft-mode stability, and crystalline persistence without passivation or hybrid-lattice redesign—relevant to strain-controlled flat-band and correlation physics and to silicon epitaxy. Strengths include a transparent soft-mode search that yields two nearly degenerate buckled minima, consistent PBE/LDA planar electronic signatures, explicit phonon hardening under tension, a clear 6 imes6 AIMD contrast (disordered unstrained vs ordered strained at 315 K), and an operational, time-resolved definition of TCA rather than an unstated melting claim. The multi-scale hierarchy (DFPT → AIMD → classical MD → substrate sketch) is a genuine asset if the classical map is properly caveated.

major comments (4)
  1. [§2.3, §3.4] §2.3 and §3.4: The strain–temperature map, the ~600 K claim at 10% strain, the ε≈0.0196 bond-population crossover, and the two-stage vs first-order-like pathway distinction rest entirely on a Tersoff Si–Si potential and a 50 ns RDF-midpoint TCA. SiKL is a high-energy, soft linked-triangle allotrope (~1.14 eV/atom above diamond Si; Table 2) outside the training domain of standard Tersoff silicon. Without validation against AIMD barriers, DFT-relaxed amorphous competitors, or at least a second potential, absolute TCA values and the pathway switch cannot be treated as load-bearing. Either add such checks or reframe §3.4 as a model-dependent kinetic survey and soften the abstract/conclusions accordingly.
  2. [§3.2, Fig. 4] §3.2 and Fig. 4: Under 10% biaxial strain the softest mode only hardens from ~−1.98 to ~−0.64 THz; residual imaginary branches remain. The paper correctly notes anharmonic renormalization and finite-T metastability, but the central claim that strain “suppresses the dominant structural instability” still leans on a 10 ps 6×6 AIMD trajectory. Longer AIMD, larger cells, or anharmonic free-energy estimates are needed before asserting robust room-temperature metastability of freestanding strained SiKL.
  3. [§3.3, Fig. 5] §3.3 and Fig. 5: The strained bandwidth is 0.47 eV along the sampled path—narrowed but not flat. The text already notes that correlation physics depends on U/W rather than a semantic flatness threshold, yet the abstract and title still frame the result as “flat-band tuning.” Report full-BZ bandwidths, orbital character, and at least a rough interaction-scale estimate (or Hubbard U scan) so the reader can judge whether 0.47 eV is interaction-relevant, and align the abstract language with the actual dispersion.
  4. [§3.5, Fig. 12] §3.5 and Fig. 12: The Ag(111) proof-of-concept uses a classical EAM+LJ interface and reports a peeled-off SiKL band narrowed to ~0.2 eV without Si-projected unfolded spectral weight. Hybridization with metallic Ag can destroy or bury the near-Fermi Si feature. First-principles interface bands (or at least projected DOS/unfolding) and registry/adhesion checks are required before claiming that epitaxial mismatch “retains a narrow SiKL band.”
minor comments (5)
  1. [Fig. 1] Fig. 1 caption/layout shows duplicated “(a) (b)” labels; clean the panel labeling.
  2. [Table 2] Table 2 cohesive energies are quoted to five decimals while differences between HB and LB are 0.04 meV/atom—state numerical precision and whether zero-point energy was considered.
  3. [§2.1] §2.1: Specify which PAW/pseudopotential files (library versions) were used for LDA vs PBE for reproducibility.
  4. [Appendix A] Appendix A: Planar C/Ge/Sn/Pb analogues are interesting but undiscussed for dynamical stability; either add a brief phonon note or mark them clearly as electronic-structure comparisons only.
  5. [Abstract, §3.4] Throughout: “first-order-like” / “second-order-like” are kinetic descriptors; keep that qualification in the abstract as well as in §3.4.

Circularity Check

0 steps flagged

No circularity: SiKL results are forward DFT/DFPT/AIMD/classical-MD computations on a proposed structure, not identities forced by definition or self-citation.

full rationale

The paper proposes an unfunctionalized buckled silicon kagome lattice and reports its properties via standard first-principles and classical simulations. Soft-mode displacements of the planar parent are relaxed to obtain HB/LB structures; biaxial strain is then applied and bandwidths, phonons, AIMD trajectories, and classical 50 ns RDFs are computed forward. TCA is defined operationally as the midpoint of an error-function fit to a selected RDF peak at fixed observation time, and the text explicitly labels it a finite-time, model-dependent metric rather than an equilibrium melting temperature—so the definition is transparent bookkeeping, not a prediction smuggled from a fit to the same quantity. Self-citations ([11]–[18]) concern related nanotube/flat-band methodology and do not supply a uniqueness theorem or ansatz that forces the SiKL bandwidth, phonon hardening, or transition map. The Tersoff potential and finite-time TCA are external-model limitations (correctness risk), not circular reductions of outputs to inputs. No step reduces by construction to its own premises.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 1 invented entities

The central claim is a multi-scale computational materials proposal. It inherits standard DFT/MD machinery and literature potentials, invents the SiKL allotrope as the object of study, and leans on free modeling choices (Tersoff, 50 ns TCA definition, LJ Si–Ag, selected 10% strain window) that control the quantitative transition map and substrate outlook without independent experimental anchors.

free parameters (4)
  • Tersoff Si–Si potential parameters
    All 50 ns 36×36 transition maps and TCA(ε) values use literature Tersoff parameters not refitted to SiKL; absolute transition temperatures and pathway order are potential-dependent.
  • TCA observation time (50 ns RDF midpoint)
    Transition temperature is defined as the error-function midpoint of a selected RDF peak at a chosen finite time; the map shifts with observation time as the paper itself shows.
  • Si–Ag Lennard-Jones ε and σ (0.0774 eV, 3.4175 Å)
    Ag-supported classical order (~446 K) depends on Lorentz–Berthelot mixed LJ parameters from tabulated sources, not a reactive Si–Ag potential.
  • Representative strain window (especially 10% biaxial)
    Electronic and AIMD highlights are reported at a hand-selected 10% tension near the bandwidth minimum; the continuous strain scan exists but the headline numbers fix this operating point.
axioms (4)
  • domain assumption PBE (and LDA cross-check for planar parent) Kohn–Sham DFT adequately ranks bonding, soft modes, and near-Fermi band dispersion for low-coordinated Si allotropes.
    All structural relaxations, bands, DFPT phonons, and AIMD use this level; residual imaginary modes and bandwidths inherit DFT limitations.
  • domain assumption Finite-time classical MD order parameters (RDF peak intensity, potential-energy jumps) are a valid comparative metric for strain-dependent crystalline–amorphous kinetics of SiKL.
    §2.3 explicitly defines TCA as finite-time and model-dependent; the non-monotonic map and pathway claims rest on this proxy.
  • ad hoc to paper Soft-mode subspace search around the planar saddle (Γ1, Γ2, Γ3 displacements) captures the relevant buckled minima for this topology.
    §3.1 states the search is targeted, not a global allotrope search; other lower-energy Si networks with different connectivity are not ruled out as competitors under growth conditions.
  • ad hoc to paper Epitaxial mismatch on Ag(111) can impose ~10% tension without destroying the linked-triangle registry or the near-Fermi SiKL band via hybridization.
    §3.5 is labeled exploratory; adhesion, nucleation barriers, and Si-projected unfolded bands are deferred.
invented entities (1)
  • Buckled silicon kagome lattice (SiKL), HB/LB forms no independent evidence
    purpose: Provide an unfunctionalized elemental 2D linked-triangle Si network that retains a near-Fermi kagome-derived band and a strain-tunable soft-mode/amorphization response.
    The allotrope is proposed computationally; no experimental realization or independent observation is claimed. Cohesive energy places it high on the Si landscape, so existence is a prediction, not an established material.

pith-pipeline@v1.1.0-grok45 · 20182 in / 3854 out tokens · 37004 ms · 2026-07-14T03:09:04.136339+00:00 · methodology

0 comments
read the original abstract

Electronic flat bands in an elemental two-dimensional material provide an attractive setting for electron interactions competing with suppressed kinetic energy. Here we propose a buckled silicon kagome lattice (SiKL), an unfunctionalized six-atom monolayer of bond-linked Si$_3$ triangles and dodecagonal pores. Its planar parent hosts a dispersionless Kohn--Sham band near the Fermi level but is unstable to out-of-plane distortions. Following three soft zone-centre phonons and relaxing displaced structures yields two nearly degenerate buckled forms. The high-buckling form retains a partially flat kagome-derived band near the Fermi level. Biaxial tension controls lattice dynamics and electronic dispersion: at 10% strain, the bandwidth decreases significantly, the density-of-states peak approaches the Fermi level, and the softest phonon hardens. At 315 K, $6\times6$ ab initio MD shows the unstrained network disordering while the strained network remains ordered, indicating finite-temperature metastability. Fifty-nanosecond classical MD of $36\times36$ sheets reveals a strain-controlled crystalline--amorphous transition and local-bonding crossover near 2% strain. Low-strain trajectories show gradual, two-stage disordering; higher strains undergo an abrupt, first-order-like collapse, with the transition temperature reaching approximately 600 K at 10% strain. An exploratory Ag(111) substrate model suggests epitaxial mismatch could supply comparable tension, retain a narrow SiKL band, and preserve crystalline order above room temperature. Unlike passivated or hybrid-lattice silicon kagome proposals aimed mainly at conventional semiconductors, SiKL is elemental and uses strain alone to couple thermal metastability, bond rearrangement, and near-Fermi flat-band tuning. Buckled SiKL is a candidate platform for strain-controlled flat-band and electronic correlation physics.

Figures

Figures reproduced from arXiv: 2607.11790 by Amartya S. Banerjee, Chenhaoyue Wang.

Figure 1
Figure 1. Figure 1: (a) Top view of the planar SiKL parent lattice. The dashed polygon marks the primitive cell. (b) Kohn–Sham band structure (left) and density of states (right) calculated with LDA (red dashed curves) and PBE (black solid curves). Energies are referenced to the Fermi level. Jones form ϕ(r) = 4ϵ " σ r 12 −  σ r 6 # , (1) with a 10 ˚A cutoff. Lorentz–Berthelot mixing based on the tabulated parameters [49, … view at source ↗
Figure 2
Figure 2. Figure 2: (a) PBE phonon dispersion of planar SiKL. Negative plotted frequencies denote imaginary modes. (b) Representative relaxed-energy landscape for a fixed Γ3 = 0 slice as a function of the Γ1 and Γ2 mode amplitudes. The red point marks the lowest-energy structure within the displayed slice. configuration was fully relaxed. Figure 2b shows one representative two-dimensional slice of this energy landscape. This … view at source ↗
Figure 3
Figure 3. Figure 3: Relaxed buckled SiKL structures: (a) top view of a 3×3 supercell and side views of (b) the HB and (c) the LB primitive cells. The indicated buckling height is the vertical separation between neighbouring triangular units [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: PBE phonon dispersions and AIMD potential-energy histories for HB SiKL at 315 K: (a,b) unstrained and (c,d) under 10% biaxial tensile strain. The AIMD cells contain 6 × 6 primitive cells. Insets show configurations averaged over the final 2 ps of the respective trajectories. Strain strongly hardens the soft modes and preserves the ordered linked-triangle network over the 10 ps strained trajectory. (figure … view at source ↗
Figure 5
Figure 5. Figure 5: Kohn–Sham band structures and densities of states of HB SiKL (a) without strain and (b) under 10% biaxial tensile strain. The solid red line denotes the Fermi level in the DOS panel. The dashed red line and arrow indicate the energy interval used to quote the kagome-derived bandwidth. Biaxial strain reduces the displayed bandwidth from 0.86 to 0.47 eV. (a) (b) [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Folded 6 × 6-supercell band structures (left) and densities of states (right) calculated from coordinates averaged over the final 3 ps of the 315 K AIMD trajectories: (a) unstrained and (b) under 10% biaxial tension. The red line marks the Fermi level. The strained structure retains a pronounced near-Fermi DOS feature together with the ordered network. 3.4. Strain-controlled crystalline–amorphous transitio… view at source ↗
Figure 7
Figure 7. Figure 7: RDFs and final top-view configurations of unstrained 36 × 36 HB sheets after 50 ns at (a) 90 K and (b) 110 K. The sharp low-temperature peaks correspond to the two characteristic Si–Si bond populations of crystalline SiKL; their broadening accompanies loss of the periodic pore network. peaks broaden and lose intensity while the long-range pore pattern disappears, indicating transformation to an amorphous-l… view at source ↗
Figure 8
Figure 8. Figure 8: Representative 50 ns classical-MD trajectories for ε = 0.07 HB SiKL. (a) At T = 460 K, the crystalline SiKL network is retained and the potential energy remains stationary after initial equilibration. (b) At T = 530 K, the sheet undergoes an abrupt transformation to an amorphous-like network, accompanied by a sharp drop in potential energy. 80 90 100 110 120 Temperature (K) 10 15 20 25 30 Average g(r) t 10… view at source ↗
Figure 9
Figure 9. Figure 9: (a) Selected RDF-peak intensity of unstrained HB SiKL versus temperature at several observation times; dashed curves are error-function fits. (b) Strain– temperature crystalline–amorphous transition map constructed from the 50 ns fit midpoints. Purple and red shading denote crystalline and amorphous-like outcomes, respectively, within the finite-time classical model. model. Across this narrow range, weight… view at source ↗
Figure 10
Figure 10. Figure 10: Potential-energy histories over 50 ns at 120 K for (a) 1% and (b) 3% biaxial strain. Insets show RDFs at 50 ns for structures initially equilibrated at 80 K. (c) Ratio of the RDF-peak intensities near 1.8 and 2.1 ˚A versus strain at 70 K. The dashed vertical line marks the fitted inflection at ε = 0.0196, corresponding physically to a local-bonding crossover near 2%. trajectory before a final abrupt drop,… view at source ↗
Figure 11
Figure 11. Figure 11: Temperature dependence of the selected RDF-peak intensity at observation times from 10 to 50 ns for HB SiKL under (a) 1% and (b) 3% biaxial strain. The low-strain sheet loses order gradually over a broad interval, whereas the high-strain sheet shows a sharper transition. state, suppresses gradual disorder accumulation, and dramatically raises its finite-time transition temperature. This coupling among bon… view at source ↗
Figure 12
Figure 12. Figure 12: (a) Band structures of the free-standing unstrained HB-SiKL reference (black solid curves) and the free-standing peeled-off SiKL geometry obtained after relaxation on Ag (red dashed curves). The red curve includes the spontaneous strain generated during Ag-supported relaxation. (b) Selected crystalline RDF-peak amplitude versus temperature for the Ag-related classical-MD model at observation times from 10… view at source ↗

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

62 extracted references · 2 linked inside Pith

  1. [1]

    Leykam D, Andreanov A and Flach S 2018Advances in Physics: X31473052

  2. [2]

    Neupert T, Denner M M, Yin J X, Thomale R and Hasan M Z 2022Nature Physics18137–143

  3. [3]

    Po H C, Zou L, Vishwanath A and Senthil T 2018Physical Review X8031089

  4. [4]

    Yankowitz M, Chen S, Polshyn H, Zhang Y, Watanabe K, Taniguchi T, Graf D, Young A F and Dean C R 2019Science3631059–1064

  5. [5]

    Cao Y, Fatemi V, Demir A, Fang S, Tomarken S L, Luo J Y, Sanchez-Yamagishi J D, Watanabe K, Taniguchi T, Kaxiras E, Ashoori R C and Jarillo-Herrero P 2018Nature55680–84

  6. [6]

    M, Watanabe K, Taniguchi T, Rademaker L, Van Der Molen S J, Allan M P, Efetov D K and Baumberger F 2021Nature Physics17189–193

    Lisi S, Lu X, Benschop T, De Jong T A, Stepanov P, Duran J R, Margot F, Cucchi I, Cappelli E, Hunter A, Tamai A, Kandyba V, Giampietri A, Barinov A, Jobst J, Stalman V, Leeuwenhoek Strain-controlled transition and flat-band tuning20 (a) (c) (b) (d) Figure A2.Band structures (left) and densities of states (right) of planar (a) Si, (b) Ge, (c) Sn, and (d) P...

  7. [7]

    Mielke A 1991Journal of Physics A: Mathematical and General24L73–L77

  8. [8]

    Zhou M, Liu Z, Ming W, Wang Z and Liu F 2014Physical Review Letters113236802

  9. [9]

    Bergholtz E J and Liu Z 2013International Journal of Modern Physics B271330017

  10. [10]

    Yin J X, Lian B and Hasan M Z 2022Nature612647–657

  11. [11]

    Sharma S and Banerjee A S 2025arXiv preprint arXiv:2501.11783

    Pith/arXiv arXiv

  12. [12]

    Ghosh S, Banerjee A S and Suryanarayana P 2019Physical Review B100125143

  13. [13]

    Banerjee A S 2021Journal of the Mechanics and Physics of Solids154104515

  14. [14]

    Yu H M and Banerjee A S 2022Journal of Computational Physics456111023

  15. [15]

    Agarwal S and Banerjee A S 2024Journal of Computational Physics496112551

  16. [16]

    Pathrudkar S, Yu H M, Ghosh S and Banerjee A S 2022Physical Review B105195141

  17. [17]

    Yu H M, Sharma S, Agarwal S, Liebman O and Banerjee A S 2024RSC Advances14963–981

  18. [18]

    Sharma S, Wang C, Yu H M and Banerjee A S 2026The Journal of Physical Chemistry Letters 174352–4364 Strain-controlled transition and flat-band tuning21

  19. [19]

    Zhong C, Xie Y, Chen Y and Zhang S 2016Carbon9965–70

  20. [20]

    Sarikavak-Lisesivdin B, Lisesivdin S, Ozbay E and Jelezko F 2020Chemical physics letters760 138006

  21. [21]

    Chen Y, Xu S, Xie Y, Zhong C, Wu C and Zhang S B 2018Physical Review B98035135

  22. [22]

    Vogt P, De Padova P, Quaresima C, Avila J, Frantzeskakis E, Asensio M C, Resta A, Ealet B and Le Lay G 2012Physical Review Letters108155501

  23. [23]

    Leenaerts O, Schoeters B and Partoens B 2015Physical Review B91115202

  24. [24]

    Matusalem F, Marques M, Teles L K and Bechstedt F 2015Physical Review B92045436

  25. [25]

    Li Z, Zhuang J, Wang L, Feng H, Gao Q, Xu X, Hao W, Wang X, Zhang C, Wu K, Dou S X, Chen L, Hu Z and Du Y 2018Science Advances4eaau4511

  26. [26]

    Lee J H, Kim G W, Song I, Kim Y, Lee Y, Yoo S J, Cho D Y, Rhim J W, Jung J, Kim G and Kim C 2024ACS Nano1825535–25541

  27. [27]

    Sassa Y, Johansson F O, Lindblad A, Yazdi M G, Simonov K, Weissenrieder J, Muntwiler M, Iyikanat F, Sahin H, Angot T, Salomon E and Le Lay G 2020Applied Surface Science530 147195

  28. [28]

    Sato Y, Fukaya Y, Nakano A, Hoshi T, Lee C C, Yoshimi K, Ozaki T, Nakashima T, Ando Y, Aoyama H, Abukawa T, Tsujikawa Y, Horio M, Niibe M, Komori F and Matsuda I 2025Physical Review Materials9014002

  29. [29]

    Sang P, Wang Q, Wei W, Wang F, Li Y and Chen J 2021ACS Materials Letters31181–1188

  30. [30]

    Sang P, Wang Q, Wu J, Yi G, Li Y and Chen J 2023ACS Applied Nano Materials66849–6857

  31. [31]

    Zhang Y, Gao W, Shi X, Liu H and Gao J 2025Applied Physics Letters127022103

  32. [32]

    Huder L, Artaud A, Le Quang T, Trambly de Laissardi` ere G, Jansen A G M, Lapertot G, Chapelier C and Renard V T 2018Physical Review Letters120156405

  33. [33]

    Qiao J B, Yin L J and He L 2018Physical Review B98235402

  34. [34]

    Mao J, Milovanovi´ c S P, Andjelkovi´ c M, Lai X, Cao Y, Watanabe K, Taniguchi T, Covaci L, Peeters F M, Geim A K, Jiang Y and Andrei E Y 2020Nature584215–220

  35. [35]

    Yang L Z, Tong L H, Liao C S, Wu Q, Fu X, Zhou Y Y, Tian Y, Zhang L, Zhang L, Cai M Q, He L, Qin Z and Yin L J 2022Physical Review Materials6L041001

  36. [36]

    Kim H J, Kim M J, Lee J, Ok J M and Kang C J 2024Physical Review B110024504

  37. [37]

    Xing S, Zhao T, Zhou J and Sun Z 2024The Journal of Physical Chemistry C1282618–2624

  38. [38]

    Fleurence A, Friedlein R, Ozaki T, Kawai H, Wang Y and Yamada-Takamura Y 2012Physical Review Letters108245501

  39. [39]

    K¨ uchle J T, Baklanov A, Seitsonen A P, Ryan P T P, Feulner P, Pendem P, Lee T L, Muntwiler M, Schwarz M, Haag F, Barth J V, Auw¨ arter W, Duncan D A and Allegretti F 20222D Materials 9045021

  40. [40]

    Giannozzi P, Baroni S, Bonini N, Calandra M, Car R, Cavazzoni C, Ceresoli D, Chiarotti G L, Cococcioni M, Dabo I, Dal Corso A, De Gironcoli S, Fabris S, Fratesi G, Gebauer R, Gerstmann U, Gougoussis C, Kokalj A, Lazzeri M, Martin-Samos L, Marzari N, Mauri F, Mazzarello R, Paolini S, Pasquarello A, Paulatto L, Sbraccia C, Scandolo S, Sclauzero G, Seitsonen...

  41. [41]

    Perdew J P, Burke K and Ernzerhof M 1996Physical Review Letters773865–3868

  42. [42]

    Perdew J P and Zunger A 1981Physical Review B235048–5079

  43. [43]

    Baroni S, De Gironcoli S, Dal Corso A and Giannozzi P 2001Reviews of Modern Physics73 515–562

  44. [44]

    Ghosh S and Suryanarayana P 2017Computer Physics Communications212189–204

  45. [45]

    Ghosh S and Suryanarayana P 2017Computer Physics Communications216109–125

  46. [46]

    Plimpton S 1995Journal of Computational Physics1171–19

  47. [47]

    Tersoff J 1988Physical Review B389902–9905

  48. [48]

    Foiles S M, Baskes M I and Daw M S 1986Phys. Rev. B337983–7991

  49. [49]

    Rappe A K, Casewit C J, Colwell K S, Goddard W A and Skiff W M 1992Journal of the American Strain-controlled transition and flat-band tuning22 Chemical Society11410024–10035

  50. [50]

    Timpel D, Scheerschmidt K and Garofalini S H 1997Journal of Non-Crystalline Solids221187– 198

  51. [51]

    Peng Q, Wen X and De S 2013RSC Advances313772

  52. [52]

    G¨ urb¨ uz E and Sanyal B 2023arXiv preprint: 2305.08134

    Pith/arXiv arXiv

  53. [53]

    Wyckoff R W G 1963Crystal Structures2nd ed vol 1 (New York: Interscience Publishers)

  54. [54]

    Stich I, Car R and Parrinello M 1991Physical Review B4411092–11104

  55. [55]

    Petretto G, Gonze X, Hautier G and Rignanese G M 2018Computational Materials Science144 331–337

  56. [56]

    Lin C, Ponc´ e S and Marzari N 2022npj Computational Materials8236

  57. [57]

    Pallikara I, Kayastha P, Skelton J M and Whalley L D 2022Electronic Structure4033002

  58. [58]

    Galashev A Y and Vorob’ev A S 2021Physics Letters A408127487

  59. [59]

    Pr´ evot G, Hogan C, Leoni T, Bernard R, Moyen E and Masson L 2016Physical Review Letters 117276102

  60. [60]

    Kim D Y, Stefanoski S, Kurakevych O O and Strobel T A 2015Nature Materials14169–173

  61. [61]

    Qiu Y, Bender H, Richard O, Kim M S, Van Besien E, Vos I, de Potter de ten Broeck M, Mocuta D and Vandervorst W 2015Scientific Reports512692

  62. [62]

    Chen M X, Zhong Z and Weinert M 2016Physical Review B94075409