Four-probe magnetoresistance of current-perpendicular-to-plane structures
- PDF / 210,020 Bytes
- 12 Pages / 612 x 792 pts (letter) Page_size
- 1 Downloads / 182 Views
0906-HH01-07.1
Four-probe magnetoresistance of current-perpendicular-to-plane structures Hua Zhou*, M. Covington, and Michael A. Seigler Seagate Research, 1251 Waterfront Place, Pittsburgh, PA 15222 Abstract The resistance and magnetoresistance (MR) of three-dimensional currentperpendicular-to-plane (CPP) structures have been calculated via numerical finite element solutions of the Laplace equation. This model accounts for the non-uniform current paths in a four-probe geometry that can yield MR that differs from the intrinsic MR of the isolated CPP pillar with spatially uniform current flow. We calculated the four-probe MR for various geometries and resistivities of both the normal metal leads and the magnetoresistive pillar. From a single, unified approach, we are able to consistently account for the disparate behavior that has been previously published. In particular, we identify conditions that produce four-probe MR that differs from the intrinsic MR of the CPP pillar and highlight those situations where the four-probe resistance is negative. Finally, we present a simple analytical formula for the MR ratio that is applicable to narrow CPP pillars with wide, thin leads.
Introduction Current perpendicular to plane (CPP) magnetoresistance (MR) devices are believed to give higher MR ratio [1, 2]. However, these devices were not previously considered for use as read head sensors because of their large device area resulting in small resistance, which in turn leads to small signal. However, as lithography technology advances, CPP devices with sub-micron sizes can be made, greatly increasing device resistance, making CPP devices a viable option for next generation readback sensors [3]. One of the major concerns in measuring the room-temperature MR effects for CPP GMR devices as well as TMR devices is that the measured MR ratio may not represent the intrinsic MR ratio. Lenczowski et al [4] showed that in general current distribution effect needs to be corrected for CPP devices with pillar height smaller than the pillar width. Moodera et al [5] observed spurious large MR ratio and negative 4-point resistance for a TMR device in a cross leads geometry, where the junction area is determined by the line width of the leads. Their results are consistent with earlier findings by Pedersen & Vernon for similar devices [6]. Van de Veerdonk et al [7] explained the results of Moodera et al by a 2D finite element model. For a similar TMR device but with junction size smaller than leads line width, Matsuda et al [8] observed a decrease of MR ratio with increasing area in addition to the above spurious large MR ratio and negative 4-point resistance phenomena at larger area end. The corresponding RA product first increases with area and then decreases to negative value. Chen et al [9] observed similar decrease of MR ratio with increasing area for junction size much smaller than leads width. For CPP GMR devices, Bussmann et al [10] also observed spurious large measured MR ratio relative to the intrinsic value. Spallas et al [2, 11] had
Data Loading...
