Surface-anisotropy and training effects of exchange bias in nanoparticles with inverted ferromagnetic-antiferromagnetic core-shell morphology

A modified Monte Carlo method using the Metropolis algorithm is performed to simulate the hysteresis behaviors of the nanoparticles with an inverted antiferromagnetic (core)/ferromagnetic (shell) morphology at low temperature after field cooling. We have examined the dependence of exchange bias on the hard ferromagnetic surface anisotropy and the training effect. Our simulations reveal that, besides the antiferromagnetic core, another pinning source, namely, the hard ferromagnetic surface, can also contribute to the exchange bias in such a special structure. Above a critical surface anisotropy, the exchange bias field has a steep increase by means of the change of the magnetization reversal mechanisms, which are affected by the surface anisotropy. During the consecutive hysteresis loops, the exchange bias field decreases gradually to a constant value. The phenomena have been interpreted well by considering the combination of locking, releasing, and stabilizing of the spins on the antiferromagnetic core surface and the energy competition between Zeeman and antiferromagnetic anisotropy. Our results are in good agreement with the experimental findings. (C) 2011 American Institute of Physics. [doi: 10.1063/1.3611417]