Magneto-Optical Kerr Effect Microscope

by
Anton Geiler, Heath Marvin, Michael Zartarian, Paul Head, Adam Brandow, and
Rui Loura
 TABLE OF CONTENTS

Magnetic Domains..........................................................................................................- 3 -
Domain Observation Techniques................................................................................- 3 -
Magneto-Optical Kerr Effect Microscopy..................................................................- 3 -
Magneto-Optical Kerr Effect..........................................................................................- 4 -
Visual Contrast Optimization..........................................................................................- 6 -
Applications....................................................................................................................- 9 -
Other Methods.............................................................................................................- 9 -
Optical Design...............................................................................................................- 10 -
Optics........................................................................................................................- 10 -
Angle of Reflection...................................................................................................- 11 -
Spot Size and Resolution..........................................................................................- 11 -
Light Source..............................................................................................................- 13 -
Computer Interface.......................................................................................................- 14 -
Computer Hardware Requirements...........................................................................- 14 -
Biasing the Sample with an Appropriate Magnetic Field.............................................- 15 -
Electromagnet...........................................................................................................- 15 -
Power Supply............................................................................................................- 16 -
Gauss Meter..............................................................................................................- 16 -
Digital Image Processing..............................................................................................- 18 -
Budget...........................................................................................................................- 20 -
Schedule........................................................................................................................- 20 -
References.....................................................................................................................- 22 -

2
Magnetic Domains
In order to gain insight into the properties of magnetic materials and devices it is
important to be able to study the magnetic domain structure of the material or device
under investigation. Magnetic domains are regions of unidirectional magnetization which
are governed by one of the fundamental laws of nature – minimization of energy in the
system. Invisible to the naked eye, these are microscopic structures within otherwise
unstructured magnetic materials (figure 1). The understanding of the structure and
dynamics of magnetic domains is becoming increasingly important in various
applications of magnetic materials, such as thin-film recording heads in the magnetic
recording industry and spin-electronic devices in information technology.

Figure 1 – Domain pattern (left) and the substructure of the domain walls
(right) on a (100)-oriented SiFe crystal.

Domain Observation Techniques

A number of different techniques to study the domain structure of magnetic materials
exist today. One of these techniques is the magneto-optical imaging technique which has
significant advantages of being relatively inexpensive, non-invasive, non-contaminating,
and able to handle a broad range of magnetic samples. Magneto-optical imaging in
reflective mode takes advantage of the magneto-optical Kerr Effect while imaging in
transmission mode takes advantage of the Faraday Effect. The underlying principle of
both effects is similar. The light reflected from the surface of a magnetic sample or
transmitted through a magnetic sample will interact with the magnetization within the
sample. Through this interaction, the polarization state of the light will change and the
difference between the incident and reflected (transmitted) beams can be used to study
the magnetization within different regions of the sample.

Magneto-Optical Kerr Effect Microscopy

In our Capstone project we intend to build a Magneto-Optical Kerr Effect Microscope
which will be used in development and characterization of magnetic materials and
devices in the Center of Microwave Magnetic Materials and Integrated Circuits at
Northeastern University. The project will initially concentrate on the study of thin

3
metallic magnetic films which exhibit a more pronounced magneto-optical effect
producing a stronger signal which is easier to detect and analyze. Depending on the
progress in this stage of the project we also intend to extend our inquiry into materials
which present more challenges in Kerr Effect Microscopy such as ceramic thin films and
bulk samples.

Optimum optical contrast conditions for different materials are obtained through three
major types of magneto-optical Kerr effect. These types are classified into polar,
longitudinal, and transverse Kerr effects depending on the vector relationships between
incident light polarization, plane of incidence, and the magnetization within the sample.
An example of directional relationships between these physical quantities is shown in
figure 2 for the longitudinal case.

Figure 2 - Vector diagram for the longitudinal magneto-optical

Kerr effect. K vector is the Kerr amplitude generated as a
result of interaction of the light beam with the magnetization
vector M.

Magneto-Optical Kerr Effect

In order to detect the changes introduced into the polarization state of the light by
reflection from a magnetized surface a pair of nearly crossed polarizers is located in the
optical path. The linear polarizer on the incident beam restricts the polarization to one
direction. Through the interaction of the light with the magnetized surface polarization
components perpendicular to that of the incident beam are generated. The reflected beam
is passed through an analyzer which is nearly crossed with the incident light polarizer. As
a result, the component of the light that maintained its original polarization state after the
interaction with the magnetized surface is attenuated while the component generated
through the magneto-optical (Kerr) interaction is allowed to pass to the detector [1].

Maximum attenuation of the incident light is achieved when the polarizers are perfectly
crossed, however Kerr component of the light is very weak and would require an
extremely sensitive detector. By allowing some of the incident light to reach the detector
the Kerr signal and the background (incident) signal are allowed to add coherently

4
contributing to a higher signal which is easily registered by conventional detectors [2].
The magneto-optical Kerr effect is conveniently described by Jones matrix approach. The
incident light beam is described by the following polarization vector:
 E 
E X (1)
 EY 
We can represent the polarizer on the incident beam by the following matrix (assuming
ideal polarizer for simplicity):
0 1
P  (2)
0 0

In this setup, the light incident upon the magnetized surface will have linear polarization
in the y-direction. The analyzer set to angle Θ from the x-direction (for Θ = 0 the two
polarizers will be perfectly crossed):
 cos 2   cos  sin 
A  (3)
 cos  sin  sin 2  

The interaction of the light with the magnetized surface can be represented as a
summation of the oblique reflection matrix and the magneto-optical effect matrix [3]:
 0   0  SPL  QV  sin  
J   SS    L 2 
(4)
 0  PP    SP  QV  sin   PP sin    PP  QV / cos  cos  
2 T

In the above expression the first term represents the oblique reflection matrix where ρSS
and ρPP are the Fresnel coefficients for correspondingly polarized light components. The
second term describes the magneto-optical effect where ρSPL and ρPPT are the reflection
coefficients responsible for the Kerr rotation of the polarization state of the light, φ is the
angle between the magnetization vector and the plane of incidence, and QV is the complex
Voigt magneto-optical constant. For the longitudinal Kerr effect equation (4) simplifies
to:
  SS 0   0  SPL  QV  
J  J 0  J1    L  (5)
 0  PP    SP  QV  0 
The magneto-optical reflection coefficient ρSPL for the longitudinal case is given by [4]:
 in cos 0 mY sin 1QV
 SP
L
 (6)
 n cos 0  cos1  cos 0  n cos1  cos1
In this case, the light is incident from air at an angle θ0 upon a magnetic medium with
index of refraction n. The magnetization is given by mY in the y-direction. The angle of
refraction within the magnetic medium θ1 is obtained by application of Snell’s Law.
Combining the above components we can summarize the input-output relationship of the
magneto-optical

Kerr effect microscope as follows:
EOUT  A   J 0  J1   EIN (7)
Based on equation 6 we can now write the power relationship for normalized electric
field amplitudes:

5
 
EOUT
2 T 
 EOUT
T
EOUT  E IN
T
 T
 
 J 0  J 1  A T  A   J 0  J 1   E IN (8)

Equation 7 contains 4 terms as follows:

 2 T T

EOUT  E IN  J 0  A T  A  J 0  E IN 
 T

E INT  J 0  A T  A  J 1  E IN 
  (9)
T
E INT  J 1  A T  A  J 0  E IN 
 T

E INT  J 1  A T  A  J 1  E IN
The first three terms in equation 9 will determine the intensity ratio between the
background component of the light and the Kerr effect component of the light. Hence
these terms are critical in optimizing the contrast when an image of the magnetic domain
structures is captured with a digital camera. First term will determine the background
light intensity while the two subsequent terms will determine the intensity of the Kerr
effect light.

Visual Contrast Optimization

To gain quantitative insight into the advantages of slightly uncrossing the incident and
the reflected light polarizers in the light path of the Kerr effect microscope consider the
following simplified form of equation 9:
  2  2    2
~ ~ ~ ~ ~ ~ ~ ~
E E
0  E
K E E
0 E E  E
0 K

0

K K (10)

~
Here, E 0
is the amplitude of the reflected light which maintained the original

polarization state after the incidence (background) and E ~ is the amplitude of the Kerr
K

component of the reflected light. These terms were organized in the order corresponding
to the terms of equation 9. Hence, the first term represents the irradiance of the
background signal, the second and the third terms (conjugate pair) represent the coherent
addition of the Kerr signal and the background signal while the third term is the
irradiance of the Kerr component of the light. To make this analysis more quantitative we
will consider a high permeability metallic Permalloy film (Ni-Fe alloy). This film
represents a major challenge in magneto-optical investigations due to very weak
polarization rotations. The Kerr rotation angle defined by the arcsine of the ratio between
Kerr and background components of the light was reported to be on the order of 0.001°
for Permalloy films [5]. Consequently, the irradiance of the Kerr component of the light
is on the order of 10-6 with respect to the irradiance of the background component of the
light. With such a ratio generating a visual contrast would be quite challenging. However,
due to the coherent nature of the addition of the Kerr and the background components of
the light much higher contrast can be generated due to the second and the third terms in
equations 9 and 10. This approach is represented graphically in figure 3:

6
 Irradiance at detector vs. analyzer setting 
0.8

0.7

0.6

0.5
2
|E0 + EK|
0.4

0.3

0.2

0.1

0
0 0.2 0.4 0.6 0.8 1

Figure 3 - Irradiance dependence on analyzer setting θ in degrees

~
The background light amplitude E 0
is proportional to sin(θ) where θ is the analyzer
setting (as expected for θ = 0 perfect extinction of the background light occurs).

~
Consequently, for fixed Kerr amplitude E K the irradiance detected after the analyzer
exhibits a parabolic dependence on the angle θ. Since the slope of the curve in figure 3
increases with increasing θ, for higher values of θ a small change in the Kerr amplitude

~ will produce a higher change in the detected irradiance. Based on this observation
EK
we can conclude that by increasing the angle θ we can increase the visual contrast
introduced into the image by the appearance of the Kerr signal. This conclusion can also
be verified by substituting the dependence of the background light intensity and the
relative Kerr amplitude into equation 10. The ratio between the background irradiance
and the Kerr component irradiance is then given by:
 2
~
E0
sin 2  sin  (11)
DR       
~ ~  ~ ~ sin   10 3
10 3
E0 E K  E0 E K
In this equation DR is the dynamic range of the image obtained by the magneto-optical
setup described above. This is a noticeable improvement from the 106 ratio in the
perfectly crossed polarizers scenario. In order to achieve visual contrast in magneto-
optical effect imaging the Kerr signal has to be sufficient to produce several counts in the
corresponding pixels of a CCD camera. In the previous example, for Permalloy film,
setting the angle θ to 6° will create a dynamic range of about 100. Hence, if the
experimental setup is configured such that the background light generates 200 counts in
the pixels of the CCD camera the Kerr signal will contribute to 200/DR= ±2 pixels
(depending on the orientation of the magnetization in the given area of the sample) which
will translate into visual contrast in the magneto-optical effect image. According to this
analysis an 8-bit CCD camera with a count range of 0-255 will be able to produce an

7
image where the magnetic domain structure of the material is visible. Considering the
fact that the noise level of a typical (un-intensified) CCD camera is on the order of 2-4
counts digital image processing will be required to eliminate the noise contribution to the
image. Upgrading to a 10-bit CCD camera will provide a dynamic range of 1024.
Configuring the experiment such that the background contributes to 900 counts will cause
the Kerr effect signal to produce as many as ±9 counts which is considerably higher than
the CCD noise level. The images can be further improved by the digital image processing
methods.

Based on the theoretical analysis of the magneto-optical Kerr effect microscope carried
out in this section we conclude that this approach can applied in the study and
characterization of magnetic materials by providing a visual image of the magnetic
domain structure of the material under investigation. Applying various biasing fields to
the magnetic materials in this setup will allow us to study the dynamic behavior of
magnetic domains. Case study of Permalloy has shown that visual contrast can be
obtained even for this material which has proven to be particularly challenging in
magneto-optical investigations due to very weak Kerr rotations.

8
Applications
Magnetic domain studies are instrumental in many electrical applications, including
magnetic storage devices, transformers and motors. Understanding magnetic domains
means higher performance and efficiency in such devices. In order to better understand
these magnetic domains the Kerr effect can be used to study their structure. As the
technology to manufacture the devices dependant on magnetic domains advances, so does
the difficulty in observing their structure.

One such application is magnetic recording media. Now that manufacturing technologies
have advanced to the point where a much thinner storage medium can be made we are
able to increase the storage density of devices that use this technology. Also by
decreasing the thickness of such a medium we are now able to decreasing the coercivity
and thus the same amount of data can be stored with a weaker magnetic field [1]. The
Kerr effect microscope can be used to observe the domain walls and find the correct
medium thickness and coercivity combination for a stable information element.

Other Methods
It is important to consider other the methods of observing magnetic domains besides the
Kerr effect microscope. One such method is the Bitter technique which maps the stray
fields of a domain pattern with the use of colloidal magnetite particles. One main
disadvantage to this method is the fact that the application of the colloid to the magnetic
sample could actually affect the domains and give a false image of their structure. Also it
is important to note that the bitter effect measures the stray fields of magnetic domains.
So when it is necessary to observe the domain structure of a higher permeable material,
or any material where the stray fields are weak, the Bitter method would be less than
ideal [1].

One of the main advantages of observing domains through the magneto-optical Kerr
effect is that it can observe a variety of sample shapes, sizes, and compositions all while
leaving it uninfluenced during observation. Another method for domain observation,
Transmission Electron Microscopy [TEM], is limited in it’s ability to observe different
shapes and sizes of samples. It is also difficult to apply outside influences, such as
magnetic fields and physical stresses, to the sample while observing it [1]. The TEM also
has a smaller field of view than the Kerr Microscope, which is less expensive than the
TEM, and uses materials and items that can be adapted for other purposes.

The Kerr microscope is the one of the most versatile domain observations techniques
considering the cost. It can observe a wide range of magnetic samples that can be used
for a variety of magnetic devices. While there are other observation techniques available,
one that employs the Kerr effect is an effective solution in today’s modern magnetic
material laboratories.

9
Optical Design
In the design of the optical system of the Kerr Effect Microscope we rely on the
components integrated into the Mitutoyo FS110 Inspection Microscope. The microscope
has several limitations which will have to be addressed in order to adapt it in the Kerr
Effect Microscope design. One of the main limitations is the limited infinity space
available to us in the microscope.

Figure 4 - Mitutoyo FS110 Inspection Microscope [6]

While the microscope has infinity corrected optics it is not designed to allow the users to
incorporate additional optical components into the frame of the microscope. In order to
achieve a high signal to noise ratio and eliminate as much background light as possible
we will have to use high quality optical components. For example, a pair of polarizers
(Glan-Thompson prisms) with an extinction ratio of 10-6 will occupy a volume of 1/8
cubic inches each. Sheet polarizers currently used in the microscope have an extinction
ratio of 10-4 at best. Additionally, the original optical path in the microscope utilizes a
non-polarizing beam splitter with a transmission ratio of about 50%. Such losses cannot
be tolerated in such a sensitive design. This beam splitter will have to be replaced with a
polarizing beam splitter with transmission ratio of as much as 95% which will occupy a
volume similar to that of Glan-Thompson prisms (~1/8 cubic inches). As a result it was
decided to rebuild the component of the microscope that houses the infinity space to
create more room for the necessary optical components. This will bring the illumination
straight into infinity space avoiding any effects introduced by transmission through
various components currently installed in the microscope.

Optics
Our light source—discussed shortly—will follow the path in the image below. It will
shine though a polarizer which will polarize all the light directed toward our magnetic
sample. This light will be reflected by the sample with changes in polarization due to the
magneto-optical Kerr effect, and this light with changes in polarization, amplitude, and
phase (these changes are dependent on the orientation of the magnetization) will be
reflected through another polarizer (or analyzer).

10
 CCD Camera
Polarizer
Tube Lense
Mirrors
Polarizer/Analyzer

Polarizing Beam Splitter

Laser
Objective Lense

Figure 5: The Light Path of the Proposed Microscope

The second polarizer, which will be nearly perpendicularly crossed with the first, will
function as a filter for the light which reflects off the sample without changes in
polarization. This is discussed in further detail in the previous sections. As also discussed
in the previous sections on magnetism, the rotation angle of the polarization is very low
(0.01 degrees or lower depending on the sample under observation), and the signal of the
Kerr-affected light is weak. This weak signal will pass though the second polarizer—also
called the analyzer—and through the tube lense into the camera and to the image
processing in the computer interface for enhancement.

Angle of Reflection
The angle of reflection of the polarized light will be an element of our study. The light
path diagram above shows that vertical adjustments to the second mirror will facilitate
changes in the angle of reflection. Our study will involve research of the range of
possible angles using this method. We will also discuss the use of a fiber-optic cable to
direct the laser light source. This not only eliminates the need for mirror adjustments, but
it also reduces the coherence of the laser light. Coherence is an effect caused by light
having defined phase relationships. This can cause the light to add up destructively and
produce spots on the viewed area [7].

Changing the angle of incidence will allow us to find the angle that has the largest effect
on the polarization angle and thereby gives the best contrast in our domain imaging. This
angle flexibility is only possible of the beam size is sufficiently small. The problems that
may arise with a small beam size are discussed in the following section.

Spot Size and Resolution

Some of the critical parameters that need to be considered in the design of the optical
system are the field of view and the resolution. Field of view defines the area of the

11
sample that can be viewed simultaneously by the microscope. This area is given by the
following formula:

4 
b F (1)
 B

where b is the diameter of the visible spot, λ is the wavelength, B is the diameter of the
beam of light incident upon the objective lens, and F is the effective focal length. For the
red HeNe laser the wavelength is 633nm. Such as laser can have the diameter of the beam
on the order of 0.5mm. If, for example, we use the x50 Mitutoyo M Plan Apo objective
with focal length of 4.0mm, equation (1) yields a spot diameter of ~6.5μm. This spot size
is smaller then what would typically be needed. To achieve larger spot size (on the order
of 20-40μm diameter) the laser beam will have to be condensed to a smaller diameter.

Resolution is a critical parameter of our microscope since the domain structures for
certain materials of interest can be on the order of microns or even fractions of a micron.
Due to the diffraction limit, the resolution is ultimately limited by the numerical aperture
of the objective lens. Consequently, for the aforementioned x50 objective lens (NA =
0.55) the maximum achievable resolution is given by the following equation:

C 
  (2)
NA

to be 0.7μm (C assumed to be 0.61, wavelength of 633nm and 0.55 NA). However, in

order to be able to conduct longitudinal and transverse Kerr Effect domain observations,
a well defined oblique angle of incidence is required. In order to realize such an angle of
incidence a very narrow beam of light is required. In order to take into account the
dimensions of the light beam equation (2) needs to be modified by the introduction of the
“Filling Factor” defined as the ratio of light beam diameter to the objective lens diameter
in the denominator. The maximum resolution as determined by equation (1) needs to be
multiplied by the inverse of the “Filling Factor” to yield the maximum resolution
achievable in the scenario described above. Consequently, to achieve the resolution of
1μm the “Filling Factor” needs to be 0.7. This result is concerning because if the beam of
light will occupy 0.7 of the objective it will be impossible to have a well defined angle of
incidence. The situation can be improved by utilizing a shorter wavelength; however
additional design tradeoffs will have to be made in order to achieve a satisfactory
resolution.

Because of the initial concerns about the trade-off between resolution and spot size, we
have and will explore means of resolving this issue. We plan to research the laser light
more extensively, and we may find that there is a balance adequate to define an angle of
incidence while still getting a resolution high enough. If this proves to be impossible, we
will explore other options. These options include mechanically dithering a multimodal
fiber [3], scanning the sample at high speeds, or using a shorter wavelength. These
options will be delved into as it become necessary during the design process.

12
Light Source
Because the extinction is required to be so high, it appears that the best approach is to
begin our experiments with a laser light source. We would like our laser to be green,
linearly polarized, and highly stabilized. This would be the most ideal choice, but
currently we have a red laser which is polarized but not stabilized (varies in intensity up
to 5%). These specifications are required for the following reasons:
 It should be green because the wavelength of green is 543 nm and that would enable
us to have a viewing resolution of half that distance (Red light has a wavelength of
about 633 nm).
 It must be linearly polarized to prevent intensity fluctuations that occur when a non-
linearly polarized laser changes polarization.
 It must be highly stabilized to minimize power fluctuation during operation.
The cost of a laser with these qualifications may be priced in the range of $4,000, so it is
important that we attempt to use the available red laser.

13
Computer Interface
The PC in our setup will communicate with, regulate, and capture data from the various
components. Specifically, the PC will perform the following tasks:

 Regulate and monitor power supply voltage, current and polarity

 Monitor magnetic field at the sample with a Gauss meter
 Monitor ambient light
 Acquire image from CCD camera
 Perform image processing on captured data and display results

These tasks will be executed within a LabView program that will be written for the
application. The test equipment will interface with the computer via standard industry
communication protocols such as GPIB, RS-232, or USB.

Computer Hardware Requirements

A computer may be readily available to us in the lab to perform the aforementioned tasks,
but will require some degree of modification. Alternatively, we could construct a
dedicated custom PC for the application for relatively low cost.

Approximate Hardware Requirements:

 Pentium 4, 2.5 GHz, 533 MHz FSB or equivalent

 1GB or greater of memory
 60GB Hard Drive or greater
 Dual monitor capable video card
 USB 2.0 ports
 RS-232 Ports
 GPIB Interface PCI Card

Required Software:

 Windows XP or equivalent
 LabView
 MatLAB

14
Biasing the Sample with an Appropriate Magnetic Field
The sample to be observed by the Magneto-Optical Kerr Effect must be placed in an
appropriate magnetic field to excite the magnetic domains. For our setup, this field must
also be measured and regulated via computer control. The field strength required varies
greatly depending on the sample under test. The components required to create and
regulate this field are show in Figure 4:

Figure 6: A diagram of the major subsystem components

The basic requirements of this subsystem are as follows:

 Produce a precise field from tens of Gauss to hundreds of Gauss

 Have the ability to quickly and easily change field polarity
 Ability to precisely measure and adjust the field strength
 Ability to be controlled closed-loop via computer and LabView software

Electromagnet
A suitable C-shaped electromagnet may be constructed out of readily available materials
and techniques. A quick calculation yields an approximate result:

The strength of an electromagnet in air is governed by Ampere’s Law:

B  * gap length  = μ * N * I

15
 B is in Tesla (10,000 gauss)
gap is in meters (This is the opening of the "C ".)
μ = 4π * 10-7
N is an integer equal to the number of turns in coil
I is the current in Amps

Assuming a maximum 200 Gauss desired field, and 2 cm air gap, approximately 320
ampere-turns would be required. In other words, this would require 320 turns of wire
carrying 1 ampere, etc. These calculations should yield actual results within ±20%.

Observing that this result is easily within the grasp of our construction capabilities, a
more formal analysis will be made before fabrication. Factors which will need to be
considered include:

 Core metal type

 Exact core size and shape
 Number of turns in coil
 Optimizing coil for heat, size and total resistance
 Physical mount into the MOKE setup
 Mounting a Gauss meter within the gap to monitor field strength

The electromagnet will be mounted on the microscope sample stage, held by a hard
plastic housing. Further considerations will need to be given if there are any optical
components which need to be accommodated.
Further research and trials will be conducted – we anticipate building a prototype
electromagnet and then a final electromagnet further along in the project.

Power Supply
To power the above electromagnet and work in this subsystem, our power supply must
have the following specifications:

 50 Volts, 5 Amperes (DC)

 Bi-polar
 Fully programmable to interface with LabView
 High resolution (12 bit or greater) current control

Attempts will be made to find a suitable power supply in the laboratory.

Gauss Meter
A Gauss meter for the subsystem will need to meet the following specifications:

 12 bit resolution with a wide dynamic range

16
  Able to easily interface with GPIB/IEEE-488 and LabView

A suitable Gauss meter should be available in the laboratory.

17
Digital Image Processing
In observing the magnetic domains of material, one of the most challenging issues
will be to enhance the captured image through digital image processing. This will be
done in two ways; first a baseline image of the material will be taken while the sample is
magnetically saturated. This baseline image will be subtracted from all following images
in order to distinguish contrasts observed from background light. Secondly we will use
multiple images averaged to obtain a clearer, noise free image. If calibrated correctly the
system will be able to distinguish the magnetization direction and a color code will be
used to illustrate this in the images displayed on screen.

The challenges we expect to face in this section of the project involve vibrations,
processing time, and system calibration. If the vibrations occur during an image capture
the clarity of the image will be degraded, this may also introduce unwanted noise [1].
For this reason, the image processing will need to be relatively quick to ensure stability of
the system. System calibration will be done by optimizing the system to utilize the full
dynamic range of the camera, this will be adjusted after the system is set up and we begin
experimenting with image capture. Concurrently with this effort we will attempt to add a
color code to distinguish the direction of magnetization.

As discussed earlier, the dynamic range of our system is dependent on the number
of bits our camera uses in the digital output. Using the case study of Permalloy we
calculate that we would probably register about 2-3 counts with an 8-bit camera, to be
sure that we are out of the noise floor we will look to obtain a 10-bit camera. Our aim is
to obtain resolutions on the order of 0.5um; to do this with an objective lens
magnification of 200x, we will need a camera with at least 1 Mega pixel resolution.
Cameras like this are available on the commercial market at a cost of about $2000.

In order to see good contrasts between the different magnetic domains we will
need to process our images by subtracting a baseline image taken while the sample is
magnetically saturated [1]. In addition to this image subtraction we will need to take
several pictures and average the images; this will reduce the noise in our measurement as
well as helping to develop a color code to identify the direction of magnetization. This
processing can be done easily in matlab by using the imread command and performing
the different mathematical functions on the resulting matrix. This is not the expertise of
any of our group members, however after an initial research we are confident this can be
done and exported to our labview control program.

18
Figure 7 - Design Block Diagram

19
Budget
Order Materials
Laser $1500 BORROW  
Camera w/ Software $1800 PURCHASE 10 bit ~1MPixel
BORROW
Computer w/ GPIB $200  
Or BUILD

$150 BUILD  
Horeshoe Mag (Soft Fe)
bi-
Power Supply DC $2000 BORROW
polar/switching?
PURCHASE/
Gauss Meter w/GPIB $800  
REPAIR?
Current meter   BORROW  
Glan-Thomsons   BORROW  
Polarized Beam Splitter $300 PURCHASE  
Lenses/Apperatures $500 PURCHASE  
Stages/Mounting
$ PURCHASE ferrite material
Apparatus
Raw Metal $200 PURCHASE  

Schedule
Purchase Materials 8/31
Camera Paul
Laser Anton
Magnet Mike
Polarizing Beam
Anton/Heath
Splitter
ALL Lenses/Apperature Heath
Raw Metal Mike
Computer/BPIB Rui
Power
Adam
Supply/Meter
Gauss Meter Anton
Modify Microscope 9/21
Lead: Anton Machine Microscope + Breadboard

20
Assist: Adam + Mike
Stage Modification
Setup Optics 10/1
Lead: Anton + Heath
 
Assist: Paul
Magnet/Stage 10/1
Lead: Mike
 
Assist: Rui
Write/Use DIP 10/1
Lead: Paul
 
Assist: Adam + Rui
Write GUI 10/1
Lead: Rui
 
Assist: Paul
Start Taking Images 10/31
ALL  
Enhance Image 11/15
ALL  
Final Enhancement 11/31
ALL  

21
References

[1] A. Hubert, R. Schäfer, Magnetic domains: the analysis of magnetic microstructures,

New York, NY: Springer-Verlag New York, 1998.
[2] B. E. Argyle,J. G. McCord, “New laser illumination method for Kerr microscopy,” J.
Appl. Phys., vol. 87-9, pp. 6487-6489, May 2000.
[3] S. A. Oliver, C. A. DiMarzio, S. C. Lindberg, S. W. McKnight, A. B. Kale,
“Measurement of magnetic fields using the magneto-optic Kerr effect,” Appl. Phys.
Lett., vol. 63-3, pp. 415-417, July 1993.
[4] C. Y. You, S. C. Shin, “Generalized analytic formulae for magneto-optical Kerr
effects,” J. Appl. Phys., vol. 84-1, pp. 541-546, July 1998.
[5] B. E. Argyle, B. Petek, D. A. Herman, “Optical imaging of magnetic domains in
motion,” J. Appl. Phys., vol. 61-8, pp. 4303-4306, Apr. 1987.
[6] Mitutoyo FS110 Specification Sheet. (n.d). Available:
http://www.mitutoyo.com/catalog/pdf/F-09-10.pdf.
[7] F. Joachim, Three-dimensional Electron Microscopy of Macromolecular Assemblies.
New York: Academic Press, 1996, pp. 28-31.

22