Concentration behavior and lattice structure of 3D surface superconductivity in the half space (Q1741776)

It is well understood that in a suitable range of magnetic field strengths, the solutions to the Ginzburg-Landau (GL) equations are localized near the boundary. While a detailed study of the GL model in a 2D domain was the subject of numerous papers, the study of the model in a 3D domain is much less complete. In order to improve the understanding of this boundary layer, the authors examine the energy contribution of the order parameter $\psi$ in the vicinity of the domain boundary. It is proposed that, in the surface superconducting state, the order parameter will exhibit a certain lattice structure in the vicinity of the domain boundary similar to the Abrikosov lattices of 2D samples. Understanding the lattice structure near the boundary helps to understand the vortex lattices of superconductivity when the applied magnetic field decreases and approaches the second critical field. The pair $(\psi, A)$, where $A$ presents the magnetic potential, is a critical point of a specific energy (GL) functional. After rescaling and taking limits, the behavior of the solutions to the GL equations in a boundary layer are studied on the base of the limiting equations in half-space. This means that the boundary curvature has disappeared. It has been shown by \textit{S. Fournais} and \textit{A. Kachmar} [Commun. Partial Differ. Equations 38, No. 1--3, 339--383 (2013; Zbl 1267.82140)] that the curvature does not affect the energy to leading order -- one might expect curvature effects to appear in more precise descriptions of the surface superconductivity, as in the 2D case. In this boundary layer, the magnetic field can be considered constant, because on short length scales any sufficiently smooth function is well approximated by a constant. The ground state energy is a function of the inclination of the magnetic field and the result gives that the energy density increases when the magnetic field tends to be perpendicular to the surface of the sample. This complements results from the linear analysis. It is well known that superconductivity survives the longest (when the exterior magnetic field is increased) in the boundary region where the magnetic field is parallel. Theorem 1.6 on an explicitly constructed boundary energy density $e(b,\nu)$, to being a monotone non-decreasing function (where $\nu(x)$ denotes the angle between the tangent plane at $x \in\partial\Omega$ and the magnetic field, $b \in(0.59,1[$) shows that also before this happens superconductivity is the strongest where the magnetic field is parallel and monotonically decreases as the angle $\nu$ increases. This interpretation of Theorem 1.6 follows from the correspondence between $e(b,\nu)$ and the density of superconductivity. As the main result, the authors construct solutions with 3D lattice structure in the case when $\nu\neq0$. The corresponding proposition is proved and suggests that a bounded superconductor, subjected to an applied magnetic field with strength between the second critical field and the third critical field, will be in a surface superconducting state, and vortex lattices will appear along some surface layers where the applied magnetic field is not tangential to the surface. The bounded solutions are constructed with lattice structure, i.e., for which the physical quantities -- the density of Cooper pairs, the magnetic field and the magnetic current are periodic. This corresponds to states of $\psi$ satisfying the magnetic periodic conditions. The construction of the lattice solutions is completely similar to the Abrikosov solutions in 2D. The result, proved in Theorem 5.5, implies an upper bound on the energies (infimum of reduced GL type energy functional and $e(b,\nu)$). This situation complements Theorem 1.5. For $\nu=0$, there is a separation of variables implying that the modulus of a minimizer is constant along the boundary. As soon as $\nu\neq0$, it could be expected (in analogy with the 2D-case) this continuous symmetry to be broken and be replaced by the discrete (lattice) symmetry. In particular, for $\nu\neq0$, the minimizers will display a vortex lattice structure. However, just as this remains unproven in 2D, the authors cannot give a rigorous proof of this expectation in the present 3D setting. The results of this paper demonstrate how the surface superconductivity varies along the boundary of a finite sample when the external magnetic field is between the second and third critical fields. Superconductivity is the strongest where the external magnetic field is parallel to the boundary and decreases monotonically with the angle $\nu$. Moreover, where the field is parallel, the structure is precisely given in terms of a 1D-model by separation of variables. For non-zero angle $\nu$, this separation of variables seems to fail and is probably replaces by a lattice-type structure similar to the 2D-situation of Abrikosov lattices.