Exchange bias and magneto-resistance in an all-oxide spin valve with multi-ferroic BiFeO3 as the pinning layer
Modern microscopy techniques indicate that the electrical switching of magnetic domains in multi-ferroic materials is possible. However, the application of such functionality in a real device has yet to be proven. In this work we fabricated an all-oxide spin valve with the ferroelectric anti-ferroma...
Gespeichert in:
Veröffentlicht in: | Acta materialia 2013-11, Vol.61 (19), p.7444-7453 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Modern microscopy techniques indicate that the electrical switching of magnetic domains in multi-ferroic materials is possible. However, the application of such functionality in a real device has yet to be proven. In this work we fabricated an all-oxide spin valve with the ferroelectric anti-ferromagnet BiFeO3 (BFO) as the pinning layer. The multi-layered heterostructure was grown epitaxially on a (001) SrTiO3 substrate and magneto-resistance was achieved at room temperature, which was switchable magnetically in a similar way to conventional metallic spin valves. Some key physical and material issues for building up such a novel device were addressed, in particular the hetero-epitaxy-induced strain effects on the electrical and magnetic properties of each layer and the establishment of exchange bias between BFO and an oxide ferrimagnet, e.g. Zn0.7Ni0.3Fe2O4 (ZNFO). The strains caused a significant increase in the coercivity but a decrease in the saturation magnetization of the ferrimagnet used. The former is particularly undesirable because it increases the required switching field. The all-oxide architecture allowed the spin valve to be field annealed from a temperature above the high Néel point of BFO (∼660K), after which a very large exchange bias field (Hex) was achieved at 5K and kept at a decent value at room temperature. The Hex–T curve did not follow the widely observed (1−T/TN)β temperature dependence, but could be explained by the random field model with one-dimensional (1-D) anti-ferromagnetic sublattice magnetization derived from the spin wave theory. Based on the observed 1-D spin wave behavior and the geometric arrangements of the paramagnetic ions at the (001) surface we propose an atomic model in which only a part of the spin along the diagonal lines in the BFO (001) surface was strongly exchange coupled with ZNFO. |
---|---|
ISSN: | 1359-6454 1873-2453 |
DOI: | 10.1016/j.actamat.2013.08.054 |