Vacancies in artificial spin ice

We study artificial spin-ice systems (in the form of square arrays of islands of magnetic materials) from the numerical point of view. Typical Permalloy elongated and anisotropic islands [1] occupy the positions completely filling a layer L×L, with variable L; several equivalent layers are then piled up. We focus on the vertices (that are the convergence points of islands) as the internal layers are diluted upon introducing vacancies (lack of islands). This is both a random and controlled process for a single internal layer; the upper-most and lower-most layers remain without vacancies. The introduction of vacancies generated several new types of vertices that can be characterized according to their distribution depending of vacancy concentration. We found that a small amount of vacancies has strong influence on the vertex distributions. Moreover, we considered a code based format for the different vertices and found that is possible to write on each layer complete and robust messages. The propagation of information in a transverse way (from top to lower layer) is also explored.
[1] S. A. Morley, et al. Phys. Rev. B 95, 104422 (2017)