The superadiabatic projectors method applied to the spectral theory of magnetic operators

L. Benedetto, C. Fermanian Kammerer, N. Raymond, É. Vacelet · 2025

Read The superadiabatic projectors method applied to the spectral theory of magnetic operators by L. Benedetto, C. Fermanian Kammerer, N. Raymond, É. Vacelet (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14388.txt.

Introduction

Lino Benedetto and co-authors present a 2025 math paper on magnetic operators. Such operators describe particles in a magnetic field, and their spectra matter in quantum theory. A classic tool builds near-perfect projectors step by step. The authors widen that tool to a general frame. They give a full account for the rank-one case, where a neat split occurs.

The team then applies the abstract results to spectral theory. They recover or improve recent results on the magnetic Laplacian in the semiclassical limit. That limit is where quantum rules shade toward classic motion. Analysts get a cleaner path to known results and some stronger ones. The paper is formal theory with no data. Its value is a shared method that may serve other operator problems beyond the magnetic case.

Catalog record

CREATORL. Benedetto, C. Fermanian Kammerer, N. Raymond, É. Vacelet
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE70 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-14388

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The superadiabatic projectors method applied to the spectral theory of magnetic operators L. Benedetto, C. Fermanian Kammerer, N. Raymond, É. Vacelet Abstract . This article deals with a generalization of the superadiabatic projectors method. In a general framework, the well-known superadiabatic projectors are constructed and accurately described in the case of rank one, when a remarkable factorization occurs. We apply these ideas to spectral theory and we explain how our abstract results allow to recover or improve recent results about the semiclassical magnetic Laplacian. 1. Introduction 1.1. Motivation and context This article deals with a generalization of the superadiabatic projectors method. In essence, this method aims at diagonalizing pseudodifferential operators H^{w} by finding approximate eigenspaces. To do so, under suitable assumptions, we build a microlocal projection \Pi^{w} that commutes microlocally with H^{w} . This leads to consider an effective operator, formally given by \Pi^{w}H^{w}\Pi^{w} , attached to the spectral window we are interested in. The projection is constructed by means of the spectral elements of the operator-valued symbol H and their quantization. In the present article, we establish a generalized version of this method (to operator-valued symbols in a non-necessarily selfadjoint context), and we apply it to spectral theory. In particular, we recover (and even improve), by elementary means, recent results about the semiclassical magnetic Laplacian in two dimensions (see [ 20 , 13 ] ). It turns out that this method is an alternative to the famous Grushin method (see the review article [ 27 ] and the concise presentation in [ 17 ] ), which has recently shown its power in the semiclassical spectral analysis of magnetic Laplacians (see the Ph.D. thesis [ 16 ] and, for instance, [ 6 , 1 ] ). The strategy that we propose in the present article simplifies a little the Grushin method in the sense that the latter relates the bijectivity of H^{w}-z to that of an effective operator a^{w}_{\mathrm{eff},z} that implicitly depends on z , whereas the microlocal projection method directly reduces to an effective operator independent of z . Hopefully, this article will inspire a new view point on the spectral analysis of magnetic Laplacians and suggest further applications. 1.2. Superadiabatic projectors We consider two Hilbert spaces \mathscr{A} and \mathscr{B} , with continuous embedding \mathscr{A}\subset\mathscr{B} , and we denote by h>0 a small semiclassical parameter. We consider the symbol class S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B}))=\{H\in C^{\infty}(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B}))\,,\forall\alpha\in\mathbb{N}^{2d}\,,\exists C>0:\\ \|\partial^{\alpha}H\|_{\mathscr{L}(\mathscr{A},\mathscr{B})}\leqslant C_{\alpha}\}\,. The operator-valued symbols can depend on h , as soon as the estimates defining the symbol class hold uniformly for h small. Of course, contrary to the scalar case, this class of symbol is not an algebra. If H\in S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B})) , the semiclassical Weyl quantization of H is given by H^{w}u(x)=\frac{1}{(2\pi h)^{n}}\iint_{\mathbb{R}^{2d}}e^{i\frac{(x-y)\cdot\xi}{h}}H\left(\frac{x+y}{2},\xi\right)u(y)\,\mathrm{d}y\mathrm{d}\xi\,, for u\in\mathscr{S}(\mathbb{R}^{d},\mathscr{A}) . For details about pseudodifferential operators with operator-valued symbols, one refers to [ 16 , Chapitre 2] . The Moyal product induced by the composition of such operators (when it is well-defined) is denoted by \circledast . In what follows, we consider a symbol H\in S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B})) which admits the following asymptotic expansion H\sim\sum_{j\in\mathbb{N}}h^{j}H_{j}. We call admissible symbol any element of S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B})) having a similar asymptotic expansion. Assumption 1.1 . We work in the following framework. (i) For all (x,\xi)\in\mathbb{R}^{2d} , the principal symbol H_{0}(x,\xi) is a selfadjoint operator on the Hilbert space \mathscr{B} with domain \mathscr{A} and we denote by \sigma(x,\xi)\subset\mathbb{R} its spectrum. (ii) There exists \delta>0 such that the following holds: for all (x,\xi)\in\mathbb{R}^{2d} , there exists a subset \sigma_{0}(x,\xi)\subset\sigma(x,\xi) of the spectrum of H_{0}(x,\xi) such that (1.1) {\rm dist}(\sigma_{0}(x,\xi),\sigma(x,\xi)\setminus\sigma_{0}(x,\xi))>\delta, and such that the corresponding spectral projector (1.2) \Pi_{0}(x,\xi)\coloneqq\mathbf{1}_{\sigma_{0}(x,\xi)}\left(H_{0}(x,\xi)\right) defines a symbol in S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{B},\mathscr{A})) . (iii) There exist m,M\in\mathbb{R} such that, for all (x,\xi)\in\mathbb{R}^{2d} , \sigma_{0}(x,\xi)\subset[m,M] . Note that \Pi_{0}(x,\xi) is bounded and thus define an operator of \mathscr{L}(\mathscr{A}) . As such, it is an element of \mathscr{L}(\mathscr{A})\cap\mathscr{L}(\mathscr{B}) . The first main result in this article is the following theorem. Theorem 1.2 . Let H\in S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathscr{B})) be an admissible operator-valued symbol and H^{w} its Weyl quantization. We assume that Assumption 1.1 holds. Then, there exists an admissible operator-valued symbol \Pi\in S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{B},\mathscr{A})) with principal symbol \Pi_{0} and unique up to a remainder of order \mathscr{O}(h^{\infty}) , satisfying \Pi^{w}\Pi^{w}=\Pi^{w}+\mathscr{O}(h^{\infty}), where the remainder \mathscr{O}(h^{\infty}) is a pseudodifferential operator whose symbol is in the same class as \Pi , and [H^{w},\Pi^{w}]=\mathscr{O}(h^{\infty})\,, where the remainder \mathscr{O}(h^{\infty}) is a pseudodifferential operator whose symbol is in the class as H . Moreover, if H is selfadjoint then \Pi is selfadjoint as well. Theorem 1.2 is part of a rather long story, which originates from the famous Born-Oppenheimer approximation for molecules [ 7 ] . On the mathematical side, this approximation method is often named adiabatic theory. In the early eighties, the adiabatic reduction for systems of Schrödinger equations has been performed by Nenciu in the context of the time-dependent Born-Oppenheimer approximation, see [ 21 , 22 ] . The reader might want to consider the monographs [ 29 ] and [ 19 ] , which give detailed accounts of more recent approaches. Many insights and fundamental ideas are also described in [ 28 , 18 , 11 ] . The idea of constructing superadiabatic projectors, that is almost projectors as quantization \Pi^{w} of matrix-valued symbols, originates from the seminal work [ 12 ] and the detailed construction in [ 3 ] . This strategy to study elliptic systems has also been used in [ 4 , 5 ] and even more recently in [ 8 , 9 ] in the case of matrix-valued symbols to obtain spectral results. Theorem 1.2 can be seen as a generalization of the above mentioned results to a more general class of (non necessarily selfadjoint) operator-valued symbols. As we will see, our proof will inspire an elegant way to derive an effective operator – the adiabatic Hamiltonian – having remarkable applications motivated by spectral theory (especially of semiclassical magnetic Laplacians), see Section 1.3 below. When \sigma_{0}(x,\xi) is made of disjoint parts, the corresponding superadiabatic projectors are orthogonal to each other. Proposition 1.3 . Let \Pi^{1} and \Pi^{2} be the symbols of superadiabatic projectors given by Theorem 1.2 and corresponding to disjoint parts of the spectrum of H_{0} that satisfy Assumption 1.1 , denoted respectively by \sigma_{0}^{1}(x,\xi) and \sigma_{0}^{2}(x,\xi) for (x,\xi)\in\mathbb{R}^{2d} . Then, (\Pi^{1})^{w} and (\Pi^{2})^{w} are orthogonal up to \mathscr{O}(h^{\infty}) in the sense that we have (\Pi^{1})^{w}(\Pi^{2})^{w}=\mathscr{O}(h^{\infty}). 1.3. Towards spectral theory Let us now turn to the applications of Theorem 1.2 . When \Pi_{0} is of rank one, we can describe the structure of the superadiabatic projector. More precisely, we will make the following assumption. Assumption 1.4 . There exists (x,\xi)\in\mathbb{R}^{2d}\mapsto u_{0}(x,\xi)\in\mathscr{B} in C^{\infty}(\mathbb{R}^{2d},\mathscr{B}) such that \forall(x,\xi)\in\mathbb{R}^{2d},\,\;\|u_{0}(x,\xi)\|_{\mathscr{B}}=1\quad\mbox{and}\quad\Pi_{0}(x,\xi)=(u_{0}(x,\xi),\cdot)_{\mathscr{B}}\,u_{0}(x,\xi). Under this assumption, the set \sigma_{0}(x,\xi) , (x,\xi)\in\mathbb{R}^{2d} , is reduced to the point given by the eigenvalue corresponding to the eigenvector u_{0}(x,\xi) . Moreover, it allows for the following factorization of the superadiabatic projectors. Theorem 1.5 . Under Assumptions 1.1 and 1.4 , there exist two symbols \ell and L in the class S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{A},\mathbb{C}))\cap S(\mathbb{R}^{2d},\mathscr{L}(\mathscr{B},\mathbb{C})) such that (1.3) L^{w}(\ell^{w})^{*}={\rm Id}_{L^{2}(\mathbb{R}^{d},\mathbb{C})}+\mathscr{O}(h^{\infty})\quad\mbox{and}\quad(\ell^{w})^{*}L^{w}=\Pi^{w}+\mathscr{O}(h^{\infty}). Moreover, if H is selfadjoint, then L=\ell . The existence of such operators as L^{w} and \ell^{w} allows us to diagonalize formally the operator H^{w} . All along the paper, we will use the notation \Pi^{\perp} for denoting the operator \Pi^{\perp}=1-\Pi for a

Read the complete text