Estudo de sistemas magnéticos nanoestruturados via simulação computacional, com aplicação em dispositivos de memórias magnetorresistivas

Abstract

Nanostructured magnetic systems have properties that are extended to several technological applications, such as high density magnetic recording media, magnetoresistive random access memory, data processing devices, etc. There is a growing demand for increasingly fast and high storage capacity recording media, which has driven not only the study of manufacturing nanostructured magnetic systems, but also the evolution of experimental techniques to measure their properties. These nanostructures can be manufactured in several formats, using magnetic materials such as cobalt or Permalloy alloy. Depending on the material, shape and size of the sample, they may exhibit different domain regions. These domain regions define the information bits in possible memory devices. Thus, knowing the dynamic behavior of these domain regions is essential. Several spintronic memory devices use single-domain planar magnetic nanoparticles as unit cells. In this thesis we will present three proposals for memory devices. In the first part of this work, we performed micromagnetic computational simulations to investigate interparticle dipole-dipole interactions in single-domain nanoelipses matrices made from Permalloy. Assuming a certain limit of size and aspect ratio, the nano-ellipse remanent state made of Permalloy can be a single and almost uniform domain, due to the shape anisotropy. There are two directions along the ellipse major axis on which the magnetization can point (degenerate states), which define an information bit. The magnetization is alternated by applying a magnetic field perpendicular to the easy magnetization axis. In this study involving a matrix of nano-ellipses, the variation is introduced not only in the aspect ratio and in the spacing between ellipses, but to the magnetization distribution and the 2D lattice type as well. For small values of the spacing between ellipses, they interact strongly, such that the shape anisotropy is locally modified by the magnetization distribution. So, we evaluated the minimum spacing between the particles in which the single domain nano-ellipses become a non-interacting magnetic system. This kind of sample can be applied to Toggle MRAM devices, such that understanding and controlling the magnetic nano-ellipses magnetization switching is extremely important. In the second stage, we consider a transverse domain wall in a Permalloy nanostrip, pinned by geometric defects called “ anti-notches”. Due to shape anisotropy, these anti-notches have an attractive character. Thus, investigating through computer simulation the interaction of this domain wall by the geometric defect and, due to be a finite device, the edge effects that this wall suffers from, this is the first step for study. Configurations that have sufficiently small geometries, but magnetically stable, allow this sample to be used in a three-terminal device. The transverse domain walls behave like quasiparticles that move fastly from one “ anti-notch” to another by very short pulses of electrical current applied parallel to the nanostrip easy axis. The rapid stabilization of the domain wall, with well-defined magnetization states, allows the measurement of a high magnetoresistive signal by a magnetic tunnel junction and indicates the proposed device as a promising magnetoristive random access memory cell. In the third and last stage, we consider the skyrmion dynamics in nanowires made of Co/Pt multi-layers. As in the previous step, we need to understand and control, through micromagnetic simulation, this quasiparticle that moves by the application of electric current applied parallel to the nanowire major axis. During its movement, it is possible to observe the skyrmion Hall effect, mapping its position in each computational step. As it is a new subject for the group, observing well known results obtained in the literature and being able to reproduce them using our program, demonstrates the reliability of our code. We propose the use of an impurity track, combined with changing the Dzyaloshinskii-Moriya interaction parameter, in order to cancel the skyrmion deflection while it moves. We consider impurity to be a local variation of magnetic properties. It was possible to obtain configurations in which we can observe the total cancellation of skymion Hall effect. We also mapped the skyrmion topological charge during its displacement to ensure this quasiparticle continues to maintain its topological properties. We believe these results presented here have a potential technological application in future memory devices.