Magnetic Imaging in the Presence of an External Field:
Erasure Process of Thin Film Recording Medium
R. D. Gomez, I.D. Mayergoyz and E.R. Burke
Department of Electrical Engineering, University of Maryland at College Park 20740
and Laboratory for Physical Sciences, College Park, MD 20742
Abstract --- Microscopic magnetic processes occurring in the
presence of an external field have been observed with sub 100
nm resolution. Force gradient contours obtained at
progressively increasing fields, showed the local evolution of
thin film recording patterns while undergoing DC erasure. The
images exhibited characteristic changes that can be attributed to
the combined effects of probe and sample magnetization reorientation. Probe-induced image variations were first clarified
in order to stress the sample-induced contrast variations. The
microscopic behavior of the patterns were then directly
correlated with the sample's macroscopic state as defined by the
location on the magnetization curve.
I. INTRODUCTION
Recently, the distributions of remanent magnetization
structures of a thin film recording medium have been studied
while undergoing successive stages of erasure.[1] In this
work, we extend those investigations by imaging the actual
evolution of the patterns while experiencing an external
magnetic field. The goal is to relate the microscopic spatial
reordering of the magnetic moments of the medium with the
specific points along the magnetization curve.
The set-up exposed both the sample and the probe to the
same external field. It is then worthwhile to first assess the
influence of probe-induced effects on the image quality and
distinguish those artifacts from the actual transformations of
the sample. We subsequently present and discuss the
different stages that the medium has undergone en route to
bringing most of its magnetic moments aligned in the
direction of the applied field. The results and their discussion
are focused principally on a number of key points, namely, (i)
the expansion of the bright-contrast domains at the expense
of the dark-contrast domains, (ii) the increase in transition
edge ripples across the track, (iii) the "re-texturing" of the
dark contrast areas, (iv) the coalescence of neighboring
expanding domains, which causes (v) the breakup of the
unfavorably magnetized regions across the track and (vi) the
eventual collapse of the unswitched clusters at the highest
fields. The balance of this section is devoted to clarifying
these observations.
II. EXPERIMENT
An electromagnetic element was incorporated in a
commercially obtained magnetic force microscope (Digital
Instruments Nanoscope III). It was designed to have a small
gap which was positioned directly underneath the cantilever
probe during scanning. The sample was placed inside this
gap region and experienced an external magnetic field which
was assumed to be oriented primarily along the surface plane
and aligned in the direction of the recorded tracks. The
average strength of the external field was calibrated against
Manuscript received February 17, 1995.
the input current by using a miniature Hall probe and could
reach as high as 3000 Oe. Care has been taken to ensure that
no appreciable Joule heating occurred and that the
fluctuations in the applied field was kept to a minimum.
Space limitations in this paper preclude extensive discussion
of the design but will be furnished elsewhere. The sample
was prepared from a conventional removable rigid disk
whose characteristics have previously been studied.[1]
Standard magnetization curves were measured using a
vibrating sample magnetometer and the following
macroscopic properties were obtained: Hc = 926 Oe, Mr=
5.0 memu/cm2, S = 0.81, S*=0.86, and the in-plane
orientation ratio Mr(parallel)/Mr(perpendicular)~1.5. In
addition, the medium exhibited positive inter-particle
interaction. The present data consisted of nearly 100 images
taken in succession, each of which is associated to a fixed
external field.
III. RESULTS AND DISCUSSION
Fig. 1 shows the sequence as the magnetic moments
experienced a progressively stronger external field. The M-H
loop for this medium is shown at the center. Since only the
average gap field could be measured with the Hall probe, a
separate determination of the local field at the scan location
was done by assigning the coercive field to the image which
showed roughly 50% switching of the dark contrast areas.
The rest of the images were then identified relative to this and
their positions were labeled on this loop.
The narrow strips of alternating bright and dark contrasts
in (A) mark the locations of the transition regions joining two
magnetized areas in either head-to-head or tail-to-tail
configurations. With slight application of an external field
the contrast in the interior regions of the patterns intensified
as shown in (B) and the contrast was no longer confined to
the transition regions but broadened to resemble the
magnetization distribution. Since a field of 100 Oe was too
low for this medium to exhibit perturbations, this
transformation can be regarded solely as a manifestation of
probe-induced effects.
This can be understood by
considering an arbitrarily oriented probe. The contrast
mechanism of the magnetic force microscope, neglecting
probe-sample electrostatic forces, is given by,[2]
 2 Hi
F'   mi
 z2
(1)
where F' is the change in force gradient and m i and Hi
are the Cartesian components of the probe magnetization and
the surface field, respectively. Prior to imaging, the probe
was exposed to a 3.5 kOe magnet which aligned its
magnetization normal to the sample surface direction. Thus,
without any external field, the probe had selected this
component
of
the
Fig. 1. Microscopic pattern evolution correlated with its magnetization state. MFM images of a specific 65 x 65 m2 area on a thin film
recording medium under the action of an increasing DC magnetic field.
field. Image (A) ostensibly exhibits the locations of the free
magnetic poles at the transition regions, i.e., the z
component of the surface field. Upon applying a field, the
probe magnetization has reoriented toward the new direction
and has lead to a non-zero mx term in (1). Hence in (B), the
contrast had broadened as a consequence of an increased Hx
contribution. At larger fields, we therefore anticipate that
the probe would be increasingly diverted along the xdirection and this will simply result in an enhanced
sensitivity to the horizontal field component. The horizontal
component of the surface field more or less mimics the
lateral spread of the field producing component of the
magnetization,[3] so that this apparent drawback offers the
advantage in interpreting the images as that of the local magnetization distribution.
It is worthwhile to point out that while the external field
forces imaging sensitivity to a specific component,
application of a uniform field across the entire scan area will
not introduce extraneous local effects.
Hence, with
judicious design, one may be able to adapt the MFM as a
component selective imaging tool.[4]
Similarly, by
constraining the probe magnetization in a rigidly fixed
orientation, one can avoid the complications of a changing
probe in image interpretation. This approach is ideally
suited in cases where the sample magnetization is orthogonal
to the applied field and has a high anisotropy field, or when
the coercivity of the sample along the applied field is much
higher than that of the probe. The latter condition is true in
the present case.
The bright areas in these images correspond to the
magnetization parallel and oriented favorably with the
applied field. The overall high frequency background noise
has not been suppressed in order to provide a qualitative
sense on instrumental limits. Under these conditions, the
relative texture of the bright and dark areas in (B) are very
similar to each other.
This means that the surface
magnetization was equally saturated in one direction as it
was in the other. The transition region which separated
opposite magnetized regions were more or less straight with
the exception of slight spatial variations across the track.
These edge ripples arose from the combined effects of
background noise as well as the intrinsic zigzag and localized vortex formations at the transition region that tend to
minimize the magnetostatic energy.[5]
As the field was increased, no discernible effects on the
sample have been observed, consistent with the M-H loop
which shows that the magnetization was little changed from
its saturation value for small external fields. At 696 Oe,
however, changes in the patterns became apparent. The
subtle expansion of the bright areas and the distinctive
roughening of the transition region signify the initial stages
of magnetization reversal. The increase in transition edge
irregularity suggests that the leading edges of the favorably
magnetized ripples, in response to the applied field, have
extended into areas previously magnetized in the opposite
direction. Direct comparison of the bit lengths in images
(B) and (D) confirms that transition roughening intensified
and that the bright regions have extended by an average of
about 1 m, accompanied by a corresponding remission of
the dark contrast areas. In addition to these, the interiors of
the dark areas have significantly coarsened at 865 Oe in
comparison with the relatively unchanged interiors of the
bright areas. This strongly suggests that the magnetization
reduction at the unfavorable areas were accomplished by
means of the local reorientation of magnetic moments, not
only in the transition areas but over their entire regions.
This is in qualitative agreement with the results of
micromagnetic simulations at various states along the
hysteresis loop [6] and where the observed roughening may
be attributable to the formation of a large number of vortex
structures. Within instrumental limits, the local
redistribution did not appear completely random which
would have produced uncorrelated patches on the surface.
Instead, there seemed to be a tendency to form somewhat
triangular microstructures in the interior regions that
extended up to the transition regions. Finally, within this
range of applied field, we found evidence that moments at
the track edges similarly switch in the direction of the field.
This can be perceived as the slight widening of the nonmagnetized portions which separate the tracks between the
images (B) and (D). This implies the relative ease by which
the cross track magnetization components were reoriented
by an average angle of 90o.[4,7] These in turn, facilitated
the early reorientation of adjacent interior moments due to
positive exchange interaction. The overall result was an
apparent shortening of the track width.
As the field approached coercivity, adjacent growing
domains have begun to break-through the entire length of
the unfavorable domains. Localized areas, where large
neighboring transition ripples edge ripples have coalesced,
caused the breakup of the unfavorably magnetized
areas.[6,8] This can be seen at 866 Oe where light-intensity
streaks appear in most of the short-bit-length dark contrast
areas. In sections where this occurred, the brightness
intensity at the transition region has been diminished. This
is a consequence of the disappearance of free magnetic
charges at the transition regions when neighboring
identically magnetized ripples combine. At 926 Oe, the
streaks have similarly appeared on the wider dark areas and
have considerably expanded laterally in some locations. We
note however, that as soon as this manner of magnetization
reversal occurred, the average location of the transition
boundaries have remained stationary. In other words, the
average length of the patterns, neglecting the broken-through
areas, had stopped contracting. This implies that the system
found a new channel to reverse magnetic moments that was
energetically more favorable than the displacement of transition boundary. This mechanism prevailed for fields at and
above the coercivity. Starting at 865 Oe, magnetization
reversal had occurred predominantly by the widening of the
streaks along the cross track direction. By 1027 Oe, most
parts have been reoriented in the direction of the field, with
the exception of isolated remnants of unswitched clusters.
These pinned areas persisted tenaciously and required fields
significantly higher than coercivity to completely switch
them. The terminal distribution of these remaining clusters
appear to be uncorrelated with their previous patterns and
their random distribution may have reflected the local
fluctuations in the magnetic properties of the medium.
CONCLUSIONS
The capability to produce magnetic force microscopy
images under the action of an external magnetic field has
been realized and has been applied to study the fielddependent progress of magnetization reversal of thin film
media. Characteristic image alterations introduced by probe
effects were observed at very low fields and were explained
on the basis of the probe's magnetization reorientation. The
intermediate steps leading to full medium saturation were
determined by following the evolution of the patterns
starting from the initial displacement of the transition
boundary and terminating by the switching of isolated
clusters. These observations help elucidate the nature of
magnetization reversal in granular systems.
1. R.D. Gomez, L. Lising, R. Madabhushi, E.R. Burke and I.D.
Mayergoyz, J. Mag. Mgn. Mat. in press.; Conf. Digest Intermag '94,
Albuquerque, NM
2. H.J. Mamin, D. Rugar, J.E. Stern, B.D. Terris and S.E. Lambert, "Force
Microscopy of Magnetization Patterns in Longitudinal Recording Media",
Appl. Phys. Lett. Vol. 73, pp. 1564, October 1988.
3 D. Rugar, H.J. Mamin, P. Guethner, S.E. Lambert, J.E. Stern, I.
McFadyen and T. Yogi, "Magnetic force microscopy: General principles
and application to longitudinal recording media", J. Appl. Phys. Vol. 68,
pp. 1169-11-83, Aug. 1990.
4. R.D. Gomez, "Component-resolved imaging of the surface magnetic
fields", J. Appl. Phys. Vol. 75, pp. 5910-5912, May 1994.
5. H.N. Bertram, Theory of Magnetic Recording, Cambridge: Cambridge
University Press, pp. 207 1994 and references therein.
6 J.G. Zhu and H.N. Bertram, "Micromagnetic studies of thin metallic
films", J. Appl. Phys., Vol. 63, pp.3248-3253, April 1988.
7 J.L. Su and K. Ju, "Track Edge Phenomena in Thin Film Longitudinal
Media", IEEE Trans. Mag., Vol. 25, pp.3384-3386, Sept. 1989.
8. J.J. Miles and B.K. Middleton, "The Effect of Trackwidth on Transition
Noise in Longitudinal thin-film media", J. Mag. Mag. Mat.. Vol. 120, pp.
376, 1993.