Non-singular hotspots between closely spaced high-index nanoparticles

Bryn Davies; Konstantinos Alexopoulos; Pierre Millien

REVIEW 3 major objections 7 minor 31 references

Non-singular hotspots between closely spaced high-index nanoparticles

T0 review · 3 major / 7 minor · reviewed 2026-07-31 · grok-4.5

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

Pith's one-line read Gradient hotspots between high-index dielectric particles grow like 1 over gap size then saturate, from modal contrast under weak coupling rather than a true singularity.

desk verdict Clean explanation of pre-asymptotic gap hotspots in the resonant high-index scalar regime; the conditional theorem is sound, but the proved C0 bound does not by itself certify the contrast-survival hypothesis. read the letter →

arxiv 2607.24204 v1 pith:J3P4SD7M submitted 2026-07-27 math.AP math-phmath.MPmath.SPphysics.optics

Konstantinos Alexopoulos , Bryn Davies , Pierre Millien This is my paper · ORCID

classification math.APmath-phmath.MPmath.SPphysics.optics MSC 35J0535P1578A45

keywords gradienthotspotshigh-indexdielectricsweakcouplingNewtonianpotentialspectralperturbationmean-valuemechanismsubwavelengthresonatorsHelmholtztransmission

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

When two high-index dielectric nanoparticles nearly touch, experiments and simulations show a strong field gradient in the gap. Because the material contrast sits only in a lower-order term of the Helmholtz equation, the gradient cannot blow up as the gap closes; it stays bounded. This paper explains the intermediate amplification that is still observed: over a window of separations the particles are weakly coupled, so the pair modes are small perturbations of the isolated-particle modes. If an isolated mode keeps a nonzero value contrast between the two facing boundaries, a mean-value argument forces the gradient across a gap of width kappa to scale like 1/kappa. The argument needs pointwise (not merely L2) control of the spectral perturbation so the contrast survives. Once the gap is small enough that coupling is no longer weak, the modes reorganize, the contrast drops, and the growth saturates. The result reframes the hotspot as a pre-asymptotic geometric effect of modal contrast rather than a singularity of the PDE.

What carries the argument

Weak-coupling modal perturbation plus C0 lifting: the off-diagonal Newtonian interaction is treated as a small perturbation of the block-diagonal isolated operators; standard L2 spectral estimates are strengthened to uniform pointwise control so that boundary contrast survives and a mean-value theorem across the gap yields the 1/kappa lower bound (Theorem 5.1).

What would settle it

For a mode with clear facing-boundary contrast, plot the mid-gap gradient versus separation: it must track 1/kappa while the off-diagonal-to-diagonal (or spectral-gap) ratio stays much less than one, then flatten once that ratio approaches order one; absence of the intermediate 1/kappa window or continued growth after strong coupling would refute the claim.

Watch

Extended reading notes

Core claim

In the weak-coupling regime, if an isolated resonant mode has a non-vanishing contrast between the closest boundary points and the coupled mode remains uniformly close to it in C0, then the gradient of the coupled mode on the gap segment is at least order 1 over the separation kappa. The growth is therefore a mean-value consequence of persistent modal contrast; it ceases when weak coupling fails and the modes hybridize.

Load-bearing premise

The coupled mode must stay close enough in the uniform norm to the isolated mode that the original boundary contrast is not washed out; that closeness holds only while the interaction stays small relative to the spectral gap, a nontrivial intermediate window because the pointwise interaction norm does not vanish as the particles touch.

Editorial extensions

If this is right

Hotspots in this resonant dielectric setting are strong but bounded; design should target the intermediate weak-coupling window rather than the asymptotic contact limit.
Only modes that induce a nonzero facing-boundary contrast produce the 1/kappa amplification; modal selection, not geometry alone, decides whether a hotspot appears.
For identical particles the same mechanism acts inside the degenerate eigenspace via antisymmetric combinations that create a sign change across the gap.
Saturation is predicted once the interaction ceases to be perturbative, matching the observed transition from amplification to plateau.

Reading between the lines

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

The same mean-value-plus-weak-coupling picture is likely to control intermediate hotspots in other high-contrast Helmholtz or Maxwell settings where contrast sits outside the principal part.
Shape optimization of facing curvature could enlarge the contrast or widen the weak-coupling window, offering a route to stronger practical hotspots without true singularities.
Full-vector Maxwell numerics with the same diagnostics (block-norm ratio and mid-gap gradient versus scaled gap) would test whether the scalar reduction already captures the essential pre-asymptotics.

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

The one thing worth knowing: this paper explains why you still see a κ^{-1} gradient spike between high-index dielectric dimers even though the PDE cannot blow up. Contrast sits only in the zeroth-order term, so H² stays uniform; the hotspot is pre-asymptotic, from weakly coupled isolated modes that keep a facing-boundary contrast, then saturates when the off-diagonal block stops being small.

What is actually new is Theorem 5.1 for the previously open resonant scaling ε≫1, εω²∼1. Prior blow-up theory covers perfect conductors, voids, plasmons, and cases where contrast is in the principal part. Here they do the block split of the Newtonian operator, standard spectral perturbation, a direct L²→C⁰ lift, and a mean-value lower bound. That chain is clean and parameter-free. The sphere numerics (Anderson–Khavinson basis, R(κ)/S(κ), midpoint gradient vs 1/κ, eigenvalue tracks) line up with the quantities the theory defines and show the amplification-to-saturation transition. Citation pattern is honest about the open gap.

Soft spot, in proportion: the load-bearing hypothesis ‖δϕ‖_C⁰ ≤ c*/4 is not delivered with explicit constants by Corollary 4.1.2. The κ-independent term ε∞/|ρ| is O(1) from the Newtonian L²→L∞ bound and overshoots a typical c*/4 for the radial mode; Fig. 4b even rescales the predicted control. So the claim that the mean-value mechanism accounts for the observed window rests on measured pointwise error staying small, not on a closed a-priori estimate. That is a real quantitative gap, not a fatal one—the conditional implication is fine, and the numerics support the story on truncated spherical bases. Secondary limits: scalar model versus the Maxwell motivation, spheres only, no shipped code. None of that breaks the scalar result.

This is for people who work on gradient estimates in composites or mathematical nanophotonics. Worth a serious referee. I would engage with it and expect revision mainly on making the C⁰ regime explicit or leaning harder on the numerical certification of contrast survival.

Referee Report

3 major / 7 minor

Summary. The paper studies the scalar Helmholtz transmission problem for two nearly touching high-index dielectric resonators in three dimensions, in the resonant (Mie-type) scaling regime where the contrast enters only the lower-order term. Since the principal part is constant-coefficient, the gradient stays uniformly bounded as the gap κ→0; the puzzle addressed is the numerically observed pre-asymptotic regime in which the gap gradient grows like κ^{-1} before saturating. The authors formulate a weak-coupling regime via a block decomposition of the quasi-static Newtonian operator K_D^{(0)} (§3), prove standard non-degenerate and degenerate spectral perturbation estimates (Thm. 3.3, App. A.2), lift the L² perturbation control to a pointwise C⁰ estimate using the L²→C⁰ mapping property of the Newtonian potential (Prop. 4.1, Cors. 4.1.1–4.1.2), and prove the main result, Theorem 5.1: if an isolated mode has facing-boundary contrast |Φ(x2)−Φ(x1)| ≥ c* > 0 and the coupled mode remains within c*/4 in C⁰, then the mean value theorem along the gap segment yields |∇Φκ(y)| ≥ c*/(2κ) at some gap point. Numerics on a truncated Anderson–Khavinson spectral basis for two spheres (N=15 modes each) confirm the κ^{-1} window, its modal selectivity, and the saturation transition.

Significance. If the quantitative gap in Major Comment 1 is closed or honestly reframed, this is a valuable contribution to the field-concentration literature: it identifies and justifies a mechanism (weakly coupled modal contrast converted by the gap geometry into a κ^{-1} gradient) that, to my knowledge, has not been isolated before in the resonant high-index dielectric setting, and it does so with parameter-free estimates (the κ^{-1} lower bound is the mean-value theorem applied to a perturbatively stable contrast, not a fit) and concrete, checkable numerics built on the Anderson–Khavinson spectrum. The amplification/saturation picture in Figs. 8–9, together with the operator-level diagnostics R(κ) and S(κ), provides falsifiable predictions and a clear explanation of why no true blow-up occurs. The modal selectivity results (§6.6) are a nice bonus with physical content.

major comments (3)

[§5, Theorem 5.1, hypothesis (5.2) via Corollary 4.1.2, Eq. (4.6)] Theorem 5.1 is proved conditionally on hypothesis (5.2), ‖Φκ−Φ‖_{C0} ≤ c*/4, and the only analytic route to (5.2) is Corollary 4.1.2. That bound contains the κ-independent term T2 = ε∞/|ρ|, which does not shrink in any regime. With Lemma A.2 (ε∞ ≤ ½√(R/π)) and the principal radial eigenvalue of the unit ball (Sec. 6.1, j_{−1/2,1}=π/2, |ρ|=4/π²), one gets T2 ≲ π^{3/2}/8 ≈ 0.70, whereas for the L2-normalized lifted radial mode c*/4 ≈ 0.1 (c* = |f(x1)| = (2π)^{−1/2}). So the proved estimate misses the required threshold by roughly an order of magnitude even before the κ-growing terms T1, T3, T4 and the unestimated constant C are included. As written, the analysis therefore does not establish any interval of κ in which the mean-value conclusion (5.3) applies; the 'explanation' of the observed κ^{−1} window currently rests on the *measured* pointwise error, not on the proved bound. This shoul
[§6.5, Fig. 4b, and the comparison with (5.2)] This subsection is presented as numerical confirmation of the pointwise control used in Theorem 5.1, but two choices undermine that reading. First, in Fig. 4b the predicted quantity T_tot is multiplied by an ad hoc factor 0.47 to match the measured ‖Φκ−Φ‖_{L∞}; the caption acknowledges the rescaling but the text does not discuss it. Second, and more importantly, the threshold that actually matters — c*/4 in (5.2) — is never computed or plotted for any mode. The decisive diagnostic, namely measured ‖δϕ‖_{C0} versus c*/4 as a function of κ, would directly delimit the amplification window and could be correlated with the saturation onset in Figs. 8–9; without it, the claim that 'the boundary contrast used in the mean value argument persists precisely in the regime where the pointwise perturbation remains controlled' (§6.5) is not demonstrated. Please also state whether C0 and ε∞ entering T1
[§3.4–§4: constants, thresholds, and hypothesis (4.3)] The perturbation results are stated with unspecified constants C (Thm. 3.3, Cors. 4.1.1–4.1.2, Prop. A.3), the weak-coupling conditions use '≪' (Def. 3.1, (3.17)), and the absorption hypothesis |δρ|/|ρ| ≤ 1/2 in (4.3) is never verified in any regime. Since the paper's message is quantitative — a κ^{−1} law over an intermediate window bounded by two transition scales — the theory side currently provides no explicit or even order-of-magnitude delineation of that window. At minimum, the dependence of the constants on the data (spectral gap, mode, geometry) should be stated, and the numerics should indicate at which κ/δ each hypothesis ((3.11), (4.3), (5.2)) actually fails, so that the theoretically sanctioned window and the observed one can be compared.

minor comments (7)

[§3.4, §6.7] The model is a scalar Helmholtz equation (the Ez polarization in the 2D motivation), yet the text repeatedly refers to 'magnetic localization' (end of §3.4) and 'magnetic hot-spot'/'magnetic intensity' (Figs. 6–7 captions). Please clarify the physical identification or neutralize the terminology.
[§4, Eq. (4.1)] Notation: ε2 and ε∞ suggest L2 and L∞ norms, but ε2 is an L2→L2 norm and ε∞ an L2→C0 norm; the symbol C0 is used both for the norm ‖K^{(0)}_{diag}‖_{L2→C0} and for the space C0(D). Consider renaming (e.g., η2, η∞, M0) to avoid collisions.
[§3.4, Definition 3.1] Definition 3.1 defines a regime via '≪', which is not a mathematical condition; Definition 3.2 is the usable one. Either drop Def. 3.1 or state it as heuristic motivation.
[§3.5, Theorem 3.3] The proof of Theorem 3.3 is one sentence ('standard non-degenerate perturbation expansion'). Please cite a specific result (e.g., Kato, Chap. II, §2 or the analytic perturbation theorem for self-adjoint compact operators) and note that K^{(0)}_D is compact self-adjoint on L2(D) — worth one line since everything downstream depends on it.
[§6.2, Eq. (6.3)] The off-diagonal matrix entries (6.3) are six-dimensional integrals with a singular-at-coincidence kernel; the quadrature method and its accuracy as κ→0 are not described. A brief description and a convergence check in N would strengthen §6.
[§5.1] The saturation side of the story (reorganization of the coupled mode, decrease of the boundary contrast) is asserted physically in §5.1 and §7 but only observed numerically. Since the uniform boundedness of the gradient is said to be elementary, a short remark clarifying that the saturation *mechanism* is outside the proved results would help calibrate the reader.
[§6.8, Fig. 9] Fig. 9 is captioned and labeled |∇u(xmid)| but the text discusses |∂x u(xmid)|; please make these consistent. Typos: 'an elementary analysis give' (§1.2); 'these type of gradient blow up results' (§1.1); 'dimer' vs 'dimers', inconsistent hyphenation of 'hotspot/hot-spot'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: Theorem 5.1 is a conditional mean-value implication from standard spectral perturbation, not a fit or self-definitional loop.

full rationale

The load-bearing chain is: Lippmann–Schwinger/Newtonian block decomposition → weak-coupling L2 perturbation (Thm 3.3, Kato-style) → C0 lift via Newtonian L2→C0 bounds (Prop 4.1, Cor 4.1.2, App A) → if isolated facing contrast survives pointwise (hypotheses 5.1–5.2), MVT gives |∇Φκ|≥c*/(2κ) (Thm 5.1). None of these steps defines the output in terms of itself, fits a free parameter to the target κ^{-1} curve, or imports a uniqueness theorem that forces the claim. Numerics diagnose the same operator quantities the theory defines (R(κ), S(κ), T1–T4, midpoint gradient) rather than tuning constants to match a predicted hotspot. Self-citation [1] only supplies the standard integral representation and small-δ expansion; the hotspot mechanism is derived in-paper. Quantitative looseness of the C0 bound relative to c*/4 is a sharpness/correctness issue, not circularity. Derivation is self-contained and parameter-free.

Assumptions & free parameters 2 free parameters · 6 assumptions · 2 invented entities

The claim rests on standard elliptic/spectral machinery plus domain modeling choices. No fitted physical constants. The main modeling axioms are the scalar Helmholtz reduction, the quasi-static Newtonian leading term under δk0≪1, and simplicity (or explicit symmetric/antisymmetric splitting) of the isolated eigenvalue being perturbed. Invented entities are definitional regimes and diagnostics, not new physical objects.

free parameters (2)

Modal truncation N per resonator = N=15 (default)
Numerics retain N=15 Anderson–Khavinson modes per sphere to build the projected matrix KN(κ). This is a discretization cutoff, not a fitted physical constant, but hotspot amplitudes and transition locations can depend on N.
Weak-coupling smallness threshold
The regime is defined by ε2/dρ≪1 and ‖δϕ‖_C0≤c*/4. The paper does not fix a universal numerical cutoff; transition is read off where R(κ) or S(κ) approach order 1 in figures.

assumptions (6)

domain assumption Scalar Helmholtz transmission problem is an adequate model for the hotspot mechanism of high-index dielectric dimers (full Maxwell left open).
Stated in §1.1–1.2 and Conclusion; contrast enters only the zeroth-order term as in the 2D Ez reduction.
domain assumption Subwavelength expansion K_δk0 = K^(0) + O(δk0) with leading Newtonian potential governing resonant modal structure (Lemma 2.2).
Used throughout §§2–3; cited from [1].
standard math Standard non-degenerate (and degenerate) perturbation theory for compact self-adjoint operators applies to K_diag + K_off under ε2/dρ≪1 (Thm 3.3, Kato).
Section 3.5; reference [21].
standard math Newtonian potential maps L2(D)→C0(D) boundedly on bounded 3D domains, with explicit off-diagonal bounds involving min(R-bound, 1/κ) (App. A).
Lemmas A.1–A.2 and eq. (4.2); enables the C0 lift.
domain assumption Isolated eigenvalue under study is simple, or the identical-particle degeneracy is resolved into symmetric/antisymmetric combinations inside the eigenspace.
Thm 3.3 assumes simplicity; §3.3 and App. A.2 treat the identical case via 2D spectral projections.
ad hoc to paper Isolated mode has non-vanishing facing-boundary contrast |Φ(x2)−Φ(x1)|≥c*>0 at the closest points.
Hypothesis (5.1) of Thm 5.1; mode-dependent and necessary for the mean-value lower bound. Numerics show some modes violate it and produce no hotspot.

invented entities (2)

Weak-coupling regime (block-norm and spectral-gap forms) independent evidence
purpose: Delimit the separations where coupled modes remain perturbative deformations of isolated modes, enabling the contrast-persistence argument.
Defs 3.1–3.2; operationalized numerically via R(κ) and S(κ). Definitional, not a new physical particle or force.
Pre-asymptotic (non-singular) gradient hotspot mechanism independent evidence
purpose: Name the κ^{-1} amplification-then-saturation phenomenon without true blow-up.
Central interpretive claim of Thm 5.1 and §5.1; falsifiable via the numerics and, in principle, via gap-size sweeps in experiments cited in the intro.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Non-singular hotspots between closely spaced high-index nanoparticles." pith.science (2026). https://pith.science/paper/J3P4SD7M

@misc{pith2026260724204,
  author       = {Pith},
  title        = {Pith review of: Non-singular hotspots between closely spaced high-index nanoparticles},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J3P4SD7M}},
  note         = {Machine review of arXiv:2607.24204}
}

read the original abstract

We study the concentration of the field between two nearly touching high-index dielectric resonators in three dimensions. The model is a scalar Helmholtz transmission problem in the resonant regime, wherein the wavelength inside the resonators of the same order as their typical diameter. The material contrast enters only a lower-order term of the equation and not its principal part. As a consequence, the gradient of the field stays bounded independently of the distance separating the particles, and does not blow up. Nevertheless, simulations and experiments show that the gradient still concentrates in the gap: as the particles approach, it grows like the inverse of their separation over an intermediate range of distances and then saturates once they are very close. We explain this pre-asymptotic effect through a weak-coupling regime, in which the resonant modes of the pair are perturbations of the modes of each isolated particle. When such a mode keeps a nonzero contrast between the two facing boundaries, a mean value argument across the gap accounts for the growth of the gradient; this requires strengthening the standard spectral perturbation estimates from an average to a pointwise control. The growth stops once the interaction between the particles is no longer weak and the perturbation theory fails. Numerical experiments confirm the transition from amplification to saturation.

Figures

Figures reproduced from arXiv: 2607.24204 by the authors.

**Figure 2.** Gradient hotspot formation for the radial mode [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗

**Figure 3.** Weak-coupling diagnostics in three dimensions. Left: the ratio [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: Numerical verification of the pointwise control estimate. Panel [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]

**Figure 5.** Figure 5: Modal dependence of gradient hotspot formation. Each panel shows a horizontal cross-section through [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]

**Figure 6.** Figure 6: shows the first radial mode (s, m, n) = (0, 0, 0). The symmetric combination produces a smooth transition between the two resonators and only a moderate gradient in the gap. By contrast, the antisymmetric combination creates a sign change across the gap and generates a…

**Figure 7.** Figure 7: Identical resonators, higher-order mode (s, m, n) = (4, 0, 1). The antisymmetric combination again produces stronger magnetic localization in the gap. These simulations confirm that the identical-resonator case fits the same geometric amplification picture, provided th…

**Figure 8.** Figure 8: Mean value diagnostics for the weakly coupled mode [PITH_FULL_IMAGE:figures/full_fig_p017_8.png]

**Figure 9.** Figure 9: Comparison of |∇u(xmid)| for the different modes. Modes that maintain a boundary contrast exhibit inverse-gap amplification, while the mode (s, m, n) = (3, 1, 1) remains several orders of magnitude smaller. two-particle system, and the isolated-mode approximation break…

**Figure 10.** Figure 10: Tracked eigenvalue branches of the coupled operator. Solid curves show coupled eigenvalues [PITH_FULL_IMAGE:figures/full_fig_p018_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 2 linked inside Pith

[1]

Alexopoulos and B

K. Alexopoulos and B. Davies. Asymptotic analysis of subwavelength halide perovskite resonators.Partial Differential Equations and Applications, 3(4):1–28, 2022

2022
[2]

Ammari, E

H. Ammari, E. Bonnetier, F. Triki, and M. Vogelius. Elliptic estimates in composite media with smooth in- clusions: an integral equation approach. InAnnales Scientifiques de l’ ´Ecole Normale Sup´ erieure, volume 48, pages 453–495, 2015

2015
[3]

Ammari, G

H. Ammari, G. Ciraolo, H. Kang, H. Lee, and K. Yun. Spectral analysis of the neumann–poincar´ e operator and characterization of the stress concentration in anti-plane elasticity.Archive for Rational Mechanics and Analysis, 208(1):275–304, 2013

2013
[4]

Ammari, G

H. Ammari, G. Dassios, H. Kang, and M. Lim. Estimates for the electric field in the presence of adjacent perfectly conducting spheres.Quarterly of Applied Mathematics, 65(2):339–355, 2007

2007
[5]

Ammari, B

H. Ammari, B. Davies, and S. Yu. Close-to-touching acoustic subwavelength resonators: eigenfrequency separation and gradient blow-up.Multiscale Modeling & Simulation, 18(3):1299–1317, 2020

2020
[6]

Ammari, H

H. Ammari, H. Kang, and M. Lim. Gradient estimates for solutions to the conductivity problem.Mathe- matische Annalen, 332(2):277–286, 2005

2005
[7]

Ammari, B

H. Ammari, B. Li, and J. Zou. Mathematical analysis of electromagnetic scattering by dielectric nanopar- ticles with high refractive indices.Transactions of the American Mathematical Society, 376(01):39–90, 2023

2023
[8]

J. M. Anderson, D. Khavinson, and V. Lomonosov. Spectral properties of some integral operators arising in potential theory.The Quarterly Journal of Mathematics, 43(4):387–407, 1992

1992

Show all 31 references

[9]

R. M. Bakker, D. Permyakov, Y. F. Yu, D. Markovich, R. Paniagua-Dom ´ ınguez, L. Gonzaga, A. Samusev, Y. Kivshar, B. Luk’yanchuk, and A. I. Kuznetsov. Magnetic and electric hotspots with silicon nanodimers. Nano Letters, 15(3):2137–2142, 2015. 22

2015
[10]

E. S. Bao, Y. Y. Li, and B. Yin. Gradient estimates for the perfect conductivity problem.Archive for Rational Mechanics and Analysis, 193(1):195–226, 2009

2009
[11]

Bonnetier, C

E. Bonnetier, C. Dapogny, F. Triki, and H. Zhang. The plasmonic resonances of a bowtie antenna.Analysis in Theory and Applications, 35(1):85–116, 2019

2019
[12]

Bonnetier and M

E. Bonnetier and M. Vogelius. An elliptic regularity result for a composite medium with” touching” fibers of circular cross-section.SIAM Journal on Mathematical Analysis, 31(3):651–677, 2000

2000
[13]

Calandrini, A

E. Calandrini, A. Cerea, F. De Angelis, R. P. Zaccaria, and A. Toma. Magnetic hot-spot generation at optical frequencies: from plasmonic metamolecules to all-dielectric nanoclusters.Nanophotonics, 8(1):45– 62, 2018

2018
[14]

Y. Chen, H. Li, and L. Xu. Optimal gradient estimates for the perfect conductivity problem withC 1,α inclusions.Annales de l’Institut Henri Poincar´ e C, Analyse non lin´ eaire, 38(4):953–979, 2021

2021
[15]

Ciraolo and A

G. Ciraolo and A. Sciammetta. Gradient estimates for the perfect conductivity problem in anisotropic media.Journal de Math´ ematiques Pures et Appliqu´ ees, 127:268–298, 2019

2019
[16]

Y. Deng, X. Fang, and H. Liu. Gradient estimates for electric fields with multiscale inclusions in the quasi-static regime.Multiscale Modeling & Simulation, 20(2):641–656, 2022

2022
[17]

Y. Deng, Y. Hu, H. Liu, and W. Tang. Optimal estimate of field concentration between multiscale nearly- touching inclusions for 3-D Helmholtz system.arXiv preprint arXiv:2211.05609, 2022

2022 arXiv
[18]

G. H. Hardy, J. E. Littlewood, and G. P´ olya.Inequalities. Cambridge University Press, 1934

1934
[19]

Kang and S

H. Kang and S. Yu. Quantitative characterization of stress concentration in the presence of closely spaced hard inclusions in two-dimensional linear elasticity.Archive for Rational Mechanics and Analy- sis, 232(1):121–196, 2019

2019
[20]

Kang and S

H. Kang and S. Yu. Singular functions and characterizations of field concentrations: a survey.Analysis in Theory and Applications, 37(1):102–113, 2021

2021
[21]

Kato.Perturbation Theory for Linear Operators, volume 132

T. Kato.Perturbation Theory for Linear Operators, volume 132. Springer Science & Business Media, 2013

2013
[22]

Li and J

H. Li and J. Zhang. Resonant frequencies distribution for multiple closely spaced subwavelength resonators. SIAM Journal on Applied Mathematics, 85(5):2425–2457, 2025

2025
[23]

Li and L

Y. Li and L. Nirenberg. Estimates for elliptic systems from composite material.Communications on Pure and Applied Mathematics: A Journal Issued by the Courant Institute of Mathematical Sciences, 56(7):892– 925, 2003

2003
[24]

Y. Y. Li and M. Vogelius. Gradient estimates for solutions to divergence form elliptic equations with discontinuous coefficients.Archive for rational mechanics and analysis, 153(2):91–151, 2000

2000
[25]

Lim and S

M. Lim and S. Yu. Stress concentration for two nearly touching circular holes.arXiv preprint arXiv:1705.10400, 2017

2017 arXiv
[26]

Meklachi, S

T. Meklachi, S. Moskow, and J. C. Schotland. Asymptotic analysis of resonances of small volume high contrast linear and nonlinear scatterers.Journal of mathematical physics, 59(8), 2018

2018
[27]

Mirzaei and A

A. Mirzaei and A. E. Miroshnichenko. Electric and magnetic hotspots in dielectric nanowire dimers. Nanoscale, 7(14):5963–5968, 2015

2015
[28]

Moiola and E

A. Moiola and E. A. Spence. Acoustic transmission problems: wavenumber-explicit bounds and resonance- free regions.Mathematical Models and Methods in Applied Sciences, 29(02):317–354, 2019

2019
[29]

Pascale, G

M. Pascale, G. Miano, R. Tricarico, and C. Forestiere. Full-wave electromagnetic modes and hybridization in nanoparticle dimers.Scientific Reports, 9(1):14524, 2019

2019
[30]

van de Groep, T

J. van de Groep, T. Coenen, S. A. Mann, and A. Polman. Direct imaging of hybridized eigenmodes in coupled silicon nanoparticles.Optica, 3(1):93–99, 2016

2016
[31]

Yu and H

S. Yu and H. Ammari. Hybridization of singular plasmons via transformation optics.Proceedings of the National Academy of Sciences, 116(28):13785–13790, 2019. 23

2019

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.

[1] [1]

Alexopoulos and B

K. Alexopoulos and B. Davies. Asymptotic analysis of subwavelength halide perovskite resonators.Partial Differential Equations and Applications, 3(4):1–28, 2022

2022

[2] [2]

Ammari, E

H. Ammari, E. Bonnetier, F. Triki, and M. Vogelius. Elliptic estimates in composite media with smooth in- clusions: an integral equation approach. InAnnales Scientifiques de l’ ´Ecole Normale Sup´ erieure, volume 48, pages 453–495, 2015

2015

[3] [3]

Ammari, G

H. Ammari, G. Ciraolo, H. Kang, H. Lee, and K. Yun. Spectral analysis of the neumann–poincar´ e operator and characterization of the stress concentration in anti-plane elasticity.Archive for Rational Mechanics and Analysis, 208(1):275–304, 2013

2013

[4] [4]

Ammari, G

H. Ammari, G. Dassios, H. Kang, and M. Lim. Estimates for the electric field in the presence of adjacent perfectly conducting spheres.Quarterly of Applied Mathematics, 65(2):339–355, 2007

2007

[5] [5]

Ammari, B

H. Ammari, B. Davies, and S. Yu. Close-to-touching acoustic subwavelength resonators: eigenfrequency separation and gradient blow-up.Multiscale Modeling & Simulation, 18(3):1299–1317, 2020

2020

[6] [6]

Ammari, H

H. Ammari, H. Kang, and M. Lim. Gradient estimates for solutions to the conductivity problem.Mathe- matische Annalen, 332(2):277–286, 2005

2005

[7] [7]

Ammari, B

H. Ammari, B. Li, and J. Zou. Mathematical analysis of electromagnetic scattering by dielectric nanopar- ticles with high refractive indices.Transactions of the American Mathematical Society, 376(01):39–90, 2023

2023

[8] [8]

J. M. Anderson, D. Khavinson, and V. Lomonosov. Spectral properties of some integral operators arising in potential theory.The Quarterly Journal of Mathematics, 43(4):387–407, 1992

1992