L p L^{p} Estimates for Eigenfunctions of a Generalized Landau Magnetic Laplacian
Ben Goldschlager · 2025
Read L p L^{p} Estimates for Eigenfunctions of a Generalized Landau Magnetic Laplacian by Ben Goldschlager (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-2507-00183.txt.
Introduction
Ben Goldschlager bounds the size of wave patterns for an electron trapped in a magnetic field. His paper appeared in 2025. The operator he studies generalizes the Landau magnetic Laplacian, which models an electron on a plane under a perpendicular field. Its eigenspaces are huge, with levels matching those of the Hermite operator. The question is how large the peak values of these patterns can grow compared with their average size.
He shows the peak stays bounded by the average, whatever the energy level, under mild assumptions on the field. For the sixth-power average he proves an even stronger improvement. The proof adapts semiclassical tools through a new conjugation trick that works over the whole plane. The results steady the analysis of electrons in strong magnetic fields.
Catalog record
| CREATOR | Ben Goldschlager |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-00183
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
L p L^{p} Estimates for Eigenfunctions of a Generalized Landau Magnetic Laplacian Ben Goldschlager Abstract In this paper, we examine eigenfunctions of a generalized Landau Magnetic Laplacian that models the physics of an electron confined to a plane in a magnetic field orthogonal to the plane. This operator has an infinite dimensional null space and, at least in the model case, has infinite dimensional eigenspaces with eigenvalues which are essentially the same as the eigenvalues of the Hermite operator. We demonstrate that, under fairly general assumptions on the potential function of the magnetic field, the L ∞ L^{\infty} norm of these eigenfunctions is bounded by their L 2 L^{2} norm independently of the associated eigenvalue. We furthermore demonstrate an improvement in the L 6 L^{6} norm of these eigenfunctions. The method we use comes from semiclassical analysis and is inspired by the work of Koch, Tataru, and Zworski that applies locally. In our case, we use a new conjugation argument to demonstrate the result over all of ℝ 2 {\mathbb{R}}^{2} . Email: [email protected] ORCiD: 0009-0007-5365-1623 Keywords: Estimates for Eigenfunctions, Magnetic Laplacian, Semiclassical Analysis, Landau Levels 1 Introduction In this paper, we examine eigenfunctions of a generalized Landau magnetic Laplacian, H=D^{*}D , where D=\partial_{z}+\partial_{z}\phi,z\in{\mathbb{C}} with real-valued scalar potential function \phi . This operator arises as the Hamiltonian that describes an electron confined to a plane in a magnetic field orthogonal to that plane, see [ 1 ] and [ 2 ] for further background. In a suitable gauge, the vector potential, A , of the magnetic field is A=\begin{bmatrix}-\partial_{2}\phi\\ \partial_{1}\phi\\ 0\end{bmatrix} Our goal is as follows. We suppose |\partial^{\alpha}\phi|\leq C_{\alpha}\text{ for all }|\alpha|\geq 2 (1) and consider eigenfunctions u of H where Hu=\lambda^{2}u (2) Then, we will demonstrate the following two results. Theorem 1 . With H and \phi defined as above, we have ||u||_{L^{\infty}}\lesssim||u||_{L^{2}} (3) where the implicit constant is independent of the eigenvalue \lambda^{2} . and Theorem 2 . With H and \phi defined as above, we have ||u||_{L^{6}}\lesssim\lambda^{-1/3}||u||_{L^{2}} (4) Note that the second result is strictly better than what we would get by interpolation with the L^{\infty} estimate. In the specific case that \phi(z)=|z|^{2} , ?THM? is contained implicitly in the work of Folland and Thangavelu, see [ 3 ] and [ 7 ] . Folland, for example, shows that functions in the null space of this operator have the form e^{-|z|^{2}}F(\bar{z}) where F is anti-holomorphic. Then, the null space has the reproducing kernel K(z,w)=e^{2z\bar{w}} , from which the L^{\infty} estimate easily follows. He furthermore shows that the eigenspaces have eigenvalues which are the even non-negative integers (called Landau levels) and that functions in these eigenspaces can be obtained by applying the corresponding creation operator to eigenfunctions in the lower eigenspaces. Thus, the eigenspaces are infinite dimensional. Furthermore, Thangavelu’s work demonstrates that these eigenfunctions can be written as twisted convolutions with a special Hermite function, in which case the L^{\infty} result follows from a simple convolution inequality and growth estimates on the special Hermite functions. To the best of my knowledge, however, no previous work has been done on the L^{6} inequality. In the case of a general potential function, it remains true that the null space contains functions of the form e^{-\phi}F(\bar{z}) where F is anti-holomorphic, provided \phi grows fast enough that these functions are L^{2} . For example, in the case that \phi\gtrsim|z|^{\epsilon},\epsilon>0 , the null space will contain functions of the form e^{-\phi}p(\bar{z}) where p is any polynomial. Thus, the null space of H will still be infinite dimensional. We furthermore expect that, in the case of compact perturbations of the original operator with potential function \phi(z)=|z|^{2} , the higher eigenspaces may also be infinite dimensional or may split into spectral clusters. Note that, by switching to real coordinates, we can rewrite H as H=\left(\frac{D_{1}}{2}-\frac{\partial_{2}\phi}{2}\right)^{2}+\left(\frac{D_{2}}{2}+\frac{\partial_{1}\phi}{2}\right)^{2}-\frac{\Delta\phi}{4} (5) which is the form of the operator with which we will concern ourselves hereafter. For example, in the specific case that \phi(z)=|z|^{2} , this becomes H=\left(\frac{D_{1}}{2}-x_{2}\right)^{2}+\left(\frac{D_{2}}{2}+x_{1}\right)^{2}-1 (6) Thus, we see that, when x is confined to a compact ball, this operator behaves like the standard Laplacian, and we describe below how we deal with the case of x large using a conjugation argument. In order to prove these results, we take advantage of developments in semiclassical analysis; see [ 5 ] and [ 8 ] . In particular, our method of proof is inspired by work done by Koch, Tataru, and Zworski in [ 5 ] , which generalizes earlier work of Koch and Tataru on the Hermite operator, see [ 4 ] , and of Sogge on the Laplace-Beltrami operator on a compact Riemannian manifold, see [ 6 ] . Like the Laplace-Beltrami operator, the operator we consider is also of principal type. However, it has the same scaling as in the case of the Hermite operator. Koch, Tataru, and Zworski are able to prove estimates on eigenfunctions of principal type operators by factoring into an elliptic operator and a Schrodinger operator and using energy estimates and Strichartz estimates for the Schrodinger operator. In our case, we likewise factor our operator into an elliptic operator and a Schrodinger operator and use the same energy and Strichartz estimates that Koch, Tataru, and Zworski do. We then arrive at results that are analogous to the results of Sogge for the Laplace-Beltrami operator on a compact Riemannian manifold. What sets our approach apart is that we are working on all of {\mathbb{R}}^{2} and therefore some care is needed to appropriately localize our eigenfunctions. We take advantage of a multiplication operator which we denote T_{q},q\in{\mathbb{R}}^{2} , such that conjugating our operator with T_{q} allows us to approximately translate our operator in a manner which is uniformly bounded in q . This paper is organized as follows. In section 2, we prove ?THM? and ?THM? for the case when eigenvalues \lambda^{2} are sufficiently large. In section 3, we prove ?THM? and ?THM? for eigenvalues \lambda^{2} which are less than some constant. We have an appendix at the end which contains some important results of semiclassical analysis which we use in the body of this paper. 2 Large Eigenvalue Case In this section, we turn to the large eigenvalue case, which is the more difficult and interesting case. Now, if we let u_{\lambda}(x)=u(\lambda x) and \phi_{\lambda}(x)=\phi(\lambda x) , then ( 2 ) gives us \left[\left(\frac{1}{2}\frac{D_{1}}{\lambda}-\frac{\partial_{2}\phi_{\lambda}}{2\lambda}\right)^{2}+\left(\frac{1}{2}\frac{D_{2}}{\lambda}+\frac{\partial_{1}\phi_{\lambda}}{2\lambda}\right)^{2}-\frac{\Delta\phi_{\lambda}}{4\lambda^{2}}\right]u_{\lambda}=\lambda^{2}u_{\lambda} Dividing by \lambda^{2} and rearranging, we get \left[\left(\frac{1}{2}\frac{D_{1}}{\lambda^{2}}-\frac{\partial_{2}\phi_{\lambda}}{2\lambda^{2}}\right)^{2}+\left(\frac{1}{2}\frac{D_{2}}{\lambda^{2}}+\frac{\partial_{1}\phi_{\lambda}}{2\lambda^{2}}\right)^{2}-\frac{\Delta\phi_{\lambda}}{4\lambda^{4}}-1\right]u_{\lambda}=0 If h=\frac{1}{\lambda^{2}}, let u_{h}(x)=u(h^{-1/2}x)=u(\lambda x)=u_{\lambda}(x) and \phi_{h}(x)=\phi(h^{-1/2}x)=\phi(\lambda x)=\phi_{\lambda}(x) , then this becomes \left[\left(\frac{h}{2}D_{1}-\frac{h}{2}\partial_{2}\phi_{h}\right)^{2}+\left(\frac{h}{2}D_{2}+\frac{h}{2}\partial_{1}\phi_{h}\right)^{2}-\frac{h^{2}}{4}\Delta\phi_{h}-1\right]u_{h}=0 Or, by rewriting this, we have \left[\left(\frac{h}{2}D_{1}-\frac{h^{1/2}}{2}(\partial_{2}\phi)(h^{-1/2}x)\right)^{2}+\left(\frac{h}{2}D_{2}+\frac{h^{1/2}}{2}(\partial_{1}\phi)(h^{-1/2}x)\right)^{2}-\frac{h}{4}(\Delta\phi)(h^{-1/2}x)-1\right]u_{h}=:Pu_{h}=0 Note that ||u_{h}||_{L^{\infty}}=||u||_{L^{\infty}} and, by a change of variables, ||u_{h}||_{L^{2}}=h^{1/2}||u||_{L^{2}} and ||u_{h}||_{L^{6}}=h^{1/6}||u||_{L^{6}} . Thus, ?THM? can be reformulated as follows. Theorem 3 . Suppose Pu_{h}=0 , then ||u_{h}||_{L^{\infty}}\lesssim h^{-1/2}||u_{h}||_{L^{2}} (7) Likewise, ?THM? can be reformulated as Theorem 4 . Suppose Pu_{h}=0 , then ||u_{h}||_{L^{6}}\lesssim h^{-1/6}||u_{h}||_{L^{2}} (8) Define \displaystyle A \displaystyle:=\frac{h}{2}D_{1}-\frac{h}{2}(\partial_{2}\phi_{h})(x) \displaystyle B \displaystyle:=\frac{h}{2}D_{2}+\frac{h}{2}(\partial_{1}\phi_{h})(x) \displaystyle P \displaystyle:=A^{2}+B^{2}-\frac{h^{2}}{4}(\Delta\phi_{h})(x)-1 \displaystyle\tilde{A}_{q} \displaystyle:=\frac{h}{2}D_{1}-\frac{h}{2}(\partial_{2}\phi_{h})(x+q)+\frac{h}{2}(\partial_{2}\phi_{h})(q) \displaystyle\tilde{B}_{q} \displaystyle:=\frac{h}{2}D_{2}+\frac{h}{2}(\partial_{1}\phi_{h})(x+q)-\frac{h}{2}(\partial_{1}\phi_{h})(q) \displaystyle\tilde{P}_{q} \displaystyle:=\tilde{A}_{q}^{2}+\tilde{B}_{q}^{2}-\frac{h^{2}}{4}(\Delta\phi_{h})(x)-1 Likewise, we use the corresponding lowercase letters to denote the symbols corresponding to the above operators, e.g. A=a^{\mathrm{w}}(x,hD) and \tilde{A}_{q}=\tilde{a}_{q}^{\mathrm{w}}(x,hD) . Note, furthermore, that, in the specific case of a quadratic potential function, \tilde{A}_{q}=A and likewise for \tilde{B}_{q} and \tilde{P}_{q} . We likewise define \displaystyle u_{h,q}(x) \displaystyle:=u_{h}(x-q) \displaystyle\sigma(x,\xi) \displaystyle:=\langle x_{2},\xi_{1}\rangle-\langle x_{1},\xi_{2}\rangle \displaystyle T_{q} \displaystyle:=e^{i\sigma(x,h^{-1/2}\nabla\phi(h^{-1/2}q))} \displaystyle S_{\rho,\delta}(m) \displaystyle:=\{a\in C^{\infty}||\partial_{\xi}^{\alpha}\partial_{x}^{\beta}a|\lesssim h^{-\rho|\alpha|-\delta|\beta|}m\text{