Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

Zener tunnelling in biased bilayer graphene via analytic continuation of semiclassical theory

T0 review · 3 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 By analytic continuation of semiclassical wavefunctions around the turning points, this paper derives the absolute pair-production rate for Zener tunnelling in biased bilayer graphene and fixes the normalization of the tunnelling current.

desk verdict A genuinely new fully analytic Zener rate for biased bilayer graphene, but the absolute prefactor rests on an unproven finite-ky replacement rule; worth refereeing after major revision. read the letter →

arxiv 2505.24150 v1 pith:KD6NCHYO submitted 2025-05-30 cond-mat.mes-hall

Harley Scammell , Oleg P. Sushkov This is my paper
classification cond-mat.mes-hall
keywords Zenertunnellingbiasedbilayergraphenesemiclassicalanalyticcontinuationpairproductionrateabsolutenormalisationcurrentbandgapevanescentwavefunctioncomponents
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 claims that Zener tunnelling in biased bilayer graphene—a two-layer carbon sheet with a voltage-tunable band gap—can be computed fully analytically by continuing semiclassical wavefunctions around the classical turning points in the complex plane. The central output is a closed-form pair-production rate per unit area, $\frac{dn}{dt}=\frac{m\Delta^2 e^{-\sqrt{2}S_0}}{2\pi(\pi\beta_1)^{1/2}}\left(\frac{F^2}{m\Delta^3}\right)^{3/4}\left[1-\frac{\cos(\sqrt{2}S_0-\pi/8)}{2^{1/4}}\right]$, which fixes the absolute normalization of the tunnelling current rather than just its exponential suppression. Along the way the paper finds that wavefunctions in classically allowed regions must keep exponentially decaying components, otherwise the matching conditions fail. If correct, the result turns earlier exponential estimates into a definite current formula applicable to devices such as tunnel field-effect transistors.

What carries the argument

The machinery is the analytic continuation of semiclassical wavefunctions around the two turning points $x_\pm$, performed by rotating $x-x_\pm=\rho e^{i\phi}$ through $\pm\pi$ so that matching across the barrier never touches the singular points. The exponents are controlled by the actions $S_0$ and $S_y$ of Eq. (3.8); the phase factors $e^{\pm i\pi S_0/4}$ carry the tunnel suppression, and the replacement rule $e^{\pm i\pi S_0/4}\to e^{\pm i\pi(S_0\mp iS_y)/4}$ extends the $k_y=0$ solution to finite transverse momentum, turning the integral over $k_y$ into the prefactor $(F^2/m\Delta^3)^{3/4}$. Matching at $x_-$ and $x_+$ determines both the scattering phase $\phi=-\pi/4$ and the transmission amplitude $T(k_y)$, with unphysical exponentially growing components cancelled exactly.

What would settle it

Solve the two-component Schrödinger equation (3.1) at fixed $k_y$ with absorbing boundary conditions, extract $|T(k_y)|^2$ from the asymptotic running wave, integrate over $k_y$, and compare the resulting pair-production rate with Eq. (4.8) across a range of $F$ and $\Delta$; a discrepancy in the prefactor, or in the amplitude or phase of the $1-\cos(\sqrt{2}S_0-\pi/8)/2^{1/4}$ factor, would falsify the absolute normalization.

Watch

Extended reading notes

Core claim

Starting from the effective two-component Schrödinger equation for biased bilayer graphene, the paper derives the transmission amplitude $T(k_y)=-2e^{i\pi/4}e^{-(S_0+S_y)/\sqrt{2}}\sin\left(\frac{S_0-S_y}{\sqrt{2}}\right)$, with $S_0=\beta_0\sqrt{2m}\,\Delta^{3/2}/F$ and $S_y=\beta_1\frac{k_y^2}{F}\sqrt{\frac{\Delta}{2m}}$, where $\beta_0=1.748$ and $\beta_1=1.198$. Integrating $|T(k_y)|^2$ over transverse momentum yields Eq. (4.8), an absolutely normalized pair-production rate. The paper's structural claim is that the matching problem requires two independent decaying solutions in the forbidden region and that decaying components must be retained in classically allowed regions; omitting them gives inconsistent matching. The matching conditions at both turning points fix the scattering phase and the amplitudes that earlier treatments left to numerical fitting, thereby establishing the absolute normalization and a small correction to the gap dependence of the tunnelling-rate coefficient.

Load-bearing premise

The load-bearing premise is the unproved replacement rule (3.9) that converts the zero-transverse-momentum matching solution into finite transverse momentum, together with the assumption that leading-order WKB wavefunctions give the order-one prefactor and not just the exponential suppression.

Editorial extensions

If this is right

  • The pair-production rate per unit area is fixed in absolute units by Eq. (4.8), and after multiplying by the spin-valley degeneracy $g_{s\tau}=4$ the tunnelling current for a device of length $L$ and width $W$ is $I=LW\,e\,dn/dt$.
  • The tunnelling rate oscillates with gap and field through the factor $1-\cos(\sqrt{2}S_0-\pi/8)/2^{1/4}$, so the current is not a monotonic exponential of $1/F$.
  • Two linearly independent decaying solutions are required inside the barrier, and exponentially decaying components must be retained in classically allowed regions for the matching to be consistent.
  • At transverse momenta where $T(k_y)=0$ the coefficient of the decaying component remains nonzero, implying localised modes with zero transmission at those $k_y$.
  • The same normalized pair-production rate results in both the long-device (linear-dispersion) and short-device (quadratic-dispersion) regimes, so the absolute normalization is independent of device length across the two limits.

Reading between the lines

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

  • The unproved replacement rule (3.9) is the main thing to test: a direct numerical solution of the two-component Schrödinger equation at finite $k_y$ would confirm or correct the prefactor $(F^2/m\Delta^3)^{3/4}$ and the oscillation amplitude.
  • If the absolute normalization is correct, existing exponent-only estimates of Zener currents in bilayer-graphene tunnel transistors can be upgraded to full current-voltage curves including interference oscillations.
  • The same analytic-continuation strategy should transfer to other gapped systems with quartic or non-quadratic dispersion where Airy-function matching is unavailable; the Dirac limit treated in the appendix is one boundary case.
  • The evanescent components in allowed regions suggest a connection to boundary-localised modes of non-Hermitian type, an analogy the paper itself raises; transport through p-n junctions could search for the predicted transmission zeros.
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

3 major / 6 minor

Summary. The paper presents a fully analytic calculation of the Zener electron-hole pair-production rate in biased bilayer graphene, using the Zwaan analytic-continuation variant of semiclassical theory. Starting from the two-band Hamiltonian (3.1), the authors construct semiclassical wavefunctions in the three spatial regions, match them by analytic continuation around the turning points, and obtain the ky-dependent transmission amplitude T(ky) in Eq. (3.33). Integrating |T(ky)|^2 over transverse momentum gives the central result, Eq. (4.8), for the pair-production rate per unit area, including an oscillatory interference term. The paper emphasizes that this is an absolute normalization, in contrast to the earlier hybrid numerical-WKB treatment of Ref. [13], and it also discusses the device-length dependence of the wavefunction normalization.

Significance. If the result is correct, Eq. (4.8) fixes the absolute prefactor of the Zener tunnelling rate in biased bilayer graphene, including the oscillatory field dependence coming from quantum interference. This is a nontrivial advance over prior work, which relied on fitting WKB coefficients to numerical solutions. The method itself, analytic continuation in the complex plane, is an elegant and parameter-free technique that avoids turning-point matching to special functions; the constants β0 and β1 are computed from definite integrals, with no fitted parameters. The paper also clarifies the non-standard structure of the semiclassical wavefunctions, in particular the necessity of retaining decaying components in classically allowed regions. However, the central prefactor rests on a finite-ky replacement rule, Eq. (3.9), that is stated without derivation, so the absolute normalization claim is not yet established to the standard claimed.

major comments (3)
  1. [§III.A.1, Eq. (3.9)] The finite-ky replacement rule e^{±iπ/4 S0} → e^{±iπ/4(S0 ∓ i Sy)} is stated without derivation. This rule is used in §III.C to obtain T(ky), and it controls both the Gaussian width β1 in the ky integral and the interference phase S0−Sy in Eq. (4.8). Because the absolute normalization of the pair-production rate is the paper's central claim, this step needs a derivation from the ky-dependent semiclassical momentum p_F(x)^2−ky² and the associated analytic continuation around the shifted turning points, or an independent numerical check. As written, the rule is an ad hoc ansatz; a different natural extension, such as replacing S0 by S0−Sy in the matching exponentials, would change the prefactor and the interference phase.
  2. [§III.C, Eq. (3.33) and §IV.2, Eq. (4.8)] The transmission amplitude and the resulting pair-production rate are not benchmarked against any numerical solution of the bilayer Schrödinger equation (3.1), nor against the numerical WKB results of Ref. [13]. The numerical check in Appendix A is for the Dirac Hamiltonian, not for biased bilayer graphene. Since the paper claims that leading-order WKB provides the absolute prefactor and not just the exponent, an independent check is needed to rule out O(1) corrections to the prefactor in Eq. (4.8).
  3. [Eq. (3.4)] Equation (3.4) as printed, p(x)^2 = ±2m√(F²x²−Δ²−k_y²), is dimensionally inconsistent and cannot be reconciled with the definitions of p_F, p_A, S0, and Sy in Eqs. (3.6)–(3.8). The correct relation should be p(x)^2 = ±2m√(F²x²−Δ²) − k_y², with p denoting the x-component of momentum; this yields the turning points in Eq. (3.5) and the expansion used for Sy. Please correct this equation and ensure all subsequent ky expansions are consistent with it.
minor comments (6)
  1. [General presentation] The displayed text contains repeated copies of the abstract and of Section II, and an earlier draft with unresolved 'Levitov [cite]' placeholders; the published version should be cleaned of these artifacts.
  2. [§III.A.2] There is a typographical error in 'the distance from the turning point is march larger than the wavelength'; this should read 'much larger'.
  3. [§IV.3] The phrase 'multiply the wave fanction by √p−∞' should read 'multiply the wave function by √p−∞'.
  4. [§IV.2, Eq. (4.8)] The notation gsτ = 4 for the spin-valley degeneracy is not defined; please clarify, e.g. as g_{sv} = 4.
  5. [Appendix A.3, Eq. (A13)] In the clockwise continuation line, 'cl : (−ipF(x))^{−1/2} → −i(pA(x))^{1/2}' appears to have the wrong power; it should likely be (pA(x))^{−1/2} to be dimensionally consistent.
  6. [Fig. 6] The axis label 'J/v exp(πΔ²_k/vF) vs −πΔ²_k/vF' is garbled; please rewrite it so that the exponential rescaling and the horizontal variable are immediately clear.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pair-production rate follows from a parameter-free semiclassical matching calculation, with no fitted inputs or load-bearing self-citations.

full rationale

The central derivation is self-contained. The transmission amplitude T(ky) in Eq. (3.33) is obtained by analytic continuation of the semiclassical solutions (3.10)-(3.15) across the turning points, with the unknown coefficients fixed by the condition that exponentially growing components cancel, Eqs. (3.26) and (3.31); no parameter is fitted to data. The constants beta0 and beta1 in Eq. (3.8) are computed integrals, and the final rate in Eq. (4.8) follows by integrating |T(ky)|^2 over ky with densities of states from the normalized standing wave. The finite-ky replacement rule (3.9) is stated without derivation, but it is an analytic leading-order approximation in ky^2, not an input defined in terms of the predicted quantity; an unproved approximation is a correctness concern, not a circular reduction. The self-citations (Refs. [3] and [7]) are contextual references in the introduction and play no role in the derivation. The paper also benchmarks against the independent numerical-WKB treatment of Ref. [13] and, for the Dirac analogue, against direct numerical solution in Appendix A. Therefore no circular step is present.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central claim is a parameter-free analytic derivation; it rests on standard WKB/analytic-continuation mathematics plus domain assumptions about the Bblg Hamiltonian, the ky-expansion rule, and the leading-order WKB prefactor assumption. No new entities are introduced.

assumptions (4)
  • domain assumption The physical wavefunction in each spatial region is a linear combination of the listed semiclassical basis states, and exponentially growing components are discarded.
    Used throughout Sec. III; this is standard WKB/Stokes selection, but it is not proven for the quartic-dispersion ODE and the paper's key matching depends on it.
  • ad hoc to paper Finite-ky dependence enters via the replacement e^{±iπ/4 S0} → e^{±iπ/4(S0 ∓ iSy)}, retaining only leading order in ky².
    Stated in Sec. III.A.1 and applied in Sec. III.C; not derived, and it controls the ky-integral in the final rate formula.
  • ad hoc to paper The semiclassical expansion truncated at leading order gives the exact prefactor, not only the exponent.
    The derivation works at leading order in 1/pA; corrections are not estimated, yet the absolute normalization claim depends on the prefactor.
  • domain assumption In a long device, the Bblg wavefunction matches the gapped Dirac standing wave in the overlap region, fixing normalization.
    Sec. IV.1, Eq. (4.5); relies on the condition FL > γ1/2 and on the matching region being in the linear regime.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Zener tunnelling in biased bilayer graphene via analytic continuation of semiclassical theory." pith.science (2026). https://pith.science/paper/KD6NCHYO

@misc{pith2026250524150,
  author       = {Pith},
  title        = {Pith review of: Zener tunnelling in biased bilayer graphene via analytic continuation of semiclassical theory},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KD6NCHYO}},
  note         = {Machine review of arXiv:2505.24150}
}
read the original abstract

Employing a semiclassical method based on analytic continuation, we compute the electron-hole pair production rate in biased bilayer graphene subject to an in-plane electric field. This approach, originally due to Zwaan, bypasses the need for exact solutions at turning points, which are generally unavailable beyond linear or quadratic band structures. Applying this technique to biased bilayer graphene reveals non-standard features of the asymptotic wavefunctions, in particular the necessity of retaining decaying components even in classically allowed regions. By providing a fully analytic solution, this work complements and clarifies earlier results based on hybrid analytical-numerical treatments, and importantly establishes the absolute normalisation of the pair production rate -- and hence of the tunnelling current.

Figures

Figures reproduced from arXiv: 2505.24150 by the authors.

Figure 4
Figure 4. FIG. 4: Two branches of the dispersion near the [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. FIG. 5: Pair production rate per unit time and area, [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. FIG. 6: Numerical evaluation of the tunnelling current [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Exciton condensation from level repulsion: application to bilayer graphene

    cond-mat.str-el 2025-06 conditional novelty 6.0 of 10

    An in-plane electric field couples the s- and p-wave excitons of biased bilayer graphene, and the resulting level repulsion can drive the lower exciton branch below zero energy, producing an exciton condensate.

Reference graph

Works this paper leans on

37 extracted references · 36 canonical work pages · cited by 1 Pith paper

  1. [13]

    Spontaneous momentum polarization and diodicity in Bernal bilayer graphene

    J.-X. Lin, Y. Wang, N. J. Zhang, K. Watanabe, T. Taniguchi, L. Fu, and J. I. A. Li, “Spontaneous momentum polarization and diodicity in bernal bilayer graphene,” (2023), arXiv:2302.04261 [cond-mat.mes- hall]

    work page Pith review arXiv 2023

  2. [1]

    Notation and definitions Before proceeding, we define the classical momentum in the forbidden and allowed regions, denotedpF (x) and pA(x) respectively, as pF (x) ≡ √ 2m(∆2 − F 2x2)1/4, pA(x) ≡ √ 2m(F 2x2 − ∆2)1/4. (3.6) In the wave function expressions, we will also make re- peated use of αA(x) ≡ F x+ ∆ F x− ∆ 1/4 , α F (x) ≡ ∆ + F x ∆ − F x 1/4 (3.7) wh...

    work page

  3. [2]

    Semiclassical solution for the running escaping wave at x >∆/F , i.e

    Region I: x >∆/F Let us consider for now the case ky = 0. Semiclassical solution for the running escaping wave at x >∆/F , i.e. a classically allowed region, is a(x) = T 2 p pA(x) αA(x)eiS b(x) = T 2 p pA(x) αA(−x)eiS pA(x) = √ 2m(F 2x2 − ∆2)1/4 S = Z x ∆/F pA(x′)dx′. (3.10) 4 Here T is the transmission amplitude. The running wave carries current jx = |T ...

    work page

  4. [3]

    (3.12) The sign ± in the phase correlates with the sign ∓ in b(x)

    Region II: −∆/F < x <∆/F In the classically forbidden region,−∆/F < x <∆/F there are two exponentially decaying solutions a(x) = 1p 2pF (x) αF (x) exp(−e±iπ/4S) b(x) = ∓ip 2pF (x) αF (−x) exp(−e±iπ/4S) S = Z x −∆/F pF (x′)dx′. (3.12) The sign ± in the phase correlates with the sign ∓ in b(x). Of course the solutions (3.12) can be multiplied by arbitrary c...

    work page

  5. [4]

    Region III: x <−∆/F In the classically allowed region, x <−∆/F , there is the following standing wave solution a(x) = 1√pA(x) αA(x) cos(S + ϕ) b(x) = − 1√pA(x) αA(−x) cos(S + ϕ) S = Z −∆/F x pA(x′)dx′. (3.13) Landau-Zener tunnelling in biased bilayer graphene: Analytic continuation of the semicalssical theory We compute the Zener tunnelling rate in biased...

    work page

  6. [5]

    As discussed in Ref

    We will consider two regimes, long and short devices, which we make precise now. As discussed in Ref. [20], for ∆ = 0, the low-energy dispersion exhibits a crossover at momentum scale p ≈ γ1/(2v) from quadratic to linear behaviour. Let us de- note the corresponding energy scale ∆ 1 = γ1/2 ≈ 380 meV. At small momentum, the spectrum is approxi- mately quadr...

    work page

  7. [6]

    (4.1) In this case the standing wave far away on the left from the tunnelling region is best approximated as the solu- tion of an effective linear Dirac equation

    Normalisation in the linear regime (long device) Consider the sufficiently long sample, L, such that we enter the linear regime set by F L >∆1. (4.1) In this case the standing wave far away on the left from the tunnelling region is best approximated as the solu- tion of an effective linear Dirac equation. Full treatment of the Dirac equation is presented ...

    work page

  8. [7]

    Therefore the x- density of states is ρx = dpx π = dε πv = F dx πv

    Pair production rate in the linear regime Again, the wave function (4.5) asymptotically, x → −∞ is normalised as |a|2 + |b|2 = 1. Therefore the x- density of states is ρx = dpx π = dε πv = F dx πv . (4.6) Note that we divide byπ instead of 2π because this is the standing wave. We also account that dpx = d(vpx)/v = dϵ/v = F dx/v, here dx is the length that...

    work page

Show all 37 references
  1. [8]

    (4.9) According to Eq

    Normalisation and tunnelling rate in the quadratic regime (short device) Consider now the short device such that ∆ ≪ F L≪ ∆1. (4.9) According to Eq. (3.13) the wave function at x → −∞ is φ = 1√p−∞ cos(S + ϕ) 1 −1 ! (4.10) where p−∞ is momentum at x → −∞. Hence, to get the nece...

  2. [9]

    Direct observation of a widely tunable bandgap in bi- layer graphene,

    Y. Zhang, T.-T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl, M. F. Crommie, Y. R. Shen, and F. Wang, “Direct observation of a widely tunable bandgap in bi- layer graphene,” Nature 459, 820 (2009)

    2009

  3. [10]

    Electronic transport in dual-gated bilayer graphene at large dis- placement fields,

    T. Taychatanapat and P. Jarillo-Herrero, “Electronic transport in dual-gated bilayer graphene at large dis- placement fields,” Phys. Rev. Lett. 105, 166601 (2010)

    2010

  4. [11]

    Screening of the band gap in electrically biased bi- layer graphene: From hartree to hartree-fock,

    J. N. Engdahl, Z. E. Krix, and O. P. Sushkov, “Screening of the band gap in electrically biased bi- layer graphene: From hartree to hartree-fock,” (2025), arXiv:2505.04381 [cond-mat.mes-hall]

    2025 arXiv

  5. [12]

    Persistent current states in bilayer graphene,

    J. Jung, M. Polini, and A. H. MacDonald, “Persistent current states in bilayer graphene,” Phys. Rev. B 91, 155423 (2015)

    2015

  6. [14]

    Isospin- and momentum-polarized orders in bi- layer graphene,

    Z. Dong, M. Davydova, O. Ogunnaike, and L. Lev- itov, “Isospin- and momentum-polarized orders in bi- layer graphene,” Phys. Rev. B 107, 075108 (2023)

    2023

  7. [15]

    Exciton conden- sation in biased bilayer graphene,

    H. D. Scammell and O. P. Sushkov, “Exciton conden- sation in biased bilayer graphene,” Phys. Rev. Res. 5, 043176 (2023)

    2023

  8. [16]

    Cascade of isospin phase transitions in bernal-stacked bilayer graphene at zero magnetic field,

    S. C. de la Barrera, S. Aronson, Z. Zheng, K. Watan- abe, T. Taniguchi, Q. Ma, P. Jarillo-Herrero, and R. Ashoori, “Cascade of isospin phase transitions in bernal-stacked bilayer graphene at zero magnetic field,” Nature Physics 18, 771 (2022)

    2022

  9. [17]

    Isospin magnetism and spin-polarized superconductivity in bernal bilayer graphene,

    H. Zhou, L. Holleis, Y. Saito, L. Cohen, W. Huynh, C. L. Patterson, F. Yang, T. Taniguchi, K. Watan- abe, and A. F. Young, “Isospin magnetism and spin-polarized superconductivity in bernal bilayer graphene,” Science 375, 774 (2022)

    2022

  10. [18]

    A theory of the electri- cal breakdown of solid dielectrics,

    C. Zener and R. H. Fowler, “A theory of the electri- cal breakdown of solid dielectrics,” Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character 145, 523 (1934)

    1934

  11. [19]

    ¨Uber das verhalten eines elektrons im homo- genen elektrischen feld nach der relativistischen theorie diracs,

    F. Sauter, “ ¨Uber das verhalten eines elektrons im homo- genen elektrischen feld nach der relativistischen theorie diracs,” Zeitschrift f¨ ur Physik69, 742 (1931)

    1931

  12. [20]

    Consequences of dirac’s theory of positrons,

    W. Heisenberg and H. Euler, “Consequences of dirac’s theory of positrons,” Zeitschrift f¨ ur Physik 98, 714 (1936)

    1936

  13. [21]

    Common-path in- terference and oscillatory zener tunneling in bilayer graphene p-n junctions,

    R. Nandkishore and L. Levitov, “Common-path in- terference and oscillatory zener tunneling in bilayer graphene p-n junctions,” Proceedings of the National Academy of Sciences 108, 14021 (2011)

    2011

  14. [22]

    Ground-state decay rate for the zener breakdown in band and mott insulators,

    T. Oka and H. Aoki, “Ground-state decay rate for the zener breakdown in band and mott insulators,” Physical Review Letters 95, 137601 (2005)

    2005

  15. [23]

    Quantum elec- trodynamics with anisotropic scaling: Heisenberg–euler action and schwinger pair production in the bilayer graphene,

    M. I. Katsnelson and G. E. Volovik, “Quantum elec- trodynamics with anisotropic scaling: Heisenberg–euler action and schwinger pair production in the bilayer graphene,” JETP Letters 95, 411 (2012)

    2012

  16. [24]

    Steep-slope switching using bilayer graphene tunnel field effect transistors,

    K. Ghosh and S. D. Sarkar, “Steep-slope switching using bilayer graphene tunnel field effect transistors,” IEEE Transactions on Electron Devices 60, 2165 (2013)

    2013

  17. [25]

    Steep slope field effect transis- tors based on 2d materials,

    L. Qin, H. Tian, C. Li, Z. Xie, Y. Wei, Y. Li, J. He, Y. Yue, and T.-L. Ren, “Steep slope field effect transis- tors based on 2d materials,” Advanced Electronic Ma- terials 10, 2300625 (2024)

    2024

  18. [26]

    On the calculation of the transmission co- efficient by means of the phase-integral method,

    B. J. Zwaan, “On the calculation of the transmission co- efficient by means of the phase-integral method,” Phys- ica 9, 401 (1942)

    1942

  19. [27]

    L. D. Landau and E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, 3rd ed., Course of Theoretical Physics, Vol. 3 (Pergamon 10 Press, 1977) translated from the Russian by J. B. Sykes and J. S. Bell

    1977

  20. [28]

    The electronic properties of bilayer graphene,

    E. McCann and M. Koshino, “The electronic properties of bilayer graphene,” Reports on Progress in Physics76, 056503 (2013)

    2013

  21. [29]

    Common path interference in zener tunneling is a uni- versal phenomenon,

    S. Johri, R. Nandkishore, R. N. Bhatt, and E. J. Mele, “Common path interference in zener tunneling is a uni- versal phenomenon,” Phys. Rev. B 87, 235413 (2013)

    2013

  22. [30]

    The return of the quartic oscillator: The complex wkb method,

    A. Voros, “The return of the quartic oscillator: The complex wkb method,” Annales de l’I.H.P. Physique th´ eorique39, 211 (1983)

    1983

  23. [31]

    Edge states and topological in- variants of non-hermitian systems,

    S. Yao and Z. Wang, “Edge states and topological in- variants of non-hermitian systems,” Physical Review Letters 121, 086803 (2018)

    2018

  24. [32]

    Ex- ceptional topology of non-hermitian systems,

    E. J. Bergholtz, J. C. Budich, and F. K. Kunst, “Ex- ceptional topology of non-hermitian systems,” Reviews of Modern Physics 93, 015005 (2021). Appendix A: 2D GAPPED DIRAC

    2021

  25. [33]

    (A1) The full wavefunction Ψ R = φReiS+ is given by ΨR = r J 2v s F pA(x) (x − x−) 1 2 (x − x+) 1 2 ! eiS+

    x > x+ Here we consider the escaping wave, pA(x) > 0, and solve (at ε = 0, ky = 0) (pA(x)σx + ∆σz − F xσ0)φR = 0. (A1) The full wavefunction Ψ R = φReiS+ is given by ΨR = r J 2v s F pA(x) (x − x−) 1 2 (x − x+) 1 2 ! eiS+ . (A2) Normalisation is such the the current along x is J

  26. [34]

    Under this, acl : ( x − x+) 1 2 → i(x+ − x) 1 2 , (A3) cl : ( x − x+) 1 2 → −i(x+ − x) 1 2 (A4) and acl : iS+ → −S+, (A5) cl : iS+ → S+

    x− < x < x+ Starting at x ∼ x+, we perform acl and cl continua- tion via x − x+ = ρe±iπ, respectively. Under this, acl : ( x − x+) 1 2 → i(x+ − x) 1 2 , (A3) cl : ( x − x+) 1 2 → −i(x+ − x) 1 2 (A4) and acl : iS+ → −S+, (A5) cl : iS+ → S+. (A6) Running from right to left in x,...

  27. [35]

    Since we desire a standing wave for x < x−, we will need to keep both paths in the complex plain

    x < x− From Ψ F (x) we analytically continue in x − x− → (x− − x)e±iπ, with sign corresponding to acl and cl, respectively. Since we desire a standing wave for x < x−, we will need to keep both paths in the complex plain. Explicitly, these paths give acl : − S− → −iS−, (A8) cl...

  28. [36]

    (A18) One can verify the current (A18) by direct numerical methods

    Current Deep in x ≪ x−, the standing wave is normalised to unity (with A =area), and hence we can find the current J, 1 A ||ΨL||2 → 2J v e2S0 = 1 (A17) and hence J = 1 2 ve−2S0 . (A18) One can verify the current (A18) by direct numerical methods. Here we solve the Dirac Equati...

  29. [37]

    Therefore the x-density of states is ρx = dpx π = dϵ πv = F dx πv (A21) Note that we divide byπ instead of 2π because this is the standing wave

    Pair-production rate The wave function (A16) asymptotically, x → −∞is normalised as |a|2 + |b|2 = 1. Therefore the x-density of states is ρx = dpx π = dϵ πv = F dx πv (A21) Note that we divide byπ instead of 2π because this is the standing wave. We also account that dpx = d(vp...

Pith tools

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