Browse Prior Art Database

CPP Sensor With In-Stack Longitudinal Biasing of Free Layer Using Self-Stabilized LBL2 Extending Beyond Sensing Region Disclosure Number: IPCOM000016431D
Original Publication Date: 2003-Feb-28
Included in the Prior Art Database: 2003-Jun-21
Document File: 5 page(s) / 130K

Publishing Venue



Disclosed is an alternative in-stack longitudinal bias scheme for forming a resetable in-stack longitudinal bias stack using synthetic anti-parallel (AP) coupling through a thin conducting layer such as Ru and the shape anisotropy of ferromagnetic materials. Prior art in-stack bias schemes uses a second AFM and a ferromagnetic longitudinal bias layer (LBL), resulting in about 250 Angstrom of additional sensor thickness. Prior art also requires a complicated thermal annealing process to set the two AFM layers orthogonally. Also weaker magnitude of the AFM-FM exchange coupling in prior art (compared to the stronger antiferromagnetic exchange coupling across Ru in the present invention) results in a smaller range of LBL thicknesses possible. In our invention, the sense layer is stabilized by a LBL layer through magnetostatic interactions, as in the prior art. Here, however, the LBL layer is pinned through strong AP coupling through the Ru spacer to the LBL2 layer. The layer closest to the sensor free layer is patterned with the sensor stack to form a longitudinal bias layer. The other layer in the bilayer is formed into a large aspect-ratio shape which is self-stabilized due to shape anisotropy, and also stabilizes the longitudinal bias layer through antiferromagnetic exchange coupling across a thin metallic layer. The resulting in-stack biased sensor can be advantageously fabricated by this method with a much smaller shield-to-shield gap spacing, and be easily magnetically reset without the need for a high-temperature annealing step. Micromagnetic calculations have been performed which demonstrate the magnetic performance of the resulting sensor structure.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 42% of the total text.

Page 1 of 5

  CPP Sensor With In-Stack Longitudinal Biasing of Free Layer Using Self-Stabilized LBL2 Extending Beyond Sensing Region

  In the prior art (US Patent: US 6,023,395), the longitudinal bias layer (LBL) is formed by using an antiferromagnetic layer (afm2 in Fig.1a) in contact with a ferromagnetic layer as shown in the Fig.1a. As a result, two antiferromagnetic (AFM) layers have to be used: the first one for pinning the reference layer of the MTJ sensor and the second one for pinning the LBL for longitudinal stabilization. In such a scheme, a complicated two-step thermal annealing process is required to set the pinned layer into the transverse direction (as shown by the arrow in Fig.1a) but the LBL into the longitudinal direction (as shown by (X) in Fig.1a). Furthermore, the 2nd AFM layer (afm2 in Fig.1a) used in the longitudinal bias stack is usually as thick as 200A. This additional thickness will significantly impact the fabrication of narrow gap sensors. Moreover, the LBL cannot be reset after the process unless a thermal annealing combined with a high magnetic field is employed.

In this invention, the afm2 in Fig.1a is replaced by a LBL2 and a thin layer of metal such as Ru (as shown in Fig.1b) known for inducing strong antiferromagnetic coupling when inserted between two ferromagnetic layers. The LBL2 is a layer of a thin ferromagnetic material (such as Co, CoFe etc) with a large dimensional aspect ratio of track width divided by stripe height (> 3.0). After initialization with a large longitudinal field, the mid-region of the LBL2 is prepared into a self-stabilized single-domain state with a longitudinal remanent magnetization. This longitudinal state induces a stable antiparallel longitudinal state in the LBL through the thin metal (Ru..) spacer. Since the stability LBL2 is derived from its shape anisotropy, it can be reset along either longitudinal directions (x or -x) by a large enough longitudinal applied field. The longitudinal bias stack in our present invention is therefore resetable at room temperature without any thermal annealing. Furthermore, the AF coupling across the thin Ru spacer is usually strong enough to support a LBL of considerable thickness as needed for better stabilization of the sense layer. Finally, the total thickness of our longitudinal bias stack is typically (30A (CoFe as LBL2) + 8A (Ru)+25A(CoFe as LBL)=63A) much thinner than that of the prior art(200A(PtMn as afm2)+25A(LBL)=225A), rendering the fabrication of thin gap read heads much more feasible.


Page 2 of 5

Figure 1. The ABS views of the two In-stack bias schemes

top lead afm2 LBL non-magentic spacer free layer MTJ barrier pinned layer
bottom lead

top lead

Ru LBL non-magentic free layer MTJ barrier pinned layer afm1 bottom lead

LBL2 long

long. bias stack

MTJ sensor

long. bias stack

MTJ sensor

a) Prior art US 6,023,395 b) our alternative scheme

  The LBL2 layer itself can be shown to be self-stabilized at a
large enough aspect ratio (>3) and a...