Curvature-induced bound states in quantum layers over non-parabolic hypersurfaces in dimensions four and higher

A quantum layer is the tubular neighbourhood of fixed half-width $a$ around a complete hypersurface $\Sigma$ embedded in Euclidean space, equipped with the Dirichlet Laplacian. For a flat layer the spectrum is purely essential and starts at the transverse threshold $(\pi/(2a))^2$; curvature of $\Sigma$ can push the spectral bottom below this threshold, producing discrete eigenvalues (bound states), the lowest of which is the ground state. For layers over surfaces in $\mathbb{R}^3$ the existence of such curvature-induced bound states is now known in wide generality, and Lin and Lu proved existence for hypersurfaces in $\mathbb{R}^{n+1}$, $n\ge 2$, under parabolicity of $\Sigma$ and non-positive total quadratic curvature. Mao (2013) proved the first result for a non-parabolic reference hypersurface, but only for spherically symmetric hypersurfaces in $\mathbb{R}^4$ with integrable curvature combination $\mathscr{K}_2$ of non-positive total and a curvature-sup-norm width bound $a<\rho_m$. The open problem is to extend the existence of the ground state to non-parabolic reference hypersurfaces beyond this symmetric four-dimensional class — to hypersurfaces in $\mathbb{R}^4$ without spherical symmetry and to all ambient dimensions $m\ge 5$ — and to determine which of the geometric hypotheses (the sign condition $\int_\Sigma\mathscr{K}_{m-2}\,d\Sigma\le 0$, the thinness restriction $a<\rho_m$) are necessary for curvature-induced binding.

Background

Confining a quantum particle between two hard walls a distance $2a$ apart is modelled by the Dirichlet Laplacian in an unbounded planar strip; its spectrum is $[\pi^2/(2a)^2,\infty)$, built from the transverse ground mode of energy $(\pi/(2a))^2$ plus free longitudinal motion, with no discrete eigenvalues. A quantum layer is the geometric generalization in which the walls follow a complete, non-compact hypersurface $\Sigma$ embedded in $\mathbb{R}^m$: the configuration space is the layer $\Omega_a=\{x+u\,\vec n(x):\ x\in\Sigma,\ |u|<a\}$ of half-width $a$, where $\vec n$ is a unit normal field, and the Hamiltonian is the Dirichlet Laplacian $-\Delta_{\Omega}^{D}$. The geometry of $\Sigma$ is encoded in its second fundamental form $A$ (the shape operator whose eigenvalues are the principal curvatures). When $\Sigma$ is asymptotically flat, i.e. $\lVert A\rVert\to 0$ at infinity, the essential spectrum still begins at the planar threshold $(\pi/(2a))^2$, but curvature can lower the bottom of the spectrum below it, producing discrete eigenvalues called bound states; the lowest one is the ground state. Writing $\sigma_0$ and $\sigma_{\mathrm{ess}}$ for the variational infima of the spectrum and of the essential spectrum, the target inequality is $\sigma_0<(\pi/(2a))^2\le\sigma_{\mathrm{ess}}$.

The phenomenon was established for layers over non-planar surfaces in $\mathbb{R}^3$ by Duclos, Exner and Krejčiřík, who required (among alternative sufficient conditions) the total Gauss curvature to satisfy $\int_\Sigma\kappa\,d\Sigma\le 0$, together with asymptotic flatness and a thin-layer restriction; Carron, Exner and Krejčiřík extended this to topologically nontrivial layers and weakened the hypotheses. Lin and Lu generalized the theory to hypersurfaces in $\mathbb{R}^{n+1}$, $n\ge 2$ (and to quantum tubes of higher codimension): if $\Sigma$ is parabolic — meaning it admits no positive Green's function, equivalently Brownian motion on $\Sigma$ is recurrent — and asymptotically flat, and the quadratic curvature combination $\sum_k\mu_{2k}\operatorname{Tr}(R^k)$ built from the curvature tensor is integrable with non-positive total, then the ground state exists provided $a\lVert A\rVert<C_0<1$. For $n=3$ (layers in $\mathbb{R}^4$) their condition reads: the scalar curvature $\rho$ is integrable and $\int_\Sigma\rho\,d\Sigma\le 0$. They posed explicitly the question of which assumptions on $\Sigma$ are actually needed. In the opposite direction, Lu and Rowlett settled large parts of the $\mathbb{R}^3$ case of the quantum layer conjecture: for surfaces with integrable Gauss curvature, the discrete spectrum is non-empty for thin layers over weakly $\kappa$-parabolic surfaces, over surfaces of non-negative Gauss curvature, and over parabolic surfaces, so the sign condition $\int_\Sigma\kappa\,d\Sigma\le 0$ is unnecessary in $\mathbb{R}^3$. These two-dimensional arguments rest on facts special to surfaces (Huber's conformal classification of finite-total-curvature surfaces, Hartman's description of the ends, White's total curvature estimates) and have no known higher-dimensional analogue.

Mao's paper enters here. A spherically symmetric hypersurface with a pole is a hypersurface of revolution $\Sigma\subset\mathbb{R}^m$ carrying geodesic polar coordinates with warped-product metric $ds^2+r(s)^2|d\xi|^2$ over the unit sphere $\mathbb{S}^{m-2}$, with principal curvatures $k_s$ and $k_\theta$. For $m=4$ Mao proves: if the curvature combination $\mathscr{K}_2$ — a fixed positive multiple of the scalar curvature of the three-dimensional $\Sigma$, built from the even elementary symmetric functions of the principal curvatures with the coefficients $\eta_{2k}$ of the transverse eigenfunction expansion — is integrable, if $\int_\Sigma\mathscr{K}_2\,d\Sigma\le 0$, and if $\Sigma$ is not a hyperplane and satisfies assumptions A1 (the layer map is injective, so the layer is embedded), A2 (the half-width satisfies $a<\rho_m:=(\max\{\lVert k_s\rVert_\infty,\lVert k_\theta\rVert_\infty\})^{-1}$) and A3 (asymptotic flatness), then the ground state of the layer exists. The novel point is a lemma showing that integrability of $\mathscr{K}_2$ forces $\Sigma$ to be non-parabolic, with Euclidean-type (cubic) volume growth, so the result is genuinely complementary to the parabolic theory of Lin and Lu; indeed Mao's stated motivation is the question whether parabolicity of the reference submanifold is necessary for the existence of the discrete spectrum. What remains open is the extension of this non-parabolic existence theory beyond the symmetric four-dimensional class, and the necessity of the remaining geometric hypotheses.

Problem Statement

Let $m\ge 4$ and let $\Sigma\subset\mathbb{R}^m$ be a complete, connected, oriented, non-totally-geodesic $C^2$ hypersurface that is asymptotically flat ($\lVert A\rVert\to 0$ at infinity, with $A$ the second fundamental form) and non-parabolic (it admits a positive Green's function), and let $\Omega_a=\{x+u\,\vec n(x):\ x\in\Sigma,\ |u|<a\}$ be an embedded quantum layer of half-width $a<\rho_m$, where $\vec n$ is a smooth unit normal field on $\Sigma$ and $\rho_m$ is the curvature-sup-norm thinness bound of the reference theory — in Mao's rotationally symmetric setting $\rho_m=(\max\{\lVert k_s\rVert_\infty,\lVert k_\theta\rVert_\infty\})^{-1}$ with $k_s$ and $k_\theta$ the principal curvatures of $\Sigma$, and in general a sup-norm curvature bound of the type appearing in Lin and Lu's condition $a\lVert A\rVert<C_0<1$; embeddedness of the layer, i.e. injectivity of the normal map $(x,u)\mapsto x+u\,\vec n(x)$ on $\Sigma\times(-a,a)$, is a separate standing assumption (Mao's A1). By Krejčiřík and Lu, $\sigma_{\mathrm{ess}}(-\Delta_{\Omega_a}^{D})=[\pi^2/(2a)^2,\infty)$. The problem left open by Mao's theorem and by Lin and Lu's general question is:

(1) Existence beyond the symmetric class. Under Mao's curvature hypotheses — $\mathscr{K}_{m-2}$ integrable with $\int_\Sigma\mathscr{K}_{m-2}\,d\Sigma\le 0$, where $\mathscr{K}_{m-2}=\sum_{k=1}^{[(m-1)/2]}\eta_{2k}c_{2k}(A)$ is the curvature combination built from the even elementary symmetric functions $c_{2k}$ of the principal curvatures with the transverse-expansion coefficients $\eta_{2k}$ as defined in Lin and Lu (2007) and Mao (2013) — does the inequality $\sigma_0(-\Delta_{\Omega_a}^{D})<\pi^2/(2a)^2$ hold for non-parabolic reference hypersurfaces outside Mao's class, in each of the two cases Mao leaves open: (1a) in $\mathbb{R}^4$ without spherical symmetry, and (1b) in $\mathbb{R}^m$ with $m\ge 5$? Under precisely which hypotheses on $\Sigma$ does an asymptotically flat, curved layer over a non-parabolic hypersurface have a ground state? Resolving (1) requires both existence statements, (1a) and (1b), together with the general characterization of the sufficient hypotheses.

(2) Necessity of the hypotheses. Is the sign condition $\int_\Sigma\mathscr{K}_{m-2}\,d\Sigma\le 0$ necessary for the existence of the ground state when $m\ge 4$? Can the thinness restriction $a<\rho_m$ be weakened while keeping the layer embedded — replaced by a geometric condition on $\Sigma$ that allows widths at which the curvature-sup-norm bound fails but the normal exponential map is still injective? These may be answered one hypothesis at a time.

References

1. Jing Mao, A class of rotationally symmetric quantum layers of dimension 4, Journal of Mathematical Analysis and Applications 397(2):791–799, 2013. DOI: 10.1016/j.jmaa.2012.08.012; arXiv:1203.5465; https://arxiv.org/abs/1203.5465
2. P. Duclos, P. Exner, D. Krejčiřík, Bound states in curved quantum layers, Communications in Mathematical Physics 223(1):13–28, 2001. DOI: 10.1007/PL00005582; https://doi.org/10.1007/PL00005582
3. G. Carron, P. Exner, D. Krejčiřík, Topologically nontrivial quantum layers, Journal of Mathematical Physics 45(2):774–784, 2004. DOI: 10.1063/1.1635998; https://doi.org/10.1063/1.1635998
4. Christopher Lin, Zhiqin Lu, Existence of bound states for layers built over hypersurfaces in $\mathbb{R}^{n+1}$, Journal of Functional Analysis 244(1):1–25, 2007. DOI: 10.1016/j.jfa.2006.11.017; https://doi.org/10.1016/j.jfa.2006.11.017
5. Christopher Lin, Zhiqin Lu, On the discrete spectrum of generalized quantum tubes, Communications in Partial Differential Equations 31(10):1529–1546, 2006. DOI: 10.1080/03605300600635111; https://doi.org/10.1080/03605300600635111
6. Zhiqin Lu, Julie Rowlett, On the discrete spectrum of quantum layers, Journal of Mathematical Physics 53(7):073519, 2012. DOI: 10.1063/1.4736412; arXiv:1110.6807; https://arxiv.org/abs/1110.6807
7. David Krejčiřík, Zhiqin Lu, Location of the essential spectrum in curved quantum layers, Journal of Mathematical Physics 55(8):083520, 2014. DOI: 10.1063/1.4893035; arXiv:1211.2541; https://arxiv.org/abs/1211.2541
8. F. L. Bakharev, A. I. Nazarov, Existence of the discrete spectrum in the Fichera layers and crosses of arbitrary dimension, Journal of Functional Analysis 281(4):109071, 2021. DOI: 10.1016/j.jfa.2021.109071; arXiv:2006.12045; https://arxiv.org/abs/2006.12045
9. F. Bakharev, S. Matveenko, Spectra of the Dirichlet Laplacian in 3-dimensional polyhedral layers, St. Petersburg Mathematical Journal 35(4):597–610, 2024. DOI: 10.1090/spmj/1818; arXiv:2304.10575; https://arxiv.org/abs/2304.10575
10. David Krejčiřík, Jan Kříž, Bound states in soft quantum layers, Publications of the Research Institute for Mathematical Sciences 60(4):741–766, 2024. DOI: 10.4171/prims/60-4-4; arXiv:2205.04919; https://arxiv.org/abs/2205.04919
11. Pedro Freitas, David Krejčiřík, A lower bound to the spectral threshold in curved quantum layers, in Functional Analysis and Operator Theory for Quantum Physics (The Pavel Exner Anniversary Volume), EMS Series of Congress Reports, European Mathematical Society, 261–269, 2017. DOI: 10.4171/175-1/14; https://doi.org/10.4171/175-1/14
12. Zhiqin Lu, On the ground state of Quantum Layers, lecture at the AIM workshop Shape optimization with surface interactions, June 2019. https://lu.math.uci.edu/pdfs/talks/2019-AIM/2019-aim.pdf