Title: Micromagnetic Simulations of Systems with Shape-Induced Anisotropy
1Micromagnetic Simulations of Systems with
Shape-Induced Anisotropy
Florida State University
2Manufactured Single-Domain Iron Nanopillars
- Grown by STM-assisted CVD 1
200 nm
40 nm
1 S. Wirth, M. Field, D. D. Awschalom, and S. von
Molnar, Phys. Rev. B 57, R14028 (1998) J. Appl.
Phys. 85, 5249 (1999)
3Background A Typical Simulation
Mz
4Metastable Decay
F
Saddle point
Free energy barrier
Metastable minimum
Mz
Free energy vs. Mz
5(No Transcript)
6Lattice
Dynamic equation, Landau-Lifshitz-Gilbert (LLG)
on a computational lattice
7Landau-Lifshitz-Gilbert (LLG)
- uniform magnetization density
- go electron gyromagnetic ratio, 1.67x107 Hz/Oe
- a phenomenological damping parameter, 0.1
- total local field at
8Rescaling
- M M/Ms
- H H/Ms
- r r/le
- t ?oMst
9Dipole interactions
Handled by the Fast Multipole Method O(n2)
O(n)
10Realistic Models Computational Details
- 6x6x90 computational lattice -gt 3240 sites
- 10 nm x 10 nm x 150 nm Fe nanopillar
- dt 0.083 picoseconds usin first-order Euler
integration - Temperature 20 Kelvin
- Applied Field 3160 Oe at 75 degrees from the
easy axis - Fields dipole-dipole, thermal, exchange, Zeeman
- 3-4 days on IBM SP3 using 20 processors
11Cumulative Distribution Function of the Lifetime
t All Runs
12Phase Plot of the Total Magnetization
13Phase Plot of the Total Energy
Slower Mode
14Two Distributions
Fast
Slow
15Thermalization out of the T0K Metastable State
Quenched
Slower mode
16Thermalization
T 20K
T 0K
17Quench/Relax vs Slow Mode
Q/R
Slow
18Projective Dynamics
Number of visits, N
Same, N (G S)
Growth, G
Shrinkage, S
Mz
Growth Probability G/N Shrinkage Probability
S/N
19Projective Dynamics
Slower mode
Faster mode
20Location of Extrema
21Choosing the Correct Model
7x7x101 spins
1x1x17 spins
Stoner-Wohlfarth
- 1-D models are not sufficient, full 3-D models
are necessary
Wirth, et al, J. Appl. Phys. 85, 5249 (1999). Li,
et al, J. Appl Phys, 93, 7912 (2003). Li, et al,
J. Appl. Phys. Lett 80, 4644 (2002)