Micromagnetic model of noncollective magnetization reversal in ultrathin magnetic dots with in-plane uniaxial anisotropy - art. no. 174418

Citation
O. Fruchart et al., Micromagnetic model of noncollective magnetization reversal in ultrathin magnetic dots with in-plane uniaxial anisotropy - art. no. 174418, PHYS REV B, 6317(17), 2001, pp. 4418
Citations number
57
Categorie Soggetti
Apllied Physucs/Condensed Matter/Materiales Science
Journal title
PHYSICAL REVIEW B
ISSN journal
01631829 → ACNP
Volume
6317
Issue
17
Year of publication
2001
Database
ISI
SICI code
0163-1829(20010501)6317:17<4418:MMONMR>2.0.ZU;2-2
Abstract
In most magnetic systems the magnetization reversal is nonuniform, and is i nitiated in a so-called "nucleation volume," whose dimensions are by far sm aller than the total system volume. For simplicity reasons magnetization re versal theories are usually based on the assumption that coherent rotation occurs in this "nucleation volume." In this approach, self-dipolar fields a nd exchange forces are obviously not well described, because in reality the nucleation volume is coupled with the rest of the system. In the case of u ltrathin dots with in-plane uniaxial anisotropy, we could take into account dipolar fields and the exchange stiffness explicitly. The approximations u sed to derive analytical equations were suggested by experimental results o n real dots. The model yields the nonuniform micromagnetic configuration of nucleation volumes. It predicts nucleation and reversal field values, as w ell as the field dependence of the energy barrier to be overcome to reverse the dot at finite temperature. The (negative) reversal field is found to i ncrease with the dot thickness T and the volume magnetization M-s, and to d ecrease with the material anisotropy K. In the low-thickness limit, the rev ersal field H-r approaches the Stoner-Wohlfarth reversal field H-a with a l aw close to 1-\H-r\H-a\ similar toM(s)(7/2)A(-3/4)K(-1)T(3/2), where A is t he exchange constant. The relevance of the approximations used is discussed and demonstrated by the good agreement found for all predictions between e xperiment and/or numerical calculations on the one hand and the model on th e other hand.