next up previous contents
Next: Analytical calculation of the Up: Summary and outlook Previous: Summary and outlook   Contents

Summary

We have investigated the properties of micromagnetic samples through simulation paying particular attention to shape anisotropy -- the influence the shape has on the hysteresis loop and magnetic microstructure.

Most magnetic samples at a nanometre scale possess characteristics desirable for magnetic data storage applications, either as data read/write heads or sensors, where the magnetisation is sensitive to small changes in $ \ensuremath{\mathbf{B}}$, or as storage media with a high coercivity.

We note that cylinders are particularly suitable candidates for storage media when considering their comparatively large coercivity, and our phase diagram identifies the behavioural dependence on diameter and height. Spherical geometries exhibit properties useful for magnetic sensors. Our studies reveal how coercivity decreases when overall diameter is increased.

More complex shapes, such as the cone, half-sphere and the droplet are more flexible, and by subtle alterations in their sizes can be manipulated for use in any data storage application. Size dependence studies in the droplet geometry show that the transition between the single-domain and the vortex state occurs at a bounding sphere diameter of 140nm in nickel. Larger droplets which have reversible characteristics and zero coercivity are ideal for magnetic sensor applications.

Antidots show interesting characteristics when considered in arrays of their peers, and given the flexibility of the coercive field through altering the size of the dots as a fraction of the spacing distance, make good candidates for storage media.

Comparisons between our simulations and experimental results demonstrate a high degree of similarity. We can use micromagnetic simulation to observe hysteretic behaviour and coercivity trends with high confidence, and the study of samples with other geometries and sizes which are not yet feasible for experimentalists to physically produce can yield possible directions for future experimental work.

Overall, we conclude the coercivity of magnetic samples can be controlled by their physical shape and size.

Possible candidates for further simulation include the remainder of the primitive set -- torii and pyramids -- as well as constructive solid geometries based upon these primitives, for example a tear-drop shape built from a conical upper and a half-ellipsoid lower section. Shapes such as these could be created by combining the self-assembly double-template method outlined in chapter 5 with electron beam lithographic techniques. Using this method it is also possible to create films manufactured from different materials -- for example, it might be possible to create a cobalt half-sphere which has a permalloy ``hat''.

The simulations presented in this thesis were computed as zero temperature. Thermal effects, even at low temperatures, can introduce subtle changes in behaviour. As such, the contribution to the results through finite temperature would be studied in a continuation of this work. Initial studies indicate that the size of secondary hysteresis loop energy barriers in the case of the sphere is reduced at finite temperature.

The work discussed in section 5.5.2 revealed some interesting properties of vortex behaviour in the droplets. Initial studies have shown that the droplets, especially larger ones, demonstrate similar vortex orientations to those found in spheres, although they quickly fall into a more disc-like out-of-plane vortex state as the applied field is further reduced.

A more detailed study will be performed into how the variation of $ l_{z(s)}$ and $ l_{z(e)}$ from equations 5.1 and 5.1 affects the vortex formation, movement and subsequent disappearance. The effect of magnetostatic energy between adjacent nanodots will also be investigated.

Finally, in chapter 6 we compared experimental data with simulation results produced through two-dimensional simulations, Monte Carlo methods and found a strong similarity between experimental MFM and computed stray field results, corroborating the two-dimensional model presented in that chapter.


next up previous contents
Next: Analytical calculation of the Up: Summary and outlook Previous: Summary and outlook   Contents
Richard Boardman 2006-11-28