Multiscale Micromagnetismus

A detailed understanding of the underlying magnetic processes is of utmost importance for the development of novel magnetic data storage devices such as hard disc drives, patterned elements and magnetic random access memories. The functional behaviour of these devices on a timescale of years has to be known in order to guarantee a high reliability. However, the thermal stability of the stored information can not be directly accessed experimentally. Theories have to be used to interpret the measured data. Sharrock`s equation for example is used to extrapolate the thermal stability of a recorded bit pattern from time dependent coercivity measurements. Most of these theories bases on the simple Stoner-Wohlfarth theory for single domain particles. The assumption of the Stoner-Wohlfarth theory was justified for conventional recording devices where every grain consists of one single phase and the magnetization remains uniform. However, state of the art recording structures consists of complicated microstructures with different magnetic phases. New theories and models are required. The models have to be suitable to calculate the magnetization dynamics within different time scales and length scales. Field induced switching process occurs within nanoseconds, whereas typical times scales of thermally induced switching processes in recording structures are several years. In order to account for these multiscale problems sophisticated numerical methods have to be applied. Renormalization techniques are developed to improve the accuracy of Langevin dynamic calculations that describes the magnetization dynamics under the action of a thermal field on the time scale of nano seconds. For the calculation of magnetization process on a time scale of years the transition state theory is applied and extended with novel methods such as the transition path sampling. A main part of the project is the numerical calculation of the attempt frequency for large scale micromagnetic models. The large scale system matrix is compressed using hierarchical matrices (H-matrices). The calculation of the attempt frequency requires the calculation of the eigenvalues and determinants of H-matrices, which is a delicate problem. A Monte Carlo estimator for the calculation of determinants of large scale matrices is tested. The ability to calculate the energy barrier and attempt frequency accurately for complicated microstruceres with different magnetic phased is applied in the second part of the project. A Monte Carlo method using precomputed energy barriers and attempt frequencies is applied to calculate the thermal decay of a bit pattern of exchange spring media - a state of the art recording structure. These calculations are compared with experiments, where exchange spring structures are deposited onto polystyrol particles forming isolated monodisperse magnetic nanocaps. Finally, a modified Sharrock`s equation should be developed that improves the accuracy of the measured thermal stability of exchange spring media.

The rapid progress of nanotechnology allows the realization of novel concepts in magnetic recording and logic devices. The design of these smart magnetic devices and materials requires the detailed knowledge of the response of the system to external fields and temperature as a function of time. Of particular interest is the information of the long term thermal stability of magnetic devices such as hard discs, magnetic random access memories and future concepts relying on pre-patterned magnetic elements. However, the thermal stability of the stored information cannot be directly accessed experimentally and the judgment whether a new hard disc can hold its information for at least 10 years cannot be easily made. Theories and simulations have to be used to calculate the thermal stability and to interpret the measured data. Within this project it was succeeded for the first time to calculate the thermal stability of complex magnetic nanostructures without any free parameter. All parameters which are used in the simulations can be determined by experiments. This allows for a detailed description and design of new magnetic storage devices. With the help of this advanced method essential new storage concepts could be developed. One central problem in magnetic hard discs is, that with increasing storage density the magnetic bits become so small that these bits are no longer thermally stable or the bits are stable but cannot be written by the write head. A very recent solution to this problem relies on so called exchange spring media, where extremely thermally stable layers are stacked to unstable layers. Within this project this concept could be significantly improved by showing that the use of more than two layers indeed only lowers the required write field but does not decrease the thermal stability. The thermal stability of the new storage concept that relies on magnetic multilayers could be calculated and compared with conventional media without free parameters. We could demonstrate that recording on graded media is possible up to a storage density of 2.5 TeraBits/inch 2 which is about a factor of 4 higher than state of the art hard discs. One other highlight which came out of multiscale simulations of the research project was the invention of a three dimensional storage concept. Microwave fields are used in order to selectively address and write different magnetic layers. This allows that the information is not only stored in one magnetic layer but also in the third dimensions. This concept which is registered as an US patent has the potential to significantly increase the storage capacity of future hard discs that relies on pre patterned elements.