Positive Exchange Bias in FeF2-Fe Bilayers

Solid State Communications, 2000

Positive exchange bias (PEB) is a remarkable phenomenon, which was recently observed experimentally. Normal (negative) exchange bias (NEB) was discovered more than 40 years ago. Its signature is the shift of the hysteresis loop along the applied field axis by H E Ͻ 0; in systems where a ferromagnet (FM) is in close contact with an antiferromagnet (AFM). This occurs after the system is cooled below the Néel temperature in an external field H cf of a few kOe. As H cf is substantially increased H E adopts positive values. Here we explain this rather unexpected behavior on the basis of an incomplete domain wall model that develops in the FM, for Fe/FeF 2 and Fe/MnF 2 systems. A consistent and unified picture of both NEB and PEB, and satisfactory quantitative agreement with experimental results are obtained on the basis of our theory.

Physical Review B, 2002

The angular dependence of the magnetic anisotropy of exchange biased Fe/MnF 2 bilayers was measured. Below the Néel temperature of the antiferromagnetic MnF 2 layer, an exchange anisotropy is observed which consists of unidirectional, uniaxial, threefold and fourfold symmetry components. The threefold exchange anisotropy term is responsible for the asymmetric magnetization reversal process recently observed in this system.

Physical Review B, 2003

An analytical model of exchange anisotropy in epitaxial ferromagnetic/antiferromagnetic bilayers was developed. The model demonstrates that the high symmetry exchange anisotropy terms in ferromagnetic/ antiferromagnetic bilayers originate from a partial domain wall in the antiferromagnetic layer. Application of the model to the experimental data analysis enables one to separately determine the fraction of uncompensated interfacial spins in the antiferromagnetic layer and the interfacial exchange coupling energy between spins in the ferromagnet and in the antiferromagnet. The model provides a quantitative description of complex exchange anisotropy recently observed in Fe/MnF 2 bilayers.

Physical Review Letters, 2003

The exchange bias and magnetic anisotropies in a Co layer on a single-crystalline FeF 2 film have been determined between 30 and 300 K. By postulating that the coupling between the ferromagnet and the antiferromagnet persists above the Néel temperature (T N ) we develop a model that quantitatively describes the exchange bias and the anisotropies over the whole temperature range, both above and below T N . Using only the measured low temperature exchange bias and a distribution of blocking temperatures we explain (i) the temperature dependence of the bias, (ii) the magnitude of the anisotropies, (iii) the opposite sign of the first and second order anisotropies, (iv) the observed 1=T and 1=T 3 temperature dependencies of the first and second order uniaxial anisotropies above T N , and (v) the decrease of the anisotropies below T N .

Low Temperature Physics, 2009

Field dependences of the magnetization and exchange bias in ferro/antiferromagnetic systems. II. Continuum model of a ferromagnetic layer

Physical Review B, 2008

We measured directly the depth-dependent Fe spin rotation upon magnetization reversal in exchangecoupled Fe/ MnF 2 bilayers using nuclear resonant scattering of synchrotron radiation from an 57 Fe-probe layer buried at different depths within the Fe film. Our results show that the exchange-biased ferromagnetic layer develops a noncollinear spin structure along the film normal direction, reminiscent of a partial domain wall parallel to the Fe/ MnF 2 interface. This is contrary to most theoretical models of exchange bias which assume a collinear spin structure in the ferromagnetic layer.

Physical Review B, 2010

2002

Exchange-biased MnF 2 /Fe bilayers, examined by variable angle and temperature ferromagnetic resonance ͑FMR͒, exhibit a sudden onset of a unidirectional and fourfold anisotropy below the MnF 2 Néel temperature. This unexpected fourfold symmetry arises from frustrated perpendicular coupling between the MnF 2 and the Fe overlayer in the presence of twinning in the antiferromagnet layer. These data are consistent with earlier polarized-neutron-reflectometry results. The FMR data show a clear reversal in the direction of the unidirectional anisotropy as a function of cooling field, switching sign at H FC ϭ13 kOe, which is consistent with the onset of positive exchange bias observed in conventional magnetometry experiments. The low-temperature FMR linewidth reflects the in-plane symmetry of the resonance itself, exhibiting surprising divergence in the hard directions. Temperature-dependent FMR measurements reveal a sharp reduction in the resonance field below the Néel point due to the ferromagnetic/antiferromagnetic coupling.

Journal of Applied Physics, 2002

Single-crystal thin films of the antiferromagnet FeF2 have been used to exchange bias overlayers of Fe. An unexpected coercivity enhancement is observed at temperatures above the Néel temperature of the FeF2. This coercivity reaches a peak value of over 600 Oe close to the Néel temperature and persists to above 300 K. The coercivity is correlated with the growth of an anisotropy in the ferromagnet, the increase of the antiferromagnetic susceptibility and the increase of the ferromagnetic resonance linewidth. We argue that the growth of spin fluctuations in the antiferromagnet leads to an enhanced ferromagnetic anisotropy, and therefore coercivity, above the Néel temperature.

Journal of Physics D: Applied Physics, 2006

The influence of an imperfect interface on exchange bias (EB) properties is investigated. Within the framework of the domain state model, the EB field H EB and the coercive field H C are determined using computer simulations, and they are found to depend strongly on the details of the interface structure. This dependence is sensitive to the dilution of the antiferromagnet (AFM) with non-magnetic defects in the bulk. For the optimal interface structure, giving greatest EB, the optimal dilution is found to be much less than that for an ideal-interface system, taking a value in better agreement with experimental results. Even without any defects in the bulk of the AFM the interface roughness leads to EB for thin antiferromagnetic layers, in accordance with the model by Malozemoff. Finally, the thickness dependence of rough-interface systems is found to differ significantly from that of ideal-interface systems.