On a linear equation arising in the study of phase separation of Bose-Einstein condensates

Christos Sourdis

REVIEW 2 major objections 2 minor 12 references

On a linear equation arising in the study of phase separation of Bose-Einstein condensates

T0 review · 2 major / 2 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 The linearized BEC phase-separation operator has inverse norm growing exactly like $1/\omega$, with a counterexample proving the rate optimal.

desk verdict Solid technical correction with sharp counterexample and optimal omega^{-1} estimate; the imported nondegeneracy theorem needs precise statement before acceptance. read the letter →

arxiv 2506.05299 v2 pith:YLRFR7MJ submitted 2025-06-05 math.AP

Christos Sourdis This is my paper · ORCID

classification math.AP MSC 35B2535J4735Q5534B05

keywords Bose-EinsteincondensatesphaseseparationGross-PitaevskiisystemlinearizedoperatorinvertibilityestimatesmalldivisorssingularperturbationFourierdecomposition

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 studies the linearized operator that arises after Fourier decomposition in the study of phase separation of two-component Bose-Einstein condensates. The main result is a sharp a priori estimate for the Dirichlet problem $L_\omega\phi=g$ on $(-R,R)$: for large $R$, small frequency $0<\omega\ll1$, and $\omega R$ large, the solution satisfies $\|\phi\|_{C^0}\le C\omega^{-1}\|g\|_{C^0_\theta}$. The author also constructs an approximate kernel element whose residual is $O(\omega)$, proving that the $1/\omega$ rate is optimal and that the previously proposed uniform-in-$\omega$ estimate fails. If true, this gives the correct small-divisor control for the linearized inner problem without orthogonality conditions, exactly what gluing constructions need.

What carries the argument

The argument rests on a three-zone decomposition of $(-R,R)$: an inner zone $|x|<M$ where the limit operator $L_0$ acts and the nondegenerate kernel spanned by $(V_1',V_2')$ determines the limit; an outer zone $\omega^{-\beta}<|x|<R$ where the system decouples into scalar equations $-\phi''+\omega^2\phi=0$ with explicit hyperbolic-sine profiles; and an intermediate zone $M<|x|<\omega^{-\beta}$ where the author exchanges information by testing the system against $(V_1',V_2')$ and integrating by parts, using the self-adjointness of $L_\omega$. This yields a scalar first-order ODE for the quotient $\psi_1/V_1'$, connecting the inner limit to the outer amplitudes. A second integration over the symmetric interval $(-\omega^{-\beta},\omega^{-\beta})$ ties the two outer profiles together and forces the common amplitude $\lambda$ to vanish, giving the contradiction. The approximate kernel of Proposition 2.1 is built by matching the inner solution $(V_1',V_2')$ to the outer hyperbolic profiles through a cutoff, which produces a residual $O(\omega)$ and hence the sharpness.

What would settle it

Compute the smallest eigenvalue of the Dirichlet realization of $L_\omega$ on a large interval with $\omega R$ large: Theorem 1.1 implies it stays bounded below by $c\omega^2$, while the approximate kernel of Proposition 2.1 implies it is at most $C\omega^2$; a numerical value outside this two-sided power law would falsify the claimed sharp rate. Alternatively, exhibit any bounded nonzero solution of $L_0\phi=0$ on $\mathbb{R}$ not proportional to $(V_1',V_2')$, which would invalidate the nondegeneracy premise.

Watch

Extended reading notes

Core claim

The central claim, Theorem 1.1, is that for any $\theta>0$ there exist constants $C,R_0,\omega_0,P_0>0$ such that every solution of $L_\omega\phi=g$ in $(-R,R)$ with Dirichlet boundary conditions, $R>R_0$, $0<\omega<\omega_0$ and $\omega R>P_0$, satisfies $\|\phi\|_{C^0((-R,R))}\le C\omega^{-1}\|g\|_{C^0_\theta((-R,R))}$, where $\|g\|_{C^0_\theta}=\|g\cosh(\theta x)\|_{C^0}$. The proof is by contradiction: assuming bounded solutions with $\omega^{-1}\|g\|_{C^0_\theta}\to0$, it splits the interval into inner, intermediate, and outer zones, shows the solution is a constant multiple $\lambda(V_1',V_2')$ in the inner zone, matches that to explicit hyperbolic-sine profiles in the outer zones, and then uses the self-adjoint testing against $(V_1',V_2')$ to force $\lambda=0$ and finally uniform vanishing of both components. The counterexample of Proposition 2.1, built by matching $(V_1',V_2')$ to the outer profiles through a cutoff at scale $\ln R$ with $\omega=R^{-\theta}$, has residual $O(\omega)$ while converging to the nontrivial profile in $C_{\rm loc}(\mathbb{R})$; hence the estimate is optimal and (1.4) fails.

Load-bearing premise

The proof assumes that the limit operator $L_0$ on the whole line has no kernel elements except multiples of $(V_1',V_2')$; if a second independent kernel mode existed, the inner limit at (3.5) could contain extra components and the exchange estimates forcing $\lambda=0$ would not close.

Editorial extensions

If this is right

For Fourier modes with small nonzero $\omega$, the inverse of the linearized operator has norm $O(\omega^{-1})$ on $(-R,R)$, so the small divisors in the strip problem are controlled by the frequency itself.
The rate $1/\omega$ is optimal: the residual of the approximate kernel of Proposition 2.1 is $O(\omega)$, so no estimate with a slower blow-up can hold uniformly.
Together with the complementary bound for $\omega R\le P_0$, one obtains an invertibility estimate valid across the full range of frequencies.
Adding the single orthogonality condition (3.35) upgrades the estimate to $O(1)$, so the entire $\omega^{-1}$ growth is due to the component of $g$ along the kernel direction.
The proof, unlike the $\omega=0$ case, does not need the second kernel element $(xV_1'+V_1,xV_2'+V_2)$; only the translation derivative $(V_1',V_2')$ is used.

Reading between the lines

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

Read as a spectral statement, the rate says the Dirichlet $L_\omega$ on $(-R,R)$ has no small singular value below $c\omega$; the approximate kernel then shows this is attained, so the linearized interface problem has exactly one soft mode whose cost vanishes linearly with $\omega$.
This suggests that in the full strip PDE, Fourier modes with small nonzero eigenvalue $\lambda_k$ will have resolvent bounds of order $\omega^{-1}=\epsilon^{-1}\lambda_k^{-1/2}$, which is mild enough for Lyapunov–Schmidt reductions; the author leaves that application implicit.
A possible testable extension is whether the estimate holds uniformly for the periodic boundary version of (1.3), since the Fourier decomposition in the strip introduces periodic conditions in the tangential variable; the proof's outer-zone analysis would need only minor changes.

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.

Desk Editor's Note

read the letter

Short version: this is a legitimate correction to the literature, with a sharp counterexample and a matching optimal estimate. The proof is coherent; the only fragile input is an imported nondegeneracy theorem that should be stated precisely. I'd send it to review.

The paper does two real things. First, it constructs an approximate kernel element for L_omega in the regime 0>1: a matched solution near (V1',V2') that satisfies Dirichlet conditions and gives a right-hand side of size O(omega) in the weighted norm, while the solution itself has a nonzero limit at 0. This kills the uniform estimate claimed in [7, Sec. 2.3]. Second, Theorem 1.1 proves the optimal bound ||phi|| <= C omega^{-1} ||g||_{C^0_theta}, which is exactly the small-divisor rate you need for Lyapunov-Schmidt reductions. The contradiction proof is clean: inner convergence to lambda V', outer representations, an exchange step using self-adjoint testing against (V1',V2'), and a final barrier argument. The estimates check out, and the counterexample shows sharpness.

The soft spot is the imported nondegeneracy from [2, Thm 1.3]. The paper uses it to infer that the bounded inner limit is lambda V', but the theorem is not stated precisely. Because L0 also has the unbounded kernel element W=(xV1'+V1, xV2'+V2), the needed statement is not just 'kernel is spanned by V'' but 'every bounded solution of L0 psi=0 is a multiple of V''. This is likely true, but the author should quote the theorem exactly and verify this bounded version. The Gaussian asymptotics of (1.2) are also imported, but those only control exponentially small remainders and are not structurally critical. The remark about extending to the whole line is plausible but unproven, which is fine given the context.

Citation pattern is fine: [1], [4], [7] are the directly relevant prior results, and the paper situates itself accurately. No fitted parameters, no data; it's a purely analytical result.

Who's the audience: people working on strong-repulsion BEC systems, singular perturbation reductions, and small-divisor estimates. A serious referee should engage with it; the main revision is to pin down the imported nondegeneracy statement. I would accept it with minor revisions.

Referee Report

2 major / 2 minor

Summary. The paper studies the linearized operator Lω of the one-dimensional inner problem for phase-separating Bose-Einstein condensates on (-R,R) with Dirichlet data, where Lω = L0 + ω². The main result, Theorem 1.1, asserts that for fixed θ>0 and for R>R0, 0<ω<ω0, ωR>P0, every solution satisfies the a priori estimate ∥φ∥_{C0} ≤ Cω^{-1}∥g∥_{C0θ}. Section 2 constructs an approximate kernel element to show that the 1/ω rate is sharp, disproving the previously claimed property (1.4); Section 3 proves Theorem 1.1 by contradiction, combining an inner-zone limit, explicit outer-zone hyperbolic representations, a testing argument in the intermediate zone that exchanges information between inner and outer regions, and a final barrier argument.

Significance. If correct, Theorem 1.1 gives the optimal small-divisor bound for the Fourier modes of the linearized inner problem, with no orthogonality assumption; the companion estimate in Remark 3.1 under one orthogonality condition is a useful strengthening. The proof strategy is original: it avoids the second kernel element in (1.7) and uses a problem-specific exchange of information based on self-adjointness. The paper is self-contained except for the standard nondegeneracy theorem imported from [2]. The linear analysis is clean and the constants are independent of R and ω in the stated regime. However, the counterexample in Section 2, which is the basis of the sharpness claim, is not valid as written because the pieced-together function has a derivative jump at x=0; this is a load-bearing gap that must be repaired.

major comments (2)

[Section 2, Eqs. (2.1)-(2.2)] The function defined by (2.1)-(2.2) is continuous at x=0 but not C^1. For x>0 one has φ1'(0+)=V1''(0)-Aω coth(ωR) and φ2'(0+)=V2''(0), while for x<0 the extension gives φ1'(0-)=φ2'(0+) and φ2'(0-)=φ1'(0+). Since V1''(0)=V2''(0)=1, the jump vector is of size O(ω). Consequently -φ'' contains a Dirac mass of order ω at the origin, so Lωφ is not a continuous function and its C0θ norm is infinite. This invalidates Proposition 2.1 as stated. The construction can likely be repaired by smoothing the corner over an O(1) interval or by mollification, which would leave the residual O(ω) in C0θ, but this repair is not present in the manuscript.
[Section 3, Eq. (3.5)] The step φ∞≡λV′ imports the nondegeneracy of V from [2, Thm. 1.3] without stating the precise statement. Since (1.7) shows that the unbounded scaling mode W=(xV1'+V1, xV2'+V2) lies in the kernel of L0, the needed fact is that every bounded solution of L0ψ=0 on R is a multiple of V′. Please state this bounded-kernel version explicitly, or give a one-line derivation from [2, Thm. 1.3], so that the exclusion of W is transparent.

minor comments (2)

[Section 3, Eqs. (3.14)-(3.15)] The notation O_{C1}(e^{-c/ω^β}) is not defined; either write O_{C^1}(e^{-c/ω^β}) or state explicitly that the error is controlled in the C^1 norm.
[Section 2, Proposition 2.1] After the corner issue is repaired, it would be helpful to state explicitly why the construction contradicts (1.4): φ(0) converges to the positive number V1'(0) while the weighted C0θ norm of the right-hand side Lωφ tends to zero.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: Theorem 1.1 is proved by a self-contained contradiction argument with no fitted inputs; the only load-bearing imported result is the independent external nondegeneracy theorem of Berestycki–Lin–Wei–Zhao.

full rationale

The main theorem (1.11) is established by a self-contained contradiction proof. Assuming sequences with ||phi_n||_{C0} = 1 and omega_n^{-1}||g_n|| -> 0, the paper derives lambda = 0 through the inner-zone limit (3.5), the intermediate-zone identity (3.19)-(3.28), the exchange formulas (3.30)-(3.31), and the outer connection (3.33), then obtains uniform vanishing by a barrier argument, contradicting the normalization. No parameter is fitted to force (1.11); the estimate follows from the contradiction, and the sharpness counterexample (Proposition 2.1) is an independent lower bound, so it cannot make the upper bound circular. The cited self-works are not load-bearing: [1] is invoked only 'as a guideline' and all of its used estimates are rederived in-line; [4] is cited only for the complementary regime omega R <= P0, which lies outside the hypotheses of Theorem 1.1. The structurally fragile input is the nondegeneracy of V at (3.5), imported from [2, Thm. 1.3]; the precise bounded-kernel form needed is not restated, and the scaling kernel element W noted in (1.7) shows the statement is nontrivial. However, [2] is external, published, independent prior work by different authors, so this is a legitimate mathematical dependency rather than circularity. The derivation is input-free with respect to its own conclusion, and the paper is self-contained against external benchmarks, giving a low circularity score of 1 for the minor non-load-bearing self-citation in the remark on [4].

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

The central claim rests on no fitted numbers. It imports the existence, uniqueness, and asymptotics of the inner profile from [2,3], and the nondegeneracy of the linearized operator from [2, Thm. 1.3]. These are prior theorems, not assumptions invented for this paper. No new physical entities are postulated.

assumptions (3)

domain assumption The one-dimensional inner solution (V1,V2) exists, is unique under the stated normalization, and satisfies the asymptotics in (1.2), including V1(x)=Ax+B+O(e^{-cx²}) as x→+∞ and V2(x)=O(e^{-cx²}) as x→+∞.
Invoked throughout the proof; all weighted estimates for the outer zone rely on the Gaussian decay of V2 and linear growth of V1. Cited to [2] and [3].
domain assumption The limiting linearized operator L0 has a one-dimensional kernel, spanned by (V1',V2').
Imported from [2, Thm. 1.3]. It is used at (3.5) to identify the inner-zone limit as λV', which is the pivot of the exchange argument.
standard math Standard scalar elliptic estimates and the maximum principle apply to the componentwise ODE in the outer zone.
Used for local C^1 compactness and for (3.9), (3.10), and the final barrier estimate; no proof is given, which is standard.

how reviews work

0 comments

Cite this review

Pith. "Pith review of On a linear equation arising in the study of phase separation of Bose-Einstein condensates." pith.science (2026). https://pith.science/paper/YLRFR7MJ

@misc{pith2026250605299,
  author       = {Pith},
  title        = {Pith review of: On a linear equation arising in the study of phase separation of Bose-Einstein condensates},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YLRFR7MJ}},
  note         = {Machine review of arXiv:2506.05299}
}

read the original abstract

We consider the inner limit system describing the phase separation in two-component Bose-Einstein condensates linearized around the one-dimensional solution in an infinite strip with zero and periodic boundary conditions, and obtain optimal invertibility estimates for the Fourier modes without necessarily assuming orthogonality conditions.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

Aftalionand C

A. Aftalionand C. Sourdis, Interface layer of a two-component Bose-Einstein conden- sate, Commun. Contemp. Math. 19 (2017), 1650052

work page 2017
[4]

Casteras and C

J.-B. Casteras and C. Sourdis, Construction of a solution for the two-component ra- dial Gross-Pitaevskii system with a large coupling parameter , J. Funct. Anal. 279 (2020), 108674

work page 2020
[7]

Kowalczyk, A

M. Kowalczyk, A. Pistoiaand G. V aira, Phase separating solutions for two component systems in general planar domains , Calc. Var. 62, 142 (2023)

work page 2023
[2]

Berestycki, T.-C

H. Berestycki, T.-C. Lin, J. Weiand C. Zhao, On phase-separation models: asymp- totics and qualitative properties , Arch. Ration. Mech. Anal. 208 (2013), 163–200

work page 2013
[3]

Berestycki, S

H. Berestycki, S. Terracini, K. W angand J. Wei, On entire solutions of an elliptic system modeling phase separations , Adv. Math. 243 (2013), 102–126

work page 2013
[5]

D. G. De Figueiredoand E. Mitideri, Maximum principles for linear elliptic systems, In: Costa, D. (ed.) Djairo G. de Figueiredo-Selected Papers, p. 2014. Springer, Cham (1990)

work page 1990
[6]

Kowalczyk, A

M. Kowalczyk, A. Pistoia and G. V aira, Maximal solution of the Liouville equation in doubly connected domains , J. Funct. Anal. 277 (2019), 2997–3050

work page 2019
[8]

Kuehn, Multiple time scale dynamics, Applied Mathematical Sciences, Vol

C. Kuehn, Multiple time scale dynamics, Applied Mathematical Sciences, Vol. 191, Springer-Basel, 2015

work page 2015

Show all 12 references

[9]

P. D. Miller, Applied asymptotic analysis, American Mathematical Society, 2006

2006
[10]

Noris, H

B. Noris, H. Tavares, S. Terraciniand G. Verzini, Uniform H¨ older bounds for non- linear Schr¨ odinger systems with strong competition, Comm. Pure Appl. Math. 63 (2010), 267–302

2010
[11]

Soave and A

N. Soave and A. Zilio, On phase separation in systems of coupled elliptic equations: asymptotic analysis and geometric aspects , Ann. Inst. H. Poincar´ e C Anal. Non Lin´ eaire 34 (2017), 625–654

2017
[12]

Tavaresand S

H. Tavaresand S. Terracini, Regularity of the nodal set of segregated critical configu- rations under a weak reflection law , Calc. Var. 45 (2012), 273–317. General Lyceum of Moires, Crete. Email address : sourdis@uoc.gr

2012

Pith tools

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

[1] [1]

Aftalionand C

A. Aftalionand C. Sourdis, Interface layer of a two-component Bose-Einstein conden- sate, Commun. Contemp. Math. 19 (2017), 1650052

work page 2017

[2] [4]

Casteras and C

J.-B. Casteras and C. Sourdis, Construction of a solution for the two-component ra- dial Gross-Pitaevskii system with a large coupling parameter , J. Funct. Anal. 279 (2020), 108674

work page 2020

[3] [7]

Kowalczyk, A

M. Kowalczyk, A. Pistoiaand G. V aira, Phase separating solutions for two component systems in general planar domains , Calc. Var. 62, 142 (2023)

work page 2023

[4] [2]

Berestycki, T.-C

H. Berestycki, T.-C. Lin, J. Weiand C. Zhao, On phase-separation models: asymp- totics and qualitative properties , Arch. Ration. Mech. Anal. 208 (2013), 163–200

work page 2013

[5] [3]

Berestycki, S

H. Berestycki, S. Terracini, K. W angand J. Wei, On entire solutions of an elliptic system modeling phase separations , Adv. Math. 243 (2013), 102–126

work page 2013

[6] [5]

D. G. De Figueiredoand E. Mitideri, Maximum principles for linear elliptic systems, In: Costa, D. (ed.) Djairo G. de Figueiredo-Selected Papers, p. 2014. Springer, Cham (1990)

work page 1990

[7] [6]

Kowalczyk, A

M. Kowalczyk, A. Pistoia and G. V aira, Maximal solution of the Liouville equation in doubly connected domains , J. Funct. Anal. 277 (2019), 2997–3050

work page 2019

[8] [8]

Kuehn, Multiple time scale dynamics, Applied Mathematical Sciences, Vol

C. Kuehn, Multiple time scale dynamics, Applied Mathematical Sciences, Vol. 191, Springer-Basel, 2015

work page 2015