Browse Prior Art Database

Controlling Exchange Anisotropy in Biased Magnetic Devices

IP.com Disclosure Number: IPCOM000120859D
Original Publication Date: 1991-Jun-01
Included in the Prior Art Database: 2005-Apr-02
Document File: 3 page(s) / 123K

Publishing Venue

IBM

Related People

Gambino, RJ: AUTHOR [+4]

Abstract

A technique is described whereby exchange anisotropy is used in the control of domain structures of magnetic devices and provides a means to control exchange bias effects.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 48% of the total text.

Controlling Exchange Anisotropy in Biased Magnetic Devices

      A technique is described whereby exchange anisotropy is
used in the control of domain structures of magnetic devices and
provides a means to control exchange bias effects.

      Exchange anisotropy offers a way of controlling domain
structures in magnetic devices, such as thin film magnetoresistive
heads (1).  In prior art, an antiferromagnetic layer, (FeMn), in
contact with a ferromagnetic layer, (NiFe), was used.  However,
difficulties have been experienced in predicting the size of the
effect, since other parameters might be important in controlling the
size of the effect.

      The concept described herein provides four key requirements for
controlling exchange anisotropy, so as to provide a means of
fabricating magnetic devices in a consistent manner.  The four key
requirements are as follows:

      (1)  Atomistically Rough Interface - Methods of film
preparation can involve low temperature deposition and minimization
of surface bombardment, so as to reduce surface mobility, etc.
Equilibrium surfaces, in which roughness height W scales as 1nL,
where L is the lateral distance, should be avoided because steps are
relatively rare.  In principle, smooth interfaces could give an
effect of the same order as rough surfaces if they were
uncompensated.  However, such interfaces are difficult to make and
control.  Rough surfaces are more forgiving to defects and
reconstructions and, therefore, are better for consistent devices.

      (2)  Large Coercivity and Rotational Hysteresis in the
Antiferro Layer - This insures that the antiferromagnetic domain,
formed under the influence of the random interface exchange, does not
move as the ferromagnet's moment is reoriented.  The "staggered"
coercive field required is the same as calculated in the computation
in requirement (3b) below.  Large coercivity is assured by the grain
structure of the antiferromagnetic layer, by large local anisotropy,
by large magnetostriction and generally by imperfections of the
antiferromagnetic layer.  The layer could also be structured, as
described later in (4).

      (3)  Large Anisotropy - There are three cases to consider,
depending on the spin structure of the antiferromagnet:

      (a)  Simple Colinear Structure - In this case, an in-plane
uniaxial anisotropy is required both in the ferromagnet and the
antiferromagnet to prevent spin flops and to determine an
antiferromagnetic domain wall width f  A/K, where K is the anisotropy
and A is the exchange stiffness.  For example, Co hcp has such an
anisotropy and could be the mechanism for Co-CoOx interfaces.
Materials with the largest uniaxial anisotropies are most favorable.

      (b)  Non-Colinear Structure - e.g., the <111> structure of FeMn
(2).  In this case, cubic anisotropy and/or in- plane anisotropy will
be satisfactory.  In addition to choosing a material with large
intrinsic cubic anisotrop...