Description of limiting vorticities for the magnetic 2D Ginzburg-Landau equations

TL;DR

This paper characterizes the structure of limiting vorticities and magnetic fields in 2D Ginzburg-Landau equations, showing their support lies on smooth curves and ruling out excessive vortices under certain conditions.

Contribution

It provides a detailed description of the measures and functions satisfying the limiting equations, including their regularity and support properties, extending previous results by Sandier-Serfaty.

Findings

 ext{Support of } ext{ is on } \u0010 ext{C}^1 ext{-curves near points with nonzero gradient}

 ext{Vorticity measure is absolutely continuous w.r.t. 1D Hausdorff measure on support}

 ext{If }h=0 ext{ on boundary and domain is star-shaped, then }h ext{ is identically zero}

Abstract

Let $\Omega$ be a bounded open set in $\mathbb{R}^2$. The aim of this article is to describe the functions $h$ in $H^1(\Omega)$ and the Radon measures $\mu$ which satisfy $-\Delta h+h=\mu$ and $ div(T_h)=0$ in $\Omega$, where $T_h$ is a $2\times 2$ matrix given by $(T_h)_{ij}=2\partial_ih\partial_jh- (|\nabla h|^2+h^2)\delta_{ij}$ for $i,j=1,2$. These equations arise as equilibrium conditions satisfied by limiting vorticities and limiting induced magnetic fields of solutions of the magnetic Ginzburg-Landau equations as shown by Sandier-Serfaty. Let us recall that they obtained that $|\nabla h|$ is continuous in $\Omega$. We prove that if $x_0$ in $ \Omega$ belongs to $\supp \mu$ and is such that $|\nabla h(x_0)|\neq 0$ then $\mu$ is absolutely continuous with respect to the 1D-Hausdorff measure restricted to a $\mathcal{C}^1$-curve near $x_0$ whereas $\mu_{\lfloor \{| \nabla h|=0\}}=h_{|\{ |\nabla h|=0\}}$. We also prove that if $\Omega$ is smooth bounded and star-shaped and if $h=0$ on $\partial \Omega$ then $h \equiv 0$ in $\Omega$. This rules out the possibility of having critical points of the Ginzburg-Landau energy with a number of vortices much larger than the applied magnetic field $h_{ex}$ in that case.