Size and shape dependences on magnetization reversal in ferromagnetic/antiferromagnetic bilayer patterned into nano-dot arrays

Abstract

Magnetization reversal in exchange biased Co/CoO dot arrays has been investigated systematically as a function of shape anisotropy, dot size, temperature, cooling field and training effect. Polycrystalline 24 nm Co films have been sputtered on a pre-patterned Si/SiO2 substrate with the lateral dimensions of the dots varying from 200 to 900 nm at a fixed spacing of 800 nm. The Co dots have been oxidized using pure oxygen to create a Co/CoO bilayer. Below the Néel temperature, the hysteresis loop of this bilayer displays a normal unidirectional shift due to the exchange bias in the continuous film, while the patterned film displays an additional anomaly of the upper part of the hysteresis loop. This anomaly, which is shape dependent, increases with decreasing temperature and cooling field. On the other hand, it decreases for the trained loops with increasing aspect ratio (length/width) of the patterned dot. This anomaly is probably caused by incomplete biasing due to the competition between magnetostatic inter-dot interactions and exchange bias.

Exchange bias in core-shell iron-iron oxide nanoclusters

“…This type of loop has also been observed in Co/CoO dot arrays and was considered as shape anisotropy (rectangles, squares, circles) dependant. 23 The upper part of the anomaly appeared due to the competition between the magnetization reversal and the dipolar (magnetostatic) interaction. A cut-away spin configuration and a core-shell structure are shown in the inset of Fig.…”

Response to “Comment on ‘Magnetization reversal in europium sulfide nanocrystals’” [Appl. Phys. Lett. 91, 026102 (2007)]

“…[3][4][5][6][7] Other groups have observed experimentally size-dependent effects in the magnetization of nanomaterials. [8][9][10] The statements made by the authors of the Comment provide no insight into that possibility. As well, since their premise is founded on calculations that only refer to bulk paramagnetic materials ͑Pt and Dy 2 O 3 ͒, their comparisons do not appear to be appropriate for our conditions.…”