Pith. sign in

REVIEW 2 major objections 5 minor 9 references

Magnetic Double-Wells: Lower Bounds on Tunneling

T0 review · 2 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

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

Pith's one-line read Magnetic double-well tunneling is generically nonzero: for any small magnetic field and for all coupling constants outside a set of density zero, the hopping coefficient and the lowest-two eigenvalue splitting are at least exp(-λ^{1+ε}).

desk verdict A detailed proof that magnetic double-well tunneling is generically super-exponentially small, but with a missing resolvent identity for complex-λ Landau operators that the paper relies on. read the letter →

arxiv 2511.13470 v2 pith:QHK6M23S submitted 2025-11-17 math-ph cond-mat.mes-hallmath.APmath.CAmath.FAmath.MP

Charles L. Fefferman , Jacob Shapiro , Michael I. Weinstein This is my paper
classification math-phcond-mat.mes-hallmath.APmath.CAmath.FAmath.MP MSC 81Q1081Q1581Q2035P15
keywords magneticSchrödingeroperatorsdouble-wellpotentialquantumtunnelingeigenvaluesplittinghoppingcoefficientanalyticcontinuationLandauHamiltonianlowerbounds
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 establishes that exact vanishing of tunneling in magnetic double-well systems is a non-generic phenomenon. For any sufficiently small magnetic field strength b and for all large coupling constants λ outside a set of density zero, both the tunneling amplitude and the energy splitting between the two lowest states are bounded below by exp(-λ^{1+ε}). The proof proceeds by analytically continuing the single-well resolvent, acting on compactly supported functions, into a wedge of the complex λ plane, then applying a complex-analysis lemma that converts one-point lower bounds plus exponential upper bounds into average and pointwise lower bounds. If correct, the specially constructed potentials with exactly zero tunneling are isolated exceptions, and generic couplings always give at least super-exponentially small tunneling.

What carries the argument

The load-bearing object is the analytically continued one-well resolvent A_λ = (h_λ - z_λ)^{-1} Q, initially defined for real λ and extended to the wedge Ω_{Λ,ε} = {|λ|>Λ, |arg λ|<π/2-ε}; Q is the projection onto the support of the potential. Its construction passes through a cut-off Landau Hamiltonian, a dyadic partition into mesoscopic annuli, and a local elliptic pseudodifferential parametrix, together with an analytic resolvent theory for the anisotropic magnetic harmonic oscillator. Once A_λ is available, the normalized quantities C_λ C̄_λ ρ(λ) and γ_λ² Δ(λ)² are analytic in the wedge and bounded by exp(C b|λ|) up to polynomial factors. A complex-analysis lemma based on Blaschke factori

What would settle it

Compute ρ0(λ) and Δ0(λ) numerically, for a fixed very small b and many large λ in the same class of double-well potentials; if the set of λ in [Λ,M] where either quantity vanishes or falls below exp(-λ^{1.1}) does not have density tending to zero as M→∞, the theorem is false. A more direct check is to evaluate the analytically continued operator A_λ from Lemma 2.3 on a compactly supported vector for complex λ in the wedge; if its norm grows faster than exp(C b|λ|) with C independent of b, the proof's exponential bounds are violated.

Watch

Extended reading notes

Core claim

For C^3 compactly supported single-well potentials with a unique non-degenerate minimum and sufficiently separated wells, the magnetic double-well splitting Δ0(λ) and the hopping coefficient ρ0(λ) satisfy |ρ0(λ)|, Δ0(λ) ≥ exp(-λ^{1+ε}) for all λ outside a set of density zero, provided b is sufficiently small. The exceptional set can be chosen discrete, {λ_k}, with Σ λ_k^{-(1+ε)} < ∞, so persistent exact cancellations are excluded. In addition, logarithmic averages of -log|ρ0| and -log Δ0 over [λ,2λ] are bounded by C λ^{1+ε}. The method is new: it establishes analyticity of the one-well resolvent on compactly supported vectors in a wedge, derives exponential upper bounds, and uses a Blaschke–

Load-bearing premise

The whole argument leans on Lemma 2.3: that the single-well resolvent, when cut off to the support of the potential, continues analytically into a wedge of the complex coupling plane with the stated exponential bound exp(C b|λ|) and off-diagonal decay; if that continuation fails for some tiny nonzero b, the generic lower bounds do not follow.

Editorial extensions

If this is right

  • For small b, exact zero tunneling can occur only on a discrete, very sparse set of coupling constants; away from that set, both the splitting and the hopping coefficient are nonzero.
  • The lower bound exp(-λ^{1+ε}) shows that tunneling is always at least super-exponentially small in the coupling, but cannot vanish identically on an interval.
  • The averaged bounds imply that the typical logarithm of the splitting is O(λ^{1+ε}) over long intervals, giving a quantitative generic tunneling rate.
  • The result turns the known exact-vanishing counterexamples into exceptional potentials: generic non-radial wells of the same class do not eliminate tunneling.
  • The splitting is treated directly rather than deduced from the hopping coefficient, avoiding the need for the stronger lower bound that the asymptotic ratio Δ0/(2|ρ0|)→1 would require.

Reading between the lines

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

  • If the exponent 1+ε could be sharpened to 1, the lower bounds would align with the scale needed for tight-binding reductions for generic double wells; the paper explicitly leaves this improvement open.
  • The technical device of continuing a resolvent only after restricting it to compactly supported functions may apply to other spectral and transport quantities in strong magnetic fields where the full resolvent has no analytic continuation.
  • For b of order one rather than small, the same template would plausibly yield a weaker lower bound exp(-λ^{K}) outside a density-zero set, provided the relevant analytic continuation holds; the paper states this as a belief, not a theorem.
  • The genericity statement is in λ, not in b: the proof does not exclude the possibility that for finitely many special b values the hopping coefficient could vanish identically for all large λ, so a full b-phase diagram remains open.
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

2 major / 5 minor

Summary. The paper proves that, for a symmetric magnetic double well built from a C^3 compactly supported single-well potential with a unique non-degenerate minimum, and for sufficiently small relative magnetic-field strength b, the hopping coefficient rho0(lambda) and the eigenvalue splitting Delta0(lambda) are bounded below by exp(-lambda^{1+epsilon}) for all sufficiently large lambda outside a set of density zero (Theorem 1.1). The proof proceeds by analytically continuing the one-well resolvent acting on compactly supported states into a wedge Omega_{Lambda,epsilon}, using a cut-off Landau parametrix and a Landau resolvent kernel for complex magnetic fields; the resulting analytic functions are then controlled by a self-contained complex-analysis lemma based on Blaschke factorization and Herglotz representation, together with a single-point lower bound inherited from the non-magnetic case. The paper complements the authors' earlier construction of potentials with exactly vanishing tunneling.

Significance. If the central analytic-continuation step is made fully rigorous, this is a significant result: it establishes that exact vanishing of tunneling is non-generic and provides the first lower bounds of super-exponentially small type for magnetic double wells in the strong-field regime. The complex-analysis machinery in Section 3 is clean and essentially self-contained, and the appendix by Shpigel gives detailed resolvent bounds for the anisotropic magnetic harmonic oscillator that are of independent interest. The paper is also careful to separate what is proved from what is conjectured, e.g. Remark 1.3. However, as detailed below, the proof currently relies on an unproved resolvent-identity property of the Landau kernel for non-real lambda, which is load-bearing for the main theorem.

major comments (2)
  1. [§2.1.2 (Lemma 2.3); §C (Prop. C.1)] Lemma 2.3 is the load-bearing step: the analytic continuation A_lambda of (h_lambda - z_lambda)^{-1}Q is constructed by rewriting [h_lambda, chi] A^theta_lambda Q as (H_Landau - z_lambda) R^Landau_lambda(z_lambda)[h_lambda, chi] A^theta_lambda Q - lambda^2 v R^Landau_lambda(z_lambda)[h_lambda, chi] A^theta_lambda Q, and then forming A_lambda = (chi - R^Landau[h_lambda, chi]) A^theta_lambda Q (1 + K_lambda)^{-1}. This requires that R^Landau_lambda(z_lambda) be a right inverse of H_Landau - z_lambda on the compactly supported functions [h_lambda, chi] A^theta_lambda Q psi. Proposition C.1, however, only proves that the kernel (C.3)-(C.5) defines a bounded operator QL^2 -> L^2 with exponential off-diagonal decay; it does not prove the resolvent identity (H_Landau - z_lambda) R^Landau_lambda(z_lambda) g = g for non-real lambda. This is not a formality: Theorem C.2 shows that the full L^2 res
  2. [§2.2-§2.3; Theorem 1.1(2)] Theorem 1.1(2) asserts the existence of a discrete exceptional set {lambda_k} with sum lambda_k^{-(1+epsilon)} < infinity on which rho_0 and Delta_0 can vanish. The proof in Sections 2.2 and 2.3 invokes Corollary 3.10 and obtains only averaged logarithmic bounds, such as (1.9). No argument is supplied for the summability of the real zero set. Corollary 3.10's averaged statement alone is too weak to imply this; the Blaschke estimates in Section 3.1, especially (3.6) together with the conformal change of Section 3.4, would yield the needed summability, but this connection is not made. Please add an explicit lemma or proof.
minor comments (5)
  1. [Throughout] Several cross-references and headings are mismatched: 'Proof of Theorem 2.1' should be 'Proof of Lemma 2.1', 'Proof of Theorem 2.3' should be 'Proof of Lemma 2.3', and 'Theorem E.1' is actually 'Lemma E.1'. Also, Proposition C.1 is sometimes called 'Theorem C.1'.
  2. [§3] Typo: 'Blashcke' should be 'Blaschke' in the heading before Theorem 3.2.
  3. [§C] In (C.3), the condition 'Re{z}<B' is written for complex B; it should be stated as 'Re{z}<Re{B}' or clarified that the formula is initially for real B and then continued.
  4. [§2.2] The step from the averaged lower bound (1.9) to the pointwise lower bound (1.6) outside a density-zero set is not spelled out. A short Markov/Chebyshev argument over dyadic intervals would make the implication explicit.
  5. [§2.3] The one-point lower bound for Delta_0 uses (2.34), which is not proved in this paper and is cited only implicitly to [FSW25a]. Please cite it explicitly and state that the error is controlled at the chosen lambda_star.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the generic tunneling lower bound rests on a self-contained analytic-continuation construction and an independently proved non-magnetic lower bound at one point.

full rationale

The derivation chain is not circular. The main ingredients are: (i) Lemma 2.3, which constructs an analytic continuation Aλ of (hλ−zλ)−1Q via a cut-off parametrix, the Landau resolvent on compactly supported functions (Theorem C.1), and off-diagonal decay estimates; (ii) the complex-analysis Lemma 3.1/Theorem 3.10, proved from the Blaschke factorization and Herglotz/Poisson representation, giving average lower bounds from one-point lower bounds and global upper bounds; and (iii) a single-point lower bound at λ★ taken from the non-magnetic result [FLW18, Section 15.3] and transferred to small b by continuity. These are independent external inputs used as normalization anchors, not fitted parameters. The paper explicitly refuses to use the self-cited relation (2.34) from [FSW25a] to convert ρ0 lower bounds into Δ0 lower bounds, stating 'this is doomed to fail', and instead proves analyticity of Δ0 directly through the Gram-matrix identity (2.41), so self-citations are not load-bearing. The acknowledgements and Remark 1.3 are honest about limitations ('we know no way to rule out the existence of b for which, say, ρ0(λ)=0 for all large λ'), which further supports a non-circular assessment. The skeptic's concern about the resolvent identity in Lemma 2.3 is a possible correctness gap in an estimate, not a circular reduction of the claimed result to its own inputs; it does not raise the circularity score under the stipulated rules.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim is built on standard spectral-theory background and two external mathematical results (Matsumoto's eigenvalue asymptotics and the FLW18 non-magnetic lower bound). The paper introduces no new physical entities. The free parameters are proof parameters chosen by hand to make the estimates work, not fitted to data. The principal unstated burden is the heavy analytic-continuation machinery in Lemma 2.3 and Appendix D.

free parameters (4)
  • exponent η in δ = Λ^{-1/2+η}
    Introduced by hand in Section 2.1.1 (2.19) to make the partition-of-unity and pseudodifferential error terms small; it must satisfy η>0 small and Λδ^2→∞. Not fitted to data.
  • small-magnetic-field threshold b_ε
    Existential smallness threshold for b in Theorem 1.1; chosen in the proofs of Lemma 2.1, Lemma 2.3, and Section 4 to make error terms small. No quantitative value is given.
  • point λ⋆ for the lower-bound-at-a-point input
    A large real point chosen in Section 2.2 where the non-magnetic lower bound from [FLW18] applies and continuity in b gives a nonzero lower bound. Not fitted to the conclusion.
  • arbitrary ε>0 in the theorem
    Controls the size of the wedge and the exponent 1+ε in the lower bound; a proof parameter, not a fitted constant.
assumptions (5)
  • domain assumption Matsumoto [Mat94] semiclassical eigenvalue asymptotics: e_{λ,j} = -λ^2 + e_j^{MHO} λ + O(λ^{1/2}) and analogous statements for the double well (equation (2.7) and around (E.11)).
    Invoked to justify that the contour z_λ(ξ) in (2.8) encircles exactly one eigenvalue at distance of order λ from the rest of the spectrum. The paper cites [Mat94] rather than proving it.
  • domain assumption Non-magnetic hopping lower bound from [FLW18, Section 15.3]: |ρ0(λ★,0)| ≥ C λ★^2 exp(-2 c d_1 λ★) for large λ★.
    Used in Section 2.2 as the single-point lower bound that is then transferred by continuity to small b. This is an external result, not proved in the paper.
  • standard math Landau resolvent kernel formula in terms of Tricomi's U-function (equation (C.3)), quoted from [KP04, Lemma 5.1].
    The paper uses this explicit formula to construct the analytic continuation of the Landau resolvent on compactly supported states in Theorem C.1.
  • standard math Anisotropic magnetic harmonic oscillator heat-kernel formulas from [Mat95], with sign corrections stated and then reproved in Appendix D.
    Appendix D derives the needed estimates from the explicit heat kernel; the corrections to [Mat95] are identified and the derivation is self-contained within the appendix.
  • standard math Standard complex analysis facts: Blaschke factorization, Herglotz/Poisson representation for positive harmonic functions on the half-plane, and the conformal map between H and a wedge.
    Used throughout Section 3 to prove the lower-bound lemma; these are classical theorems, not invented for the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Magnetic Double-Wells: Lower Bounds on Tunneling." pith.science (2026). https://pith.science/paper/QHK6M23S

@misc{pith2026251113470,
  author       = {Pith},
  title        = {Pith review of: Magnetic Double-Wells: Lower Bounds on Tunneling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QHK6M23S}},
  note         = {Machine review of arXiv:2511.13470}
}
read the original abstract

We study double-well systems with strong magnetic fields and deep potential wells. We present lower bounds on tunneling rates for generic values of the coupling constant. This result was recently announced and complements our recent counter-example construction which exhibits vanishing tunneling for specially-constructed double-well potentials.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

9 extracted references · 2 linked inside Pith

  1. [1]

    [AHS78] Avron, J., Herbst, I., and Simon, B.: Schrödinger operators with magnetic fields. I. general interactions Duke Mathematical Journal.45(4), 847–883 (1978) [Daa25] Daalhuis, A. B. O.: Confluent Hypergeometric Functions. NIST Digital Li- brary of Mathematical Functions, §13.2 and §13.4, Version 1.2.4 (released 2025-03-15). 2025.url:https://dlmf.nist....

    1978

  2. [4]

    arXiv:2509.02857 [math-ph]. [FSW25c] Fefferman,C.L.,Shapiro,J.,andWeinstein,M.I.:Quantumtunnelingandits absence in deep wells and strong magnetic fields Proceedings of the National Academy of Sciences.122(8), e2420062122 (2025) [GT01] Gilbarg, D. and Trudinger, N. S.: Elliptic Partial Differential Equations of Second Order. Classics in Mathematics. Reprin...

    arXiv 2025

  3. [19]

    Providence, RI: American Mathematical Society, 2010, pp

    Graduate Stud- ies in Mathematics. Providence, RI: American Mathematical Society, 2010, pp. xxii+749. [FSW22] Fefferman, C., Shapiro, J., and Weinstein, M. I.: Lower Bound on Quantum Tunneling for Strong Magnetic Fields SIAM Journal on Mathematical Anal- ysis.54(1), 1105–1130 (2022) [FLW18] Fefferman,C.L.,Lee-Thorp,J.P.,andWeinstein,M.I.:HoneycombSchrödin...

    2010

  4. [1981]

    New York: McGraw-Hill,

    [Mat95] Matsumoto,H.:SemiclassicalAsymptoticsofEigenvaluesforSchrödingerOp- erators with Magnetic Fields Journal of Analysis.129(1), 168–190 (1995) [Mat94] Matsumoto,H.:SemiclassicalasymptoticsofeigenvaluedistributionsforSchrödinger operators with magnetic fields Communications in Partial Differential Equa- tions.19(5-6), 719–759 (1994) [Mor24] Morin, L.:...

    1995

  5. [1987]

    and Weinstein, M

    [SW22] Shapiro, J. and Weinstein, M. I.: Tight-binding reduction and topological equivalence in strong magnetic fields Advances in Mathematics.403, 108343 (2022) [Sim83] Simon, B.: Semiclassical analysis of low lying eigenvalues. I. Non-degenerate minima : asymptotic expansions Annales de l’I.H.P. Physique théorique.38 (3), 295–308 (1983) [Sim84] Simon, B...

    2022

  6. [1988]

    [KP04] Korotyaev, E. and Pushnitski, A.: A trace formula and high-energy spectral asymptotics for the perturbed Landau Hamiltonian Journal of Functional Analysis.217(1), 221–248 (2004) [KDR24] Krejcirik, D., Duc, T. N., and Raymond, N.: Abrupt changes in the spectra of the Laplacian with constant complex magnetic field

    2004

  7. [2014]

    Rev.134, A1602–A1606 (1964) 81

    [Zak64] Zak, J.: Magnetic Translation Group Phys. Rev.134, A1602–A1606 (1964) 81

    1964

  8. [2024]

    80 [LL81] Landau, L

    arXiv:2412.17419 [math.SP]. 80 [LL81] Landau, L. D. and Lifshitz, E.: Quantum Mechanics Non-Relativistic Theory, ThirdEdition:Volume3(CourseofTheoreticalPhysics)(Vol.3).Butterworth- Heinemann,

    arXiv

Show all 9 references
  1. [2025]

    [FSW25b] Fefferman, C

    arXiv:2506.12650 [math-ph]. [FSW25b] Fefferman, C. L., Shapiro, J., and Weinstein, M. I.: Magnetic Double-Wells: Absence of Tunneling

    arXiv

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.