Exchange Bias and Training Effect in Polycrystalline Antiferromagnetic/Ferromagnetic Bilayers
- PDF / 413,158 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 67 Downloads / 209 Views
Q7.6.1
Exchange Bias and Training Effect in Polycrystalline Antiferromagnetic/Ferromagnetic Bilayers Markus Kirschner, Dieter Suess, Thomas Schrefl and Josef Fidler Solid State Physics, Vienna University of Technology, Wiedner Haupstr. 8-10/138, A-1040 Vienna, Austria ABSTRACT Exchange bias and training effect are simulated for IrMn/NiFe bilayers. As a function of the thickness of the antiferromagnet the bias field shows a maximum for a thickness of 22 nm. For decreasing antiferromagnetic thickness the domain wall energy approaches zero. For large thicknesses the high anisotropy energy hinders switching of the antiferromagnetic grains resulting in weak bias. Starting from the field cooled state as initial configuration a bias field of about 8 mT is obtained assuming a antiferromagnetic layer thickness of 20 nm, a ferromagnetic layer thickness of 10 nm, and a grain size of 10 nm. The next hysteresis cycle shows a reduction of the bias field by about 65%. Exchange bias and training effect in fully compensated antiferromagnet/ferromagnet bilayers are explained with a simple micromagnetic model. The model assumes no defects except for grain boundaries, and coupling is due to spin flop at a perfect interface. The simulations show that a weak exchange interaction between randomly oriented antiferromagnetic grains and spin flop coupling at a perfectly compensated interface are sufficient to support exchange bias. INTRODUCTION The phenomena of exchange anisotropy and exchange bias, particularly, were discovered in the year 1956 by Meiklejohn and Bean [1] when studying Co particles surrounded with antiferromagnetic oxide (CoO). They found that the field required to switch the ferromagnet from the field cooled state into the reversed state is larger than that to rotate the ferromagnet back to its original direction. Since the introduction of the Giant Magnetic Resistance (GMR) head in magnetic recording [2] the bias effect has been used widely in modern technology. The pinned layer of a spin valve sensor is stabilized through coupling to an antiferromagnet. A common system used in GMR reading heads are IrMn/NiFe bilayers [3]. Despite the application of exhange bias in magnetic field sensors, the physical mechanisms that lead to the hysteresis shift are still a field of discussion. Various theories explain particular aspects of the bias effect [4]. Nevertheless many issues remain to be solved [5]. One of the most striking experimental facts is the presence of exchange bias at fully compensated antiferromagnetic (AF) interfaces [6,7] in which the net spin averaged over a microscopic length scale is zero. Intuitively, one might expect that for compensated interfaces the bias effect vanishes, as the spins pinning the ferromagnetic cancel. Therefore various models of exchange bias assume partly uncompensated interfaces [4]. In this paper we propose a mechanism for exchange bias at fully compensated interfaces. The numerical results obtained for IrMn/NiFe bilayers are compared with experimental data from the recent literature [8,9
Data Loading...
