Citation
A scanning tunneling microscope adapted for magnetic force microscopy

Material Information

Title:
A scanning tunneling microscope adapted for magnetic force microscopy
Creator:
Rice, Paul Scott
Publication Date:
Language:
English
Physical Description:
vii, 36 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Scanning tunneling microscopy ( lcsh )
Hard disks (Computer science) ( lcsh )
Hard disks (Computer science) ( fast )
Scanning tunneling microscopy ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 35-36).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Department of Mechanical Engineering.
Statement of Responsibility:
by Paul Scott Rice.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
22937114 ( OCLC )
ocm22937114

Full Text
A SCANNING TUNNELING MICROSCOPE ADAPTED FOR
MAGNETIC FORCE MICROSCOPY
by
Paul Scott Rice
B.F.A., University of Colorado, 1975
B.S.M.E., University of Colorado, 1988
A thesis submitted to the
Faculty of the Graduate School of the
University if Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Department of Mechanical Engineering
1990


This thesis for the Master of Science degree by
Paul Scott Rice
has been approved for the
Department of Mechanical Engineering
by
William Clohessey
Date


Rice, Paul Scott (M.S., Mechanical Engineering)
A Scanning Tunneling Microscope Adapted for Magnetic
Force Microscopy.
Thesis directed by Dr. John Moreland of The National
Institute of Standards and Technology.
We have adapted a scanning tunneling microscope
(STM) to record and image magnetic regions on the
surface of a computer hard disk. The usual rigid Ptlr
tip of the STM was replaced by a flexible Fe film tip.
Tunneling images thus obtained were combinations of the
surface topography and variations in the magnetic force
between the Fe film tip and the disk surface. The
sensitivity of the Fe film tip was such that the
magnetic force influence on the image was as prevalent
as the sample surface topography. We present images of
a hard disk surface showing bit tracks written by a
ferrite head in a computer disk drive and images of
magnetic regions that we recorded with our method. The
images shown are comparable to other magnetic images of
textured surfaces such as ferrofluid decoration. Images
of computer written bits show the misalignment of bits
across the bit track separation. We believe that the
recording process relied on maintaining the proximity of
the magnetized-Fe film tip near the disk surface.
Apparently, the magnetic field was focused near the Fe
film tip with sufficient intensity to change the surface


iv
magnetization of the disk. We have recorded spots of
the disk within a 500 nm X 500 nm area. These spots
were subsequently imaged with the same tip.
The form and content of this abstract are approved. I
recommend its publication.
Signed
/ John Moreland


V
CONTENTS
CHAPTER
I. INTRODUCTION.............................. 1
II. THE SCANNING TUNNELING MICROSCOPE .............. 3
Magnetic Force Microscopy .................... 7
III. TUNNELING STABILIZED MAGNETIC FORCE MICROSCOPY 9
Tip Fabrication ............................. 10
Characterization of the Triangular film .. 12
IV. COMPUTER WRITTEN BITS ......................... 17
V. TUNNELING STABILIZED READ/WRITE PROCESS .... 24
VI. CONCLUSIONS ................................... 31
REFERENCES ........................................ 3 5
APPENDIX
A. DERIVATION OF PRIMARY FREQUENCY
37


FIGURES
vi
FIGURES
Figure
2.1. Diagram of STM sample stage ..................... 6
3.1. SEM micrograph of the apex of the flexible
Fe triangle .................................... 11
3.2. Drawing of the MFM tip assembly ................ 13
3.3. SEM micrograph of the MFM tip assembly ......... 14
4.1. TSMFM image of the hard disk showing
the magnetic structures "bits" ................. 19
4.2. STM image of hard disk surface ................. 20
4.3. Bitter pattern of the bits on the hard disk .. 22
5.1. TS magnetic images ............................. 25
5.2. Bitter pattern of TS written squares
on bit tracks .................................. 28
6.1. TS magnetic image of a 500 nm X 500 nm
magnetic spot .................................. 34


Vll
ACKNOWLEDGEMENTS
I would like to thank John Moreland for directing
the research and helping me understand the physics of
the experiments, the National Institute of Standards and
Technology (NIST) for providing funding and equipment,
Fred Frickett, Ron Goldfarb, and Jim Brug for insight
and discussions on the research.


CHAPTER I
INTRODUCTION
Scanning Tunneling Microscopy (STM) has been used
extensively to image the surfaces of electrically
conductive materials.1,2 Atomic structure of numerous
materials can be seen with the STM. Images of the
atomic structure of highly oriented pyrolitic graphite
are easily obtained. Reverse images of biological
samples can be made by evaporatively coating an organic
sample with Au then removing and imaging the Au film.
Surface images of metals obtained with the STM are
helping develop a better understanding of friction
between surfaces.
New applications of STM technology are developing
rapidly. We have demonstrated a novel application,
using the STM, to image and write magnetic regions on
the surface of a CoCrTa sputtered thin film hard disk.
Thus far we have written and imaged, with the same tip,
magnetic regions on the disk surface with a resolution
of 20 nm. Also we have imaged ferrite head computer
written "bits" on the same disk. We would like to
describe the construction and implementation of an


2
alteration to the standard STM to obtain images of
magnetic field variations on the surface of a hard disk.
Data storage capacity in computers is limited by
the magnetic characteristics of the thin film recording
media.3 The magnetic surfaces of thin film recording
media are naturally divided into areas of oriented
magnetic moments called domains. Computers write data
to the disk with an electromagnetic coil (the recording
head), orienting a number of magnetic domains in a
similar direction. These regions are the "bits". The
transition between the bits has a width of 1 to 2 /im
limiting the density of data storage. This transition
is not fully understood/'5 The primary reason for this
research was to help characterize these transitions. We
discovered that we could also change the magnetic
patterns of the disk thus creating our own magnetic
recordings.


CHAPTER II
THE SCANNING TUNNELING MICROSCOPE
The Scanning Tunneling Microscope scans an
atomically sharp tip over the surface of a sample. As
the tip is scanned, an electric potential is applied
between the tip and sample generating a tunneling
current. Passing over the sample surface roughness
changes the tip to sample distance. The magnitude of
the tunneling current is exponentially sensitive to the
change in distance from the sample. This change is then
plotted on a computer screen as a topological image
related to the tip position on the sample.
Tunneling current is the current developed from
the quantum mechanical jump of an electron from the
sample to the needle across an energy barrier higher
than the electron's energy. The tunneling current is
only developed on a nanometer range and the magnitude of
the current is exponentially proportional to the
distance from the sample.
I oc exp (-2kd)
Where d is the gap distance and k is the decay constant
for the wave functions in the barrier.1 The current


4
changes an order of magnitude for a 0.1 nm change in
distance d. In order for the current to be generated,
the surface to be scanned must be very conductive. A
small insulating layer (100 A) will prevent tunneling.
An insulating layer can be caused by chemical change of
the sample surface, such as oxidation. The electric
potential must be kept small to keep the current just a
tunneling current. As the voltage increases, there is a
chance electric field emissions could propagate from the
tip to the sample; such a current is not proportional to
the distance from the sample.
Our STM scans the surface of the sample with a
rigid Ptlr needle.6 Positioning the STM tip's X,Y, and
Z coordinates by electrical deflections of a
piezoelectric tube, the STM creates an image of the
surface by plotting change in tunneling current (related
to the Z component) against X and Y position. Operation
of the STM is in air at room temperature. The
instrument consists of a translation stage, a coarse
positioning/sample holder base and a data acquisition
and analysis system. The translation stage contains the
piezoelectric tube mounted in- an invar steel holder. The
Ptlr tip is mounted in an electrically isolated ceramic
holder on the end of the piezoelectric tube. The


5
translation stage is held magnetically to three
triangular spaced adjustment supports mounted in the
sample holder base. Coarse positioning of the needle to
the sample is achieved by a stepper motor, located in
the sample holder base, slowly adjusting one of the
adjustment supports thus lowering the translation stage
toward the sample. Once a tunneling current has been
established the stepper motor stops and the
piezoelectric tube positions the needle (Z coordinate)
to obtain the best tunneling current (see Fig. 2.1).
Scanning X and Y in a feedback mode, the needle
encounters varying surface features. As the needle
moves, a change in the tunneling current signals the
data acquisition/analysis system to reposition the
needle in the Z direction. The new position reflects a
change in surface topography. Two tunneling current
detection modes are available on our instrument: 1) a
height mode plots the relative elongation of the
piezoelectric tube in response to changing tunneling
current, and 2) a current mode plots the changing
tunneling current itself.
The images that the STM creates are plots of one
of the two modes as a function of tip position on the
sample. There are several image enhancement features in


6
i
I
i
!
Figure 2.1. Diagram of STM sample stage. Showing (A)
the piezoelectric tube (B) the coarse positioning screws
(C) the sample and (D) the placement of the tunneling
tip.


the software that we have used to evaluate our images.
These include (1) a low pass filter to remove spurious
high frequency laboratory noise, (2) cross sectioning,
used to find magnitudes of magnetic forces, (3)
flattening, to remove the effects due to piezoelectric
tube travel, and (4) image color change, to enhance the
picture produced on the black and white video printer.
Magnetic Force Microscopy
Microscopic magnetic force images of magnetic
thin film surfaces have been accomplished with several
different methods. Ferrofluid decoration, the Bitter
method, developed in the 1930's involves the application
of an aqueous dispersion of fine magnetic Fe particles
to a magnetic surface. The Fe particles subsequently
coalesce to the magnetic field divergence on the
surface, leaving dense collections of particles at the
magnetic boundaries and few particles between. The
surface is then viewed with an optical microscope.7
Lorentz Microscopy and Electron Holography detect
the magnetic deflection of electrons from a scanning
electron microscope (SEM) to produce an image of the
variation in the magnetic fields. As the electrons pass
through the magnetic film the magnetic forces attract or


8
repel the electron and change its path. It is this
change that is detected and plotted as a function of
position.8 Electron spin is also affected by magnetic
force. In scanning electron microscopy with
polarization analysis (SEMPA) the spin of electrons
after they have passed through a magnetic material is
contrasted against the spin of electrons that have not.9
Another method recently developed detects the
change in frequency of a vibrating magnetic probe using
an optical heterodyne interferometer. Here a magnetic
cantilever wire, sharpened to a point with a 500 nm
radius of curvature, is scanned, using a STM
piezoelectric stage, a few hundred angstroms above a
magnetic surface. The cantilever is vibrated at close
to its natural frequency. As a change in magnetic force
is encountered the frequency of the cantilever changes.
This frequency is detected and plotted against position
for a mapping of changing magnetic force.10,11


CHAPTER III
TUNNELLING STABILIZED MAGNETIC FORCE MICROSCOPY
Our method replaces the usual rigid Ptlr needle
in the STM by a compliant triangular Fe film. The Fe
film needle is scanned over the sample surface in the
usual STM manner and the tunneling current is detected
and analyzed. In our scheme, the tip of the Fe
triangle is attracted or repelled (dependant on the
direction of magnetization of the tip) by the divergence
of the field at the domain boundaries. If the tip is
attracted towards the sample surface the tunneling
current increases, signaling the feedback mechanism to
withdraw the needle. This movement is then plotted as a
rise in the topography of the sample surface.
Alternatively if the tip is repelled, the current
decreases signaling the feedback to move toward the
sample. This movement shows on the plot as a depression
in the surface topography. If there is no interaction
magnetically between the Fe film tip and the sample
surface, the usual surface topography is plotted. These
movements of the tip in response to tunneling current
are plotted as a combination of surface topography and


10
magnetic field variation. We call this process
tunneling stabilized magnetic force microscopy (TSMFM).
Tip Fabrication
The Fe film was made by evaporatively depositing
99.9% Fe on a glass slide. Upon cooling differential
shrinkage between the Fe film and the glass slide caused
the film to spontaneously separate. Fe films were
between 5 and 8 /xm thick.
The Fe films were cut in the shape of a isosceles
triangle with scissors or razor blade to form a sharp
tip. Figure 3.1 is a SEM micrograph of a tip showing a
radius of curvature of less than 0.5 /xm. The edges of
the film were slightly ragged resulting from the tearing
action of the scissors. The ragged edge provided many
small protrusions, providing a tunneling tip of much
smaller radius of curvature than observed optically.
The MFM tips achieved with this technique have so
far been superior to any other method that we have
studied. These methods included stretching a Fe wire
until it breaks, electroetching a Fe wire and a
triangular Fe film with 3.5% HC1 and a voltage of 1.5
volts for 10 seconds, and grinding a Fe wire. In all
these methods the surface of the film and the wire were


11
Figure 3.1.
Fe triangle
SEM micrograph of the apex of the flexible
showing possible tunneling points.


12
pitted and the tip was rounded to a point of
approximately 1 lira as viewed under an optical
microscope.
The film was attached to a support wire with
silver paint (see Fig. 3.2 and Fig. 3.3). Current
travels from the STM electronics to the tunneling gap
through the interface between the support wire and film.
Silver paint was used because it is a good electrical
conductor and bonds the Fe film securely. Several
different configurations of film to wire were tried with
varying results. As the tip approached a vertical
position, the resolution of the image was increased
although the sensitivity to magnetic forces decreased.
This decrease in magnetic sensitivity was due to the Fe
triangle not being able to move in the Z direction as
easily as it could if it were horizontal. The increase
in image resolution was due to the film presenting less
of the broad side surfaces to the tunneling gap.
Characterization of the Triangular film
A triangular shape was chosen to give lateral
stability to the needle holding the needle rigid in X
and Y. As the disk surface was scanned the needle could
be deflected in three simultaneous directions, normal to


Figure 3.2. Drawing of the MFM tip assembly.


14
Figure 3.3. SEM micrograph of the MFM tip assembly,
silver paint is holding the film to the support wire.
The


15
the surface reflecting the magnetic domain boundary
divergence, in either X or Y direction parallel to the
surface of the disk reflecting a force along the
magnetic field lines.
Estimates of magnetic forces that we wanted to
detect are on the order of 3 X 1010 N. Beam deflection
formulas were used to find the deflection of the
triangular film, assuming that the only force acting was
on the very apex of the triangle.12 A calculation for
the displacement of the triangle apex Z due to a force P
acting on the tip follows.
I(x) = xbh3 (1)
12L
M(x) = -Px (2)
dz = fL M(x) dx
dx J0 EI(x)
z(x) = -6Px3 L
Ebh3 0
(3)
(4)
Where b denotes the base of the triangular film, h the
thickness of the film, L the length of the triangle, P
the force in the Z direction at the triangle tip, M the
moment along the X axis, E the elastic modulus and I the
moment of inertia.
Also, in order that no spurious laboratory noise


16
would affect the image, it was determined that the
natural frequency of the triangle be over 10 kHz. Using
the solution for the triangle tip displacement, the
Rayleigh Method was used to find the primary natural
frequency 13 (see appendix for derivation).
w = (Eh2/^4)* (7)
Where w is the primary natural frequency and p is the
density of the material. The estimated force and
primary frequency were used to calculate the size of the
triangle. It was difficult, with scissors, to cut the
triangle to the exact size calculated but we found that
within a range of 1 mm to 2 mm for the base and height
gave satisfactory results.


CHAPTER IV
COMPUTER WRITTEN BITS
We studied sections of a thin film hard disk from
a computer disk drive, looking for the computer written
magnetization patterns. The disk was multilayered
sputter deposited films on an aluminum disk. Including
a NiP base layer, a 1 fim magnetic CoCrTa layer, and a
thin lubricating C layer. Typical coercivities for
reversing the CoCrTa film magnetization are 48 kA/m (600
Oe).14,15 Disk surface features due to manufacturing
surface preparations were in the range of 20 to 30 nm.
CoCrTa sputtered thin film disks are called longitudinal
magnetic storage media. The magnetic field lines in
longitudinal media are generally parallel to the surface
of the film except for the divergence at the domain
boundaries.16
While scanning the disk we used a tunneling bias
voltage of 1611 mV and a tunneling current setpoint of
0.49 nA. The STM was used in a constant height mode to
position the tip relative to the disk surface. Scanning
took place in air at room temperature. We scanned an
area of 53.4 iim on a side at 1 line per second. The STM


18
took 400 tunneling current samples per line, and 400
lines per scan.
The TSMFM image shown in Fig. 4.1 is a
combination of disk surface topography and magnetic
field variations. The small parallel grooves in Fig.
4.1 (running from the top left to the lower right), also
apparent in the STM images taken with a standard rigid
Ptlr tip, are scratches in the aluminum substrate that
occur during polishing (see Fig. 4.2). The polishing
marks are parallel to the direction of rotation of the
disk in the drive. Magnetic bit tracks are also evident
in Fig. 4.1 as broad parallel ridges about 5 nm wide
running roughly perpendicular to the polishing marks on
the surface of the disk.17 The slight angle is due to
the oblique position of the ferrite recording head,
while recording, to the rotation of the disk. The
separation between the bit tracks can be seen emphasized
by a slight shift in the bit position. This separation
runs parallel to the polishing marks and goes from the
middle of the left side of the figure to the lower right
corner.
The distance traveled'by the STM piezotube was
greater for magnetic forces than for tunneling current
adjustments due to surface features. The magnetic force


19
Figure 4.1. TSMFM image of the hard disk showing the
magnetic structures "bits". Separation of the bit tracks
is shown by a phase shift of the bits running from the
middle left of the figure to the lower right. The lower
section of the figure is a cross section showing the
magnetic displacement of the Fe film tip.



20
Figure 4.2. STM image of hard disk surface. The cross
section on the bottom shows the height of the surface
roughness.


21
of the bits on the Fe film tip can be calculated by-
multiplying the distance traveled by the STM as it
withdraws to keep the tip height constant times the
spring constant of the triangular Fe film. From a
cross-section of our TSMFM image (see bottom of Fig.
4.1) we measure a 20 nm displacement. This multiplied
by our estimate for the spring constant of 0.1 N/m gives
a force of 2.0 X 10'9 N. This coincides with our
original force estimate of 1 X 1011 N.18
Figure 4.3 is an optical micrograph of a Bitter
pattern obtained using ferrofluid to decorate the disk
surface. The similarity of the features draws us to
the conclusion that indeed we are detecting magnetic bit
tracks. Dimensions of the computer written tracks in
the ferrofluid image are 30 ij.m and the width of the bits
are 5.0 nm, corresponding to similar dimensions in the
TSMFM images.
An interesting effect was also observed while
TSMFM imaging disks with a Fe film tips without an Au
coating. When the tunneling current was increased, thus
decreasing the tunneling distance, we found these tips
changed the magnetization of the disk. This is a
disadvantage if the goal is to nondestructively image
existing magnetization on the disk surface.


22
Figure 4.3. Bitter pattern of the bits on the hard disk.
The bits are the parallel ridges vertical to the image.
The separation between the bit tracks are the bright
horizontal lines.


23
However, when the Fe film is coated with Au and has no
permanent magnet attached, there is minimal disturbance
of the magnetic bit tracks on the hard disk.


CHAPTER V
TUNNELING STABILIZED READ/WRITE PROCESS
In the process of trying to image existing
magnetization patterns, we discovered that we could
magnetize the disk in a specific region. This indicates
that the magnetization induced during the tunneling
stabilized (TS) magnetic recording process is caused by
the magnetic field at the tunneling tip held in
proximity to the disk surface. To our knowledge, this
is the first demonstration of magnetic recording using a
scanned TS probe and the first demonstration of a
scanned tunneling TS probe capable of both imaging and
recording on a magnetic medium.
We studied the same hard disk surfaces as in the
previous chapter. In some cases, we coated the surfaces
of either or both of the Fe film tunneling tip and the
hard disk with thermally evaporated Au (99.99%).
Figure 5.1 shows two TS magnetic images before
and after the TS magnetic recording process. TS
magnetic recording and imaging occurred in air and at
room temperature. Images consisted of 400 line scans
with 400 samples per line scan. The images were taken


0000 3 COCO -30000 J00C0
o
fl)
tr
Figure 5.1. TS magnetic images taken before (a) and
after (b) the TS magnetic recording process on the
surface of a computer hard disk. A cross sectional view
(c) of a TS magnetic "crater" is shown below.
w
U1
40000


26
at a rate of 2.5 line scans per second. The tunneling
current and voltage were 0.10 nA and 437 mV. The record
in Fig. 5.1b was created when the STM was zoomed-in to a
8 (im X 8 /Jin square, the scan rate was decreased to 1
line scan per s, and the tunneling current was increased
to 0.40 nA, keeping the tunneling voltage constant. The
STM was then zoomed-out to the original scanning range
of 53 ^m X 53 /xm before generating the image in Fig.
5.1b. The images, taken in height mode, showed the
relative elongation of the piezoelectric tube scanner in
response to changes in surface topography and forces on
the Fe film STM tip. Notice the square crater created
during the TS magnetic recording process in Fig. 5.1b.
The depth of the crater is about 50 nm. An increase in
any of three factors typically resulted in an increase
in the depth of TS magnetic craters: (1) the strength of
the magnetization in the iron film; (2) the time
duration of the TS magnetic recording process; and (3)
the tunneling current.
The Bitter technique was also used to image the
magnetic records created using the TS magnetic recording
process. A disk section with' three square TS magnetic
craters, each spanning the full STM scanning range of 53
fim X 53 /xm, was decorated with a ferrofluid and viewed


27
under a microscope (see Fig. 5.2, the squares appear in
the center of the micrograph). The area surrounding the
squares is covered with tracks of magnetic "bits"
originally recorded with one of the ferrite heads in the
disk drive. Portions of these bit tracks were erased
during the TS magnetic process. TS magnetic squares
were also recorded on portions of the disks that did not
have bit tracks. These squares were also visible in an
optical microscope having ferrofluid decorations
concentrated at their edges. Such observations are
further evidence that the features created during TS
magnetic recording are indeed magnetized sections of the
disk surface and that the crater in Fig. 5.1b is due to
the magnetic force between the Fe film tip and the disk
surface.
The topography of the surface showing the
polishing grooves usually found on the surface of a hard
disk. Never exceeding 30 nm, the surface topography is
smaller than the 50 nm crater depth shown in Fig. 5.1b.
We also found that increasing the tunneling current to
values as high as 30 nA while using the rigid Ptlr tip,
causing the STM tip to drag on the disk surface, did not
alter the surface topography more than a few nanometers.
In addition, STM imaging using a rigid Ptlr tip does not


Figure 5.
tracks.
magnetic

bit
TS


29
leave any magnetic traces in bit tracks (or sections of
the disk without bit tracks) when viewed using either TS
magnetic imaging or the Bitter method. This means that
the TS magnetic recording process is not directly caused
by tunneling current but rather by the proximity of the
magnetized-Fe film tip set by the STM feedback system.
Apparently the magnetic field is focused near the
tunneling tip with sufficient intensity to alter the
magnetization of the disk surface.
From the discussion above we conclude that the TS
magnetic crater in Fig. 5.1b was not due to surface
topography but rather to variation in the magnetic force
between the Fe film tip and the surface of the hard
disk. If this is true, then TS magnetic craters like
that in Fig. 8b represents a repulsive force (or a
relative decrease in an attractive force) between the Fe
film tip and the disk since it appears as a local
depression in the TS magnetic image. We speculate that
a repulsive force might result from the interaction of
the Fe tip with the divergent field of the longitudinal
magnetization of the TS magnetic record.
Last, preliminary measurements show TS magnetic
imaging and recording process was unaffected by the
presence of a 100 nm layer of gold deposited on either


30
or both of the surfaces of the hard disk and the Fe film
tunneling tip. This implies that TS magnetic images are
unrelated to electrostatic charging typical of some
semi-insulating surfaces which have been observed using
an atomic force microscope by Mamin et al.


CHAPTER VI
CONCLUSIONS
Knowledge of the surface magnetization patterns
of numerous materials is of great interest lately
stemming from the magnetic storage device revolution.
Most methods of imaging destroy the sample, are limited
in the image resolution or require elaborate preparation
and instrumentation. The Bitter method destroys the
sample with the ferrofluid and the resolution is limited
by the magnetic particle size. Lorentz force, SEMPA and
electron holography all require special detectors on
SEMs. Our MFM method alters an existing instrument only
slightly and requires very little sample preparation.
Several different metal films were tried for the
triangular tip. Fe films have a strong magnetic moment,
appropriate strength for thin films and presently give
the best images. We have tried Ni films, hoping that
the lower magnetic moment might have less influence on
existing magnetization, but have been unable to attain
the results comparable to Fe.- Also we have tried Ag and
Cu films coated with a thin Fe layer. These films also
have not reproduced the results when trying to image


32
existing magnetization patterns, although they have
recorded magnetic regions on the disk better than all
the films tried thus far. Also on some films we
attached permanent magnets to the film/wire assembly to
orient the magnetic domain at the tip of the triangular
Fe film.
Future research on tip fabrication include Fe
wires and Fe films coated with a thicker layer of Au.
As explained in the Read/Write section the proximity of
the Fe to the surface of the disk is what alters the
magnetization patterns. We need to move the Fe further
away from the disk but still be close enough to have a
tunneling current. We plan to try a thick (200 nm)
layer of Au between the Fe triangle and the disk.24
The resolution of the TSMFM images needs to be
improved. Studies of MFM images we have obtained show
that image resolution is directly related to tip
sharpness. Theoretically STM tips are atomically sharp,
so the potential resolution of our MFM imaging method
can be sub nanometer. Several methods to improve the
sharpness of the tip are being studied, such as
evaporating metal on the Fe film triangle through a
small hole in a photoresist layer to form a cone, or
evaporating a small very thin triangle using photoresist


33
on top of a soluble layer and removing the triangle.
We have demonstrated a new type of tunneling-
stabilized imaging probe for detecting magnetic fields
near the surface of a thin film hard disk. The probe
can be used to both image and record magnetic records.
The low-mass TS magnetic probe attaches directly to a
STM piezoelectric tube scanner acting both as the
tunneling tip and a compliant force detector. There is
another advantage of near-field magnetic imaging and
recording in a TS mode. The resolution limitations
caused by probe lift-off from the surface of magnetic
recording medium, inherent in other magnetic imaging and
recording methods, are virtually eliminated by the tight
tracking of a TS magnetic probe to the surface. Thus
far we have been able to create magnetic spots about 500
nm on a side, with a MFM image resolution of about 20 nm
(see Fig. 6.1).
We are also investigating the effect of field
strength on the TS magnetic imaging and recording
process using a small electromagnet to magnetize the Fe
film tip. Eventually we hope to add "erase" to the TS
magnetic menu. This may be an improvement over STM
patterning techniques based on ostensibly irreversible
chemical or physical modification of surfaces.


34


0 500 1000
Figure 6.1. A TS magnetic image of a 500 nm X 500 nm
magnetic spot on the surface of a computer hard disk. As
shown in the figure, the finest distinguishable feature
has a dimension of about 20 nm.


REFERENCES
1.
2 .
3.
4 .
5.
6.
7.
8.
9.
10.
11.
12.
H. K. Wickramasinghe, Scientific American, Oct,
98, (1989) .
P. K. Hansma and J. Tersoff, J. Appl. Phys. 6_1
(1987) .
H. J. Mamin, D. Rugar, J. E. Stern, B. D. Terris,
and S. E. Lambert, Appl. Phys. Lett. 53., 1563
(1988) .
U. Hartman and C. Heiden, J. Micro., 152. 281
(1988) .
H. C. Tong, R. Ferrier, P. Chang, J. Tzeng, and
K. L. Parker, IEEE Trans. Mag. Mag-20. 1831
(1984) .
The scanning tunneling microscope used was a
NanoScope II, Digital Instuments, Santa Barbara,
Ca.
F. Bitter, Phys. Rev. 38., 1903 (1938) .
E. Koster and T. C. Arnoldussen, Magnetic
Recording Volume I: Technology, edited by C. D.
Mee and E. D. Daniel (McGraw Hill Book Co.,New
York, 1987), pp. 194-210.
R. J. Celotta and D. T. Pierce, Science, 234. 333
(1988) .
P. Grutter, E. Meyer, H. Heinzelmann, L.
Rosenthaler, H. R. Hidber, and H. J. Guntherodt,
J. Vac. Sci. Technol. A, 2, 279 (1988).
D. W. Abraham, C. C. Williams, and H. K.
Wickramasinghe, Appl. Phys. Lett. 53., 1446
(1988) .
C. W. MacGregor, J. Symonds, J. P. Vidosic, H. V.
Hawkins, W. T. Thomson, D. D. Dodge, Marks1
Standard Handbook for Mechanical Engineers,
edited by T. Baumeister, E. A. Avollone, and T.
Baumeister III (McGraw Hill Book Co.,New York,
1978), pp. 5-38.


36
13 .
14 .
15.
16.
17 .
18 .
19 .
20.
21.
22 .
23 .
24 .
S. R. Rao, Mechanical Vibrations. Addison-Wesley,
New York, p. 412, 1990.
Y. Martin and H. K. Wickramasinghe, Appl. Phys.
Lett. 50, 1455 (1987) .
E. M. Rossi, G. McDonough, A. Tietze, T.
Arnoldussen, A. Brunsch, S. Doss, M. Henneberg,
F. Lin, A. Ting, and G. Trippel, J. Appl. Phys.
55, 2254 (1984).
D. Rugar, H. J. Mamin, P. Guethner, S. E.
Lambert, J. E. Stern, I. McFadyen, T. Yogi,
submitted to J. Appl. Phys. Jan. 1990.
H. C. Tong, R. Ferrier, P. Chang, J. Tzeng, and
K. L. Parker, IEEE Trans. Mag. Mag-20. 1831
(1984) .
Y. Martin, D. Rugar, and H.K. Wickramasinghe,
Appl. Phys. Lett. 52, 244 (1988).
J. A. Dagata, J. Schneir, H. H. Harary, C. J.
Evans, M. T. Postek, and J. Bennet, Appl. Phys.
Lett., to be published.
M. A. McCord, D. P. Kern, and H. P. Chang, J. Vac
Sci. Technol. B6, 1877 (1988).
R. M. Silver, E. E. Ehrichs and A. L. de Lozanne,
Appl. Phys. Lett. 51, 247 (1987).
T. H. P. Chang, D. P. Kern, E. Kratschmer, K. Y.
Lee, H. E. Luhn, M. A. McCord, S. A. Rishton,
and Y. Vladimirsky, IBM J. Res. Develop. 32., 462
(1988) .
E. J. van Loenen, D. Dijkkamp, A. J. Hoeven, J.
M. Lenssick, and J. Dielman, Appl. Phys. Lett.
58, 1312 (1989) .
H. J. Mamin, D. Rugar, J. E. Stern, R. E.
Fontana,Jr., and P. Kasiraj, Appl. Phys. Lett.
55, 318 (1989) .


37
APPENDIX A
DERIVATION OF PRIMARY FREQUENCY
Upon Substitution of equation (1) for the moment
of intertia and equation (4) the solution for the beam
with a point load at the apex of the triangle into the
Rayleigh equation, an expresion (that is an estimation)
for the primary natural frequency is obtained.
A(x) =
G)
2_
xhb
L f LEl (x) Jo ' d2z{x) l dx2 , 2 dx
fLpA Jo r L Eh2 Jo 12 {x) z (x) 2dx f ~12PL)2xdx Ebh3 J
fLp( 6PL) Jo { Ekh3, Eh2 f x5dx
Fh 2 r L
f xdx
3 Jo
p fLx5dx
Jo
Eh2X2\ a
px6
L
0
G) =
Eh
N pl
2
4
(5)
(6)
(7)