5me% t.

“nu: i .

{war

mm .31..
“‘2 swig

.w

a... .gi‘wflawwfi
‘ iémmfimfiflhu,
rm». Mm.

. .a.....|..hv‘

3...! . 2. 1.
453.5%? y
r. .1;
a.

a.

 

w flag, A .
. Y I may}.

 

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

A Real Time Bacterial Cell Detection System

presented by

Trevor Thomas Pled Barkham

has been accepted towards fulfillment
of the requirements for the

 

 

 

PhD degree in Biochemistry and Molecular
Biology
and Chemistry
l
t "v
0 [4' V\/!\\_\ oJL .
\w ' MajoV Professbr’s Signature
6" 2 H “‘ o ,9

 

MSU is an Affirmative Action/Equal Opportunity Employer

 

_-.-._.-.-.-.-._ _

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc&Pres/ClRC/Dateoue.indd

A Real Time Bacterial Cell Detection System

By

Trevor Thomas Pled Barkham

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Biochemistry and Molecular Biology
and Chemistry

2009

 

Abstract
A Real Time Bacterial Cell Detection System
By
Trevor Thomas Pled Barkham

The detection and prevention of microbial pathogens has always been a
challenge. The development of new technologies over the years has brought with
it improvements in our ability to detect and prevent infection by microbial
pathogens, resulting in dramatic advances in the quality of public health. The first
step in the prevention of infection by and spread of microbial pathogens is
accurate detection. Our ability to quickly and precisely detect a specific microbial
pathogen greatly enhances our ability to take the proper actions. There are many
short comings in the current detection platform technologies. The many features
required of a detection platform are speed, accuracy, sensitivity, portability,
robustness, and ease-of-use. The overarching goal of this project is to design
and fabricate a functional pathogen detection platform based on basic physical
laws that uses simple electronic, chemical and biological technologies, and is
capable of the real time detection of bacteria. Real time detection of a single
bacterium is not an easy task. There is very little in the way of measurable or
inducible signals that would allow detection let alone be specific enough for
identification. The obvious solution then is to add a signal to the bacterium.
In our strategy, ferromagnetic nanooomposite particles (FMNCPs) that contain
surface recognition elements that can detect and bind to specific bacterial cell

surface structures are used even if known to contain many complex surface

antigens. The FMNCPs utilize the bacterium’s surface as a scaffold to coat the
bacterium and effectively make a micro-magnet which can be detected by
several types of magnetic sensors.

We will show that we have been able to design, fabricate, and test our
detection platform and have shown that our detection platform is capable of real
time detection of low concentrations of bacteria within five minutes.
Demonstrating ferromagnetic nanooomposite particles can be used for the

detection of bacterial cell by magnetic induction with a search coil.

Cepyright By
Trevor Thomas Pled Barkham
2009

ACKNOWLEDGMENTS

My time at Michigan State has been a learning, and eye opening
experience. Not just in the sciences, but life and the workings of the world. This
has not been an easy path for me, and I would not be here if it wasn’t for the
support and help of many people. I would like to thank everyone how has helped
me on my way.

Firstly, I would like the thank Dr. Rawle l. Hollingsworth for taking me into
his lab and providing numerous opportunities and experiences. Your patience
and guidance over the past years have been greatly appreciated. Thank you for
your confidence. My appreciation is also expressed to my Graduate Committee:
Drs. Charles Hoogstraten, Kathy Hunt, Ned Jackson, Bill Wedemeyer, and Greg
Zeikus for their valuable insight and perspectives contributing to my project and
professional development.

I would like to thank the members of the Hollingsworth Lab. As colleagues
and friends you have all enriched studies and view of the world. Thank you Kun
for always playing devils advocate even when you agree with me. Dr. Kuhn thank
you for your help, insight, and perspective over the years. Dr. Kuhn, Dorothy
Black and Matt Horn Thank you for reviewing and helping me with my paper. I
would like to acknowledge the staff in the center for advanced microscopy. The
entire staff has been most helpful, always willing to listen to and work with me. I
would also like acknowledge the entire staff of the Biochemistry department from
the repair shop to the receptionist’s offices; you have aided me in resolving many

problems and complaints.

I am most thankful for my Parents, Tom and Ruth. Thank you for all that
you have done for me over the years and all that you have not done. You have
helped become the man I am today. Thank you to all my family from grandma to
brother you have all helped to shape me.

I would like to thank my wife Emily and my sons Edward and Benjamin, for
their love and encouragement. Emily thank you for putting up with so much, I

love you and appreciate all that you have done for me and with me.

vi

TABLE OF CONTENTS

LIST OF TABLES ................................................................................. ix
LIST OF FIGURES ................................................................................ x

LIST OF ABBREVIATIONS ................................................................... xiii
CHAPTER 1

Background: Detection of Bacteria ............................................................. 1

Introduction .......................................................................................... 1

Detection Platforms ................................................................................ 9
Challenges Shared by All Detection Platforms ............................................ 22
The Potential for Magnetic Detection ........................................................ 24
References ......................................................................................... 28
CHAPTER 2

The Big Idea: Magnetic Modification of Bacteria .......................................... 32
Introduction ......................................................................................... 32
Design of FMNCPs ............................................................................... 37
Macromolecular Polymer System ............................................................. 37
Detection Platform Design ...................................................................... 41

Theoretical Calculations and Expected Values ........................................... 42
Noise ................................................................................................. 49
References ......................................................................................... 52
CHAPTER 3

Macromolecular Polymer System ............................................................. 53
Introduction ......................................................................................... 53
Experimental ....................................................................................... 56
Results and Discussion .......................................................................... 57
References ......................................................................................... 64
CHAPTER 4

Nanocomposite Material ........................................................................ 65
Introduction ........................................................................................... 65
Ferromagnetic Nanoparticles .................................................................. 66
Interaction of Ferromagnetic Nanoparticles ................................................ 69
FMNCP and Bacterium Binding ............................................................... 74
Experimental ....................................................................................... 79
Results and Discussion .......................................................................... 81
References ........................................................................................ 103

vii

CHAPTER 5

The Detection System ......................................................................... 106
Introduction ....................................................................................... 1 06
Experimental ..................................................................................... 109
Results and Discussion ........................................................................ 111
Conclusion ........................................................................................ 130
References ....................................................................................... 135
Appendix A

Magnetism ........................................................................................ 136
Introduction ....................................................................................... 1 36
Magnetic Energies .............................................................................. 141
References ........................................................................................ 148
Appendix B

Calculations ....................................................................................... 149
References ....................................................................................... 153

viii

LIST OF TABLES
Tables
1.1 Ranking of commercially available detection platforms .............................. 4
5.1 Voltages and magnetic fields generated by the FMNCP coated bacterium.124
5.2 Control samples conductivity and average voltage with standard deviation

.. ........................................................................................................................ 133

ix

LIST OF FIGURES

Figure 1.1 An example of a metabolic detection platform ............................. 10
Figure 1.2 Electronic Noise ..................................................................... 11
Figure 1.3 Steps involved in DNA preparation for PCR ................................. 14
Figure 1.4 DNA hybridization platform ...................................................... 16
Figure 1.5 ELISA based method .............................................................. 19
Figure 2.1 FMNCP architecture ............................................................... 34
Figure 2.2 The bacterium is mixed with the FMNCP suspension ..................... 36
Figure 2.3 The lay out of the general design of the detection platform ............. 38
Figure 2.4 Structures of macromolecular polymer system ............................. 40
Figure 2.5 FMNP passing through loop ..................................................... 44
Figure 2.6 Graph of equation 4 ............................................................... 45
Figure 2.7 Graph of equation 5 ............................................................... 47
Figure 2.8 Graph of equation 7 ............................................................... 48
Figure 3.1 Reaction mechanism of macromolecular polymer system ............... 55
Figure 3.2 HNMR time 0 minutes ............................................................. 58
Figure 3.3 HNMR time after 15 minutes .................................................... 59
Figure 3.4 HNMR time after 30 minutes .................................................... 60
Figure 3.5 HNMR final polymer ............................................................... 61
Figure 3.6 IR spectroscopy of macromolecular polymer system ..................... 62
Figure 3.7 Electrospray mass spectrum of macromolecular polymer system ..... 63
Figure 4.1 The Langevin equation ............................................................ 71

Figure 4.2 The energy of a cobalt nanoparticle in the Earth’s field .................. 72

Figure 4.3 Structure of capsular polysaccharide .......................................... 77
Figure 4.4 Structures of Rhizobium polysaccharides .................................... 78
Figure 4.5 Gold nanoparticle suspensions and controls ................................ 82

Figure 4.6 Fluorescent confocal microcopy images of a dried piece of the gold

nanoparticle suspension ........................................................................ 84
Figure 4.7 Cobalt nanoparticle suspensions ............................................... 85
Figure 4.8 Cobalt nanoparticle suspension with bar magnet .......................... 86
Figure 4.9 TEM image of dried cobalt nanoparticle suspension ...................... 88
Figure 4.10 The confocal image of tubules and particles .............................. 90
Figure 4.11 Antibody evalution ................................................................ 93
Figure 4.12 SEM image of heat fixed Rhizobium trifolii cells .......................... 95
Figure 4.13 SEM images of nanooomposite and Rhizobium tn'folii mix 1 .......... 97
Figure 4.14 SEM image of nanooomposite and Rhizobium tn'folii mix 2 ............ 98
Figure 4.15 TEM image of Rhizobium trifolii control ..................................... 99

Figure 4.16 TEM image of the nanooomposite and Rhizobium trifolii mix 1 ..... 100

Figure 4.17 TEM image of the nanooomposite and Rhizobium tn'folii mix 2.....101

Figure 5.1 Graph of voltage induced in coil at different velocities .................. 112
Figure 5.2 Graph of average corrected voltage vs. number of beads ............. 114
Figure 5.3 Graph of average corrected voltage vs CF Us/mL ........................ 116
Figure 5.4 Raw data of voltage induced by sample .................................... 117

Figure 5.5 Graph of all of the nanooomposite and Rhizobium trifolii data samples
with line smoothing applied to the data .................................................... 118

Figure 5.6 Induced Voltages by the nanooomposite and Rhizobium trifolii ...... 120

xi

Figure 5.7 Graph of the average angle of alignment of the nanoparticles in the
earth’s field as a function of the number of 3 nm particles ........................... 122

Figure 5.8. Average voltages of the different nanooomposite suspensions......127

Figure 5.9 Graph of the average corrected voltage vs. the number of cells for
Rhizobium Tn'folii cell ........................................................................... 128

Figure 5.10 Graph of all the corrected voltage data .................................... 129

Figure A1 The toroid shaped magnetic fields of magnetic dipoles of different
sizes ................................................................................................ 138

Figure A2 The cooperative interaction of the point magnetic dipoles in

ferromagnetic materials results in a net magnetic moment ........................... 139
Figure A3 Domain structure of a polycrystalline ferromagnetic material .......... 143
Figure 4 Domain walls ......................................................................... 144
Figure A5 Hysteresis loop ..................................................................... 145

Images in this dissertation are presented in color

xii

AMR
CF U
CPS
CWG
EDC
ELISA
EMF
Fab

F ITC
FMNCP
FMNP

GMR

ICMS
IMS
KLH
LPS

M

Ms

MALDl-TOF

MAR
MR

LIST OF ABBREVIATIONS
anisotropic magnetoresistance
colony forming units
capsular polysaccharides
cell wall glycopolymers
1-ethyI-3-[3-dimethylaminopropyl] carbodiimide
enzyme linked immunosorbent assay
electromagnetic force
antibody fragment
fluorescein isothiocyanate
ferromagnetic nanooomposite particles
ferromagnetic nanoparticle
giant magnetoresistance
integrated circuits
intact cell mass spectrometry
immunomagnetic separation
keyhole limpet hemocyanin
lipopolysaccharides

magnetization

saturated magnetization
matrix-assisted laser desorption/ionization time-of-flight
magnetic assay reader

magnetoresistance

xiii

MTJ
MWCO
NC
PBS
PCR
RT—PCR
scFV
SELEX
SEM
SPR
SQUID
SRM
TEM
THF
VOCs

magnetic tunnel junction

molecular weight cut off

nanooomposite

phosphate buffered saline

polymerase chain reaction

reverse transcriptase polymerase chain reaction
single chain fragment variable

systematic evolution of ligands by exponential enrichment
scanning electron microscope

surface plasmon resonance

superconducting quantum interference devices
surface recognition molecules

tunneling electron microscope

tetrahydrofuran

volatile organic compounds

xiv

Chapter One
Background: Detection of Bacteria
Introduction

The detection and protection against of microbial pathogens has always
been a challenge. The development of new technologies over the years has
brought with it improvements in our ability to detect and prevent infection by
microbial pathogens, resulting in dramatic advances in the quality of public
health. The first step in the prevention of infection by and spread of microbial
pathogens is accurate detection. Our ability to quickly and precisely detect a
specific microbial pathogen greatly enhances our ability to take the proper
actions, whether it is isolating the cause of food poisoning or the detection of a
biological weapons attack.

The development of detection platforms for bioterrorism and biowarfare
agents has increased since the 2001 US. Postal Service-based anthrax attacks,
in which out of a total 22 cases of cutaneous and respiratory anthrax, 5 were fatal
[6]. With the average time between exposure and onset of symptoms being 4.5
days and an average of 10 days to get diagnosis from lab tests [6], fast
identification of Bacillus anthracis spores and treatment with the appropriate
antibiotics meant the difference between life and death for postal workers.

While biological attacks are a real threat, one is far more likely to die of
food poisoning. It is estimated that in the United State there are approximately 76
million food-borne illnesses, 325,000 hospitalizations, and 5,000 deaths every

year [4]. In fact, over the past 20 years, 38% of all publications on detection of

pathogenic bacteria have been concerned with the food industry. Of the
remaining publications 18% were clinical, 16% were in water and environment,
26% in other areas, and only about 1% for defense [1]. In 2007 there were over
55 recalls issued by the USDA Food Safety and Inspection Service [3] regarding
quality control issues in food.

The fast and accurate detection of pathogens has applications in nearly
every aspect of our society, from personal home use to clinical and professional
use. Early wam‘ing and monitoring systems, biological/chemical response units
along with medical diagnostics are the most pressing applications in today's
current environment. Areas of work such as emergency response units, front-line
healthcare providers, law enforcement, and military personal, that involve
possible exposure to highly pathogenic bacteria demand fast and accurate
detection. While speed and accuracy are the most important, they are not the
only two features of current detection platforms that are in need of improvement.
The other features are sensitivity, portability, robustness, and ease-of-use.
Deficiencies in these features have limited the applications of detection
platforms. One can envision an ideal detection platform embodying all the above
features that would open the door to numerous possible applications.

There is no agreement on a standard for an ideal or preferred detection
platform but one that yields results in 5 minutes or less, with an accuracy of 99%
or better for all types of error should be close for all types of error. It would have
the sensitivity to detect a single cell; it would be a small lightweight personal

monitoring device that could be integrated into other personal electronic devices

such as a cell phone. The detection platform would also have the robustness to
maintain function in normal day-to-day use in the field as well as in a broad range
of environmental conditions especially temperature. Last but not least it would
require no user initiative beyond turning it on. The detection platform should
detect and alert the user to any new threat or condition autonomously.

While the desirable characteristics of the features of a detection platform
may be stated generally, they must be assigned some quantification in order to
make some comparisons among platforms and determine their strengths and
deficiencies. Table 1.1 gives a list of currently available detection platforms and
feature ratings. The relative importance of the features is ranked with speed as
most valuable, followed respectively by accuracy, sensitivity, portability,
robustness, and ease-of-use. The features are then rated based on a three star
scale with three stars being the best (ideal) to zero (doesn’t meet any of the
criteria). The criteria of the above features and rating are as follows:

Speed is the time required to take and analyze a sample. Zero stars for
analysis times greater than 48 hours, one star for greater than 24 hours, two
stars for 24 hours or less and three stars for 5 minutes or less. Diseases such as
anthrax can kill within 48 hr, with flu like first stage symptoms, which make
diagnosis difficult, and rapidly progress to stage two lasting about 24 hours and
ending in death [42]. Immediate diagnosis and treatment in the first stage are key
to survival, treatment after the onset of second stage symptoms has little to no

effect on the mortality of the victims.

 

 

 

 

 

 

 

 

 

 

 

i .1 .1 i i. em: Co emam

a a i i a a. i 32333.

«a is is *1 i $328
33228

i. .1 32 i i .1. i i i. >oflaoo<

i i i i i i t i i « 83m
m:_no.aocaEE_ e:_no_aoc:EE_ 330325.55 mom 3282.005 0n>h
.351 233...... «0.54.8 xw mON<m 9.5.5 Eats-n.

 

 

 

 

.mEeorma 5:098 m3m=m>m >=m§oEEoo v6 9.23m 3 03m...

 

Accuracy will be the percentage of total pathogenic samples correctly
identified. One star for 70% or better, two stars for 80% or better and three stars
for 99% and better.

The sensitivity is the limit of detection for the platform. Three stars for
detection of 1 cell, two stars for detection of 10 or fewer cells and one star for
detection of 100 cells or fewer. Low limits of detection are needed because an
infectious dose of pathogenic bacteria such as Yersinia pestis can be less then
10 organisms [44]. The “reality” is that it really only takes one viable organism to
take root to cause an infection. Another reason for low detection limits is that
samples with low concentrations of the pathogen are the norm.

For the robustness, three stars for platforms that can function in the
normal day-to-day use in the field — the same that one would expect of a cell
phone. They should be shock and impact resistant as well as functioning in a
broad range of environmental conditions like temperature and humidity. Two
stars if it is shock and impact resistant but does not function in varying
environmental conditions and vice-versa. One star if it can function in the field.

For the portability criteria, three stars for a personal monitoring device that
can be integrated into other personal electronic devices such as a cell phone.
Two stars it can be moved around in the field and one star if it cannot be moved.

For ease-of-use, three stars indicate that the platform requires no user
initiative, no sampling, no sample preparation yet the platform still detects and

alerts. Two stars, the platform requires user initiative and sampling but no

preparation. One star, the platform requires user initiative and sampling with
minimal sample preparation (such as pipetting and mixing).

Now that we have outlined an ideal detector let us take it one step further.
Here we will show you how and why such a device is needed and the impact it
would have on the world. Imagine now that everyone has this technology
integrated into their cell phones, and is constantly monitoring their environment.
While this platform can function on its own, it will actually be connected to a
larger central monitoring system. At the moment of detection a call would be sent
out from the detector to the central monitoring system, as well as notifying the
user. The central monitoring system can then be in contact (operator, or text
message) with the exposed individual, giving instructions and collecting
information. Even without contact with the exposed individual a wealth of
information can be obtained, such as location (past and present) by GPS, as well
as the individual that has been exposed via phone registers. Once an individual
is identified a database can be mined such as DMV recorders. At the same time
that information is being collected, the appropriate response teams can be
notified and begin mobilizing. The central monitoring system will then coordinate
the response by relaying information as to the location and number of possible
exposed individuals. This type of system would drastically improve our ability to
stop the spread of disease and improve the chances of survival of the exposed.
Such a system would allow the global monitoring of the origin and progression of

diseases.

Such a system brings up privacy and tampering concerns. The ethical and
moral implications of technology must not be forgotten. While we may suggest
the possibility that everyone would carry such a device, the choice should be left
up to the individual. The use of such detection platforms may not find a place in
the hands of the general public but only those who have a more realistic chance
of exposure, such as first responders and military personnel. The device could
also be employed in much the same way that surveillance cameras are on street
corners, hospital, and government buildings. While technology has provided
many improvements in our lives we must not allow it to be used to infringe upon
our liberty and freedom. With all new technologies and scientific advancements
there will always be debate over how they will be used. As scientists we strive to
better understand the world around us with hopes of improving the lives of our
fellow mankind. The knowledge that science gives us allows us the potential to
do great things both good and bad, we must all choose what we do with our
knowledge. We must also look to the future and think about what implications
that our actions will have in the future.

New advances in technology in all fields of science have yielded a
plethora of detection platforms of every possible combination and the economics
of mass manufacturing has brought them within an almost affordable price range.
Traditional methods of detection that rely (in cultivation and biochemical assays
can take anywhere from 24 hours to 1 month [7, 8]. While these methods are
reliable, highly selective and sensitive, they are not capable of yielding rapid or

real-time detection. Rapid and real-time detection is a relative term in defining a

detection platform’s performance, often only the length of time for an assay is
given excluding the time for sample processing, purification, and enrichment.
These steps not only increase the time it takes to get results, but in many cases
require a laboratory environment to operate which is neither economical nor
practical for field use and does not allow for personal monitoring.

Detection platforms are classified by one of two means, the first is the type
of identifying markers that are characteristic of a particular pathogen such as its
DNA or select surface antigens. The second classification is based on the
transducer that measures the changes in the identifier elements, such as optical,
electrochemical, and mass characteristics. The second classification is not ideal
and leads confusion due to the numerous types of transducers. For simplicity I
will use classification based on identifying elements, which are divided into four
major groups: metabolic, genetic, surface recognition, and other. Classifying by
identifying elements gives a better understanding of what is available for
detection and how best to detect it. As mentioned before, there are many
features that are demanded of these detection platforms. They must have speed,
accuracy, sensitivity, portability, robustness, and ease to use.

The ability to determine the viability of the cells is important especially for
validation testing. In the next section I will discuss some of the different detection
platform technologies, design, and function, along with their challenges. I will
also cover the major types of detection methods for each of the four areas:

metabolic, genetic, surface recognition, and other features of the pathogen cells.

Detection Platforms

Biochemical detection platforms rely on the metabolic byproducts and
enzymatic activities that are associated with a specific bacterium (figure 1.1).
Many of the assays performed in traditional laboratory detection protocols have
been miniaturized, automated, and placed on cards. This allows for several
assays to be run simultaneously with small samples and greatly reduces the time
for identification. One such platform is the VITEK2 which measures the
fluorescence of probes on the card and uses a pattern recognition database to
identify bacteria [9].

Another biochemical based platform is the electronic nose. This
technology analyzes the specific volatile organic compounds (VOCs) produced
by a bacterium, using complex pattern recognition software [8]. Electronic noses
are made up of arrays of conducting polymers or semi-conducting materials
which are coated or doped with chemical or biological reactants (figure 1.2).
Interaction of the analyte with the array causes a change in the resistance of the
conduction polymers or semi-conducting materials, which is then analyzed and
compared to a database [10, 11]. Other methods of analysis include mass
spectrometry, infrared spectroscopy, and gas chromatography [12]. An obvious
drawback to these platforms is that they require a pure isolated sample of
reasonable concentration. A mixed culture can produce the same biochemical
profile as a single bacterium, as none of the biochemical measurements by

themselves can specifically identify the bacterial strain.

 

0.000.000.0000.
00000000000000.
000.000.00.000.
000000000000...
0.00.00.00.0000
00000000000000.
00.000.000.000.
00000000000093.

 

 

 

 

Figure 1.1 An example of a metabolic detection platform each spot on the chip
represents a different metabolic assay. The white box show an example of an
assay with enzyme linked with a florescent probe. Once the sample is added to
the card, the metabolite reacts with the enzyme resulting in a change in the
florescent probe. The card is then analyzed by the instrument and a metabolic
profile is generated and compared with standards for identification.

 

Figure 1.2 Electronic Nose. In A) isolated bacteria produce VOCs when grown on
specific media. B) The VOCs react with the enzymes (lime) which results in a
change in the resistance of the conduction polymer (blue), which is measured by
the attached electrode (green).

Another biochemical detection platform is based on electrochemical
measurements of enzyme linked electrodes. Detection is typically based on the
inhibition of the enzyme by a bacterial toxin resulting in change of the
electrochemical measurements. One example is the acetycholinesterase-based
electrode biosensor that is used to detect anatoxin-a, a toxin produced by certain
cyanobacteria [13].

Detection platforms that are based on genetic material identify pathogens
by detecting species specific nucleic acid sequences of DNA or RNA. All genetic
detection platforms use some form of amplification, whether by culturing or the
polymerase chain reaction (PCR). Amplification is required due to the low
concentration of analyte typically found in a sample. The exceptions are cases
where large numbers of cells can be isolated from a sample to provide enough
genetic material for analysis. One big weakness of genetic detection is that the
cells are destroyed in order to analyze the nucleic acids. Testing with limited
sample is a one shot deal, with no second chances. This makes it difficult to
determine the pathogen’s viability if at all and impossible to do confirmation
testing.

The basic strategy of genetic detection is the hybridization of species
specific nucleic acid sequences with complementary probe sequences. This
basic strategy makes use of the complementary probe sequences that are either
free in solution or immobilized. Genetic detection platforms that make use of free
complimentary probe sequences make use of polymerase chain reaction (PCR)

methods. There are several different PCR methods that can be used in a

12

platform but only real time PCR techniques offer the speed that is needed for
rapid detection. While not being a true real time detection platform, it is near real
time. There are different real time PCR techniques of varying complexity and
sensitivity, but all are based on the same principle of monitoring the amplification
of species specific nucleic acid sequences typically by the fluorescence
emissions of a fluorophore, which changes through specific or nonspecific
interactions with the amplified nucleic acids [14]. Other methods of detection
include electrophoresis of the amplified sequences and hybridization with
immobilized probes.

Another technique that was developed to determine cell viability is reverse
transcriptase PCR (RT-PCR), in which the RNA of the target pathogen is used.
This is because RNA is rapidly degraded by endogenous RNases and would only
be present if the cells were viable. PCR is a powerful diagnostic tool, and with the
aid of automation it can efficiently and accurately analyze large numbers of
samples in a clinical lab setting [15]. In many cases, however, it is not suitable for
field work as it requires a laboratory environment where quality control can be
maintained and trained personnel can quickly perform the many preparation
procedures needed before sample analysis. In order for PCR techniques to
achieve a high level of sensitivity and specificity, it requires very pure and clean
samples, which lengthens the time it takes to get results (Figure 1.3). Sample
preparation procedures are a challenge to all detection platforms and will be
addressed in a following section. There are portable PCR detection platforms

available, one such platform is the RAZOR EX system made by Idaho

13

Centn'fuge

 

Figure 1.3 Steps involved in a DNA preparation for PCR. In step A the sample is
collected, and the cells are isolated and concentrated. In B. the cells are
resuspended and lysed to extract the DNA and the DNA is isolated. Once the
DNA has been isolated, it can then be use for PCR. Preparation and isolation
procedures vary depending on the sample matrix. Typical laboratory procedures

require a minimum of 60 min.

Technology Inc. The RAZOR EX system offers near real time detections with
results within 30 minutes and detection limits in the range of 1000 cfu/ml [16].
The 30 minutes needed for detection do not include the 5 minute minimum
sample preparation time for powders. Other sample matrices require further
enrichment and purification steps, which then also involve other equipment such
as a vortexer and microcentrifuge. Another setback to PCR based detection is
the cost. PCR is both equipment, chemical, and information intensive technology
which makes for expensive start up and operating costs.

Hybridization platforms are made up of single stranded DNA (ssDNA)
complementary probe sequences immobilized onto a matrix which is suitable for
detection of hybridization [11]. Highly sensitive transducers detect hybridization
of complementary pathogenic DNA by changes in mass, electrical charge, and
optical properties (Figure 1.4). The ssDNA is immobilized on chips, arrays, and
beads. Each is specific for its transducer used in the platform. Hybridization of
complementary sequences can be done with DNA or RNA. While direct detection
of genomic DNA is difficult at low concentrations, RNA has multiple transcripts
that allows for lower detection limits. Often, hybridization methods are coupled
with PCR to increase the number of targets that can be analyzed at the same
time, and to increase sensitivity and specificity [17]. Hybridization methods suffer
setbacks in ease-of-use and robustness similar to PCR in that highly pure and
clean samples are needed as well as properly designed probes to limit

nonspecific binding. The design of probes necessitates large databases,

15

 

Figure 1.4 DNA hybridization platform of single stranded DNA (ssDNA red) is
immobilized onto a sensitive transducer which sends a signal upon hybridization
of complementary pathogenic DNA (blue).

screening and validation of probes. It is the same for PCR primers and probes; in
both cases the results are both time consuming and costly.

Surface recognition based detection platforms make use of molecular
structures that have a high affinity for specific molecular topological structures. A
prime example of surface recognition molecules is immunoglobins (antibodies).
Immunoglobins have the ability to bind to specific molecular topological
stmctures (antigens) on the surface of pathogens. A signal is generated upon
binding of the antibody and antigen, with the signal type dependent upon the
transducer and the method used. These signals can be optical, electrochemical,
mass, and magnetic, to name a few. There are several different methods
employed by various platforms, but are typically used in one of two ways.

The first use immobilized antibodies on a transducer surface, which can
be piezoelectric, nanoelectromechanical and thin metal film. Piezoelectric and
nanoelectromechanical transducers are typically cantilevers, which are very
sensitive to small changes in mass. These cantilevers have been mainly used for
detection of proteins, DNA and small molecules as well as single cells [20, 21].
One shortcoming besides the high cost of fabrication is the lack of robustness of
the sensor - overloading the cantilevers can cause them to break [22]. Thin
metal film works by surface plasmon resonance (SPR), which measures changes
in refractive index of a thin metal film with antibodies immobilized on the surface.
Binding of the pathogen to the antibody causes a change in the refractive index
of the thin metal film [23]. There are commercial SPR instruments and even a

portable kit, the “Spreeta”, made by Texas Instruments. The Spreeta system

17

yields results within 30 minutes and has been shown to have a sensitivity of 102
to 103 colony forming units (CFU)/mL for E. Coli O157:H7.

The second and most popular method of using antibodies is a sandwich
technique. Based on the classic enzyme-linked immunosorbent assay (ELISA),
(figure 1.5) this method also makes use of immobilized antibodies on a surface
that binds the pathogen, then a probe labeled secondary antibody binds to the
pathogen. The probe on the secondary antibody signals or transduces a signal
which is picked up by the detector. The type of probe depends on the detection
platform; in some cases a third antibody is used. The probes can be fluorophors,
optical and magnetic nanoparticles, and enzymes. Common and inexpensive
types of this detector are lateral flow test strips which rely on visual detection.
Neogen’s Reveal lateral flow pathogen test is a good example of these types of
test [40]. While they yield quick results they have poor sensitivity and othenrvise
require pre-enrichment of the sample.

One portable commercially available sandwich-based platform detection
unit is the RAPTOR made by Research lntemational, Inc. The RAPTOR uses
immobilized antibodies on the surface of an optical wave guide. Once the
pathogen binds to the surface, a secondary antibody with a fluorescent probe
can subsequently bind and the fluorescence can be measured to detect the
pathogen [26]. The RAPTOR unit has an assay time of 10 to 15 minutes and a
detection limit of 30 CF U/mL for anthrax spores in water [27].

Advances in sensor and detector technologies have increased the

sensitivity of immunological detection. Antibody technology has also made

18

 

Figure 1.5 ELISA based method. Sample is added to immobilized antibodies
(orange) then it is washed to remove unbound sample. Next a secondary
antibody (pink) linked to a enzyme (lime) is added, then washed again to remove
unbound antibody. Next the enzyme substrate (blue) is added and the product is
produced signaling detection.

advances in recent years resolving many criticisms. New production techniques
eliminate the need for live animal and hybridoma cultures to produce antibodies,
which was a negative aspect of antibody production because continuous
maintenance and limited life span of animals and the associated ethical concerns
restricted long-term collection [24].

The use of antibody and antibody fragment (Fab) recombinant libraries
has allowed for their production in bacteria and phage [25]. This technology not
only increases the speed and volume of production, but also allows for large
numbers of antibodies and Fabs to be screened. The recombinant technology
can also be used to engineer the antibodies or Fabs to make them more robust
or increase their sensitivity and specificity. Single chain antibodies (single chain
fragment variable (scFV)) have been engineered as fusion proteins for various
assays. These scFV fusion proteins can be grown in a ready-to-use manner with
fusion proteins already incorporated so it is unnecessary to add coupling
enzymes or other protein probes to it. Another advantage to antibody based
platforms is the potential to recover live cells which can be used for
conformational testing and/or be cultured for further testing, which is not possible
with post PCR based detection methods. Antibody technology is sound and well
documented and continues to be developed.

Other surface recognition molecules that can be used are aptamers.
Aptamers are small single stranded DNA or RNA molecules that act as ligands to
bind to non-nucleic acid targets [8]. Aptamers are used in detection platforms in

much the same way that antibodies are used to bind a specific target. Aptamers

20

have some notable advantages over antibodies, most importantly that they are
chemically synthesized, which allows them to be made and modified on demand.
Aptamers are discovered using a high throughput selection process called
systematic evolution of ligands by exponential enrichment (SELEX) [19]. This
technique allows one to screen a large library of aptamers against a specific
target and select for those aptamers with high affinity for the targets. Some
aptamers used for pathogen detection range in length from 17to 78 nucleotides
long [18]. Aptamers do not remain linear polymer chains but form secondary
three dimensional structures. Aptamers may replace antibodies in some
applications because of their structural and functional robustness. While
aptamers have been shown to perform well in laboratory testing, they have yet to
be fully tested. They must be tested in field conditions where sample matrixes
could contain inhibitors. This technology is still under development but holds
promise.

Other detection platforms take various forms such as tissue or cell based
detection platforms, and mass spectrometry. One tissue or cell based detection
platform makes use of B-Iymphocytes that have been engineered to emit light
when they come in contact with their target pathogen [28]. The B-lymphocytes
are engineered to express a pathogen specific membrane bond antibody along
with calcium sensitive bioluminescent protein. Binding of antibodies with the
pathogen increases intracellular calcium causing the protein to luminesce. One
obvious problem with this is the amount of time and money that is required in the

upkeep of the cell line as well as the luminometer needed to take measurements.

21

There are several mass spectrometry techniques that are used for
pathogen detection, with matrix-assisted laser desorption/ionization time-of-flight
(MALDI-TOF) mass spectrometry topping the list. MALDI-TOF popularity is due
to its ability to do intact cell mass spectrometry (ICMS) for direct profiling of
pathogens [43], with little sample preparation. The ICMS spectrum that is
generated gives a profile of the protein composition of the pathogen. This
spectral fingerprint can then be compared with reference spectra as well as
protein databases for identification. Other pathogen specific biomarkers besides
proteins can be used in mass spectrometry detection platforms such as nucleic
acid sequences, metabolites, lipids, and polysaccharides. In all cases data bases
must be made for profile comparison. The data bases contain spectral profiles
from known pathogens. Mass spectrometry techniques that do not use ICMS
typically require lengthy sample preparation and/or chromographic separation.
One drawback is the need for concentrated clean samples. The more complex
the sample matrix the higher the limit of detection. Complex sample matrices also
require more complex identification software [8]. As with most chemical
spectrometry, instrument cost is steep and is restricted to laboratory use, due to
the instrument size.

Challenges Shared by All Detection Platforms

The challenges for detection platforms are many. The greatest of the
difficulties is sensitivity; the detection platform must be able to detect low
concentrations of the target pathogen. High sensitivity can be a double-edged

sword: while it allows for detection of low concentrations of the analyte, at the

22

same time, it increases their sensitivity to inhibitors as well, and requires very
clean or pure samples. Such is the case with PCR based detection platforms.
Most detection platforms can analyze powder samples (e.g, spores) with little or
no preparation, but complex matrix samples such as blood, stool, mucus, soil,
food stocks, and water all contain inhibitors and competing agents. This makes
sensitive and specific detection difficult and creates issues with reproducibility as
well [8]. Another problem with smaller detectors is that the sample volume
decreases as well. This can limit the platfomn’s detection limit in low pathogen
concentrations. Specificity is important as well — the ability to specifically identify
the target pathogen from the background and limit false-positives and negatives.
Pathogens must be isolated and purified from the sample matrix before analysis,
and this slows the time it takes to detect. There are several methods for
processing samples, many of which can only be done in a laboratory setting.

Another challenge is making a user-friendly device, which requires little-
to-no training. Many manufacturers have made automated platforms which
eliminate the need for manual preparation. The obstacles to these features are
reduced portability and high cost - not just the initial cost, but also the cost of
continual operation. Mass production and the savings generated by an economy
of scale will reduce prices but not until one platform proves its utility and has the
features discussed in Section 1.
The Potential for Magnetic Detection

Ferromagnetic nanoparticles have a wide range of applications from

magnetic memory storage devices to biomedical probes for MRI. This is due to

23

the unique magnetic properties of ferromagnetic nanoparticles, which are
considerably different from that of their bulk material. Further discussion on
properties of ferromagnetic nanoparticles can be found in Chapter 2.

lmmunomagnetic separation (IMS) is used extensively in the separation
and enrichment techniques used in many detection platforms [30, 31].
lmmunomagnetic separation uses micron or nano size ferromagnetic particles
that have been functionalized with antibodies. When mixed with the target
bacteria and placed in a magnetic field the particles will be pulled to the magnet
along with whatever is attached to the antibody. IMS greatly reduces detection
times by easily removing bacteria from complex heterogeneous mixtures [32].
While several groups have designed sensors for the detection of ferromagnetic
nanoparticles [33, 34, 35, 36, 37, 38] and have proposed possible biosensing
applications, there has been relatively little work done on the use of
ferromagnetic particles for detection of pathogens. This may be mainly because
of the limited use/exposure of magnetic sensing instrumentation in biological
sciences, leaving many unaware of its great potential.

Several groups have been able to detect ferromagnetic nanoparticles
using different magnetic field sensors. The types of magnetic field sensors used
are magnetoresistance, Hall Effect sensors, superconducting quantum
interference devices (SQUID) and pick up coil devices. SQUID is used in a broad I
range of applications from characterizing properties of a material to biomedical
measurements such as magnetocardiographs. SQUID is a very sensitive type of

magnetic sensor with detection limits demonstrated as low as 1017 T [45]. While

24

SQUID are capable of magnetic characterization of materials as well as detection
of FMPs, the instrument is far too large, and its requirement of liquid helium or
nitrogen (for high temperature SQUID) is impractical for portable detection
platforms.

Magnetoresistance (MR) is a change in a material’s resistance in
response to an external magnetic field. There are many MR materials and
devices such as anisotropic magnetoresistance (AMR), magnetic tunnel junction
(MTJ) and giant magnetoresistive (GMR) sensors [39]. Researchers have
developed a GMR sensor capable of detecting 14 superparamagnetic magnetite
(16 nm) nanoparticles at room temperature [46].

Hall Effect sensors measure the transverse voltage caused by a magnetic
field on a current carrying conductor. Japanese researchers have fabricated Hall
sensors for use in biosensor arrays that can detect a single superparamagnetic
nanoparticle (200 nm) [47]. Both of the above sensors are highly sensitive
microfabricated magnetic field sensors. These magnetic field sensors are
capable of detecting single ferromagnetic nanoparticles, and are designed to be
used in a chip based format. These platforms have a static design wherein the
particles must be immobilized on the sensor, which limits the throughput of the
detection platform. These platforms have other issues that must be dealt with
such as removal of nonbonding particles and nonspecific interacting FMPs.

Pick up coils, also called induction coils, are based on Faraday’s Law of
induction, in which a voltage is induced in a loop by a magnet moving through it.

There are many types of sensors that make use of coils, and vary in their design

25

and complexity depending on the method. Researchers have developed a pick
up coil sensor capable of detecting 7 ng of 25nm FMPs by AC susceptibility
measurements [37]. Pick up coils have an advantage over other sensors in that
they are simple and robust making them ideal for sensors in detection platforms.
While the sensors are capable of detecting single FMPs they need to be worked
into detection platforms that are able to detect and discriminate between bound
and non-bound FMPs. One commercially available magnetic detector is the
Magnetic Assay Reader (MAR) by MagnaBioSciences [41]. MAR is a lateral flow
type test with superparamagnetic nanoparticles to label the antibodies. The MAR
uses magnetic field sensors in a pattern called a gradiometer, which move along
the strip. Once the target is immobilized, it can detect the change in the magnetic
field. While this detection platform offers greater sensitivity than other lateral flow
detectors its detection limit is still much higher then what is needed.
Ferromagnetic nanoparticles have many advantages over other molecular
tags or labels. Methods used for the detection of ferromagnetic nanoparticles are
passive (non-destructive). That is detection does not damage the FMPs or the
sample, unlike fluorescence detection methods that use damaging UV light. The
ferromagnetic nanoparticles are chemically stable and maintain their magnetic
signal over time. They do not have the problems of optical systems of bleaching,
and quenching. Magnetic detection systems of ferromagnetic nanoparticle can
offer more quantitative measurements than optical systems. The magnetic
properties of ferromagnetic nanoparticles are proportional to the total amount of

material. Measurements based on magnetization are proportional to the

26

magnetic moment per unit volume, and measure the total volume, not just the
surfaces as in other detection systems. Another major advantage of magnetic
sensing and measurement is that it is a nondestructive method of testing and
evaluating of samples.

As mentioned before the true challenge for all detection platforms is
speed. The time from the moment a sample is taken to the point of detection is
crucial. We will show that our detection platforms can perform at the desired
speeds, and are capable of embodying all of the features demanded of such a
device. Accuracy is also a challenge that must be balanced with speed. We must
keep in mind that there will always be trade offs between speed and accuracy.
With infinite time one could know with absolute certainty, but time is not infinite

and decisions must be made immediately, and actions taken.

27

REFERENCES

1. O. Lazcka, F.J. Del Campo, F.X. Munoz, Pathogen Detection: A
perspective of traditional methods and biosensors. Biosensors and
Bioelectronics Vol:22, pgz1205-1217, 2007.

2. httpzllwww.state.qov/t/ac/trt/4718.htm#treaty Convention on the Prohibition
of the Development, Production and Stockpiling of Bacteriological
(Biological) and Toxin Weapons and on Their Destruction 2008.

3. http://www.fsis.usda.qov/Fsis Recalls/Recall Case Archive/index.ag2
USDA Food Safety and Inspection Service Recall case Archive 2007.

4. PS. Mead, L. Slutsker, V. Dietz, L.F. McCaig, J.S. Bresee, C.Shapiro,
P.M. Griffin, R.V. Tauxe, Food-Related Illness and Death in the United
States, Journal of Envionmental Health. Vol:62, Issuez7, Pg:9-18, 2000

5. J. Guamer', J.A. Jemigan, W.J. Shieh, K. Tatti , L.M. Flannagan, D.S.
Stephens, T. Popovic, D.A. Ashford, B.A. Perkins and SR. Zaki,
Pathology and Pathogenesis of bioterrorism-related inhalational Anthrax
Amercan Journal of Pathology, Vol:163, N02, 2003.

6. National Anthrax Epidemiologic Investigation Team, Investigation of
Bioterrorism-related anthrax, United States 2001: epidemiologic findings.
Emerging Infectious Diseases Vol:8 No:10, 2002

7. PR. Murray, E.J.Baron, J.H. Jorgensen, M.L. Landry, M.A. Pfaller. Manual
of clinical microbiology 9th edition, ASM Press, 2007.

8. D.V. Lim, J.M. Simpson, E.A. Kearns, M.F. Kramer. Current and
developing technologies for monitoring agents of bioterrorism and
biowarfare. Clinical Microbiology Reviews, Vol:18, No:4, pg: 583-607,
2005

9. www.biomeriwux-usa.com 2008.

10. AK. Deisingh, M. Thompson, Biosensors for the detection of bacteria,
Can. J. Microbiol. Vol:50(2) pg:69—77, 2004.

11.K. Arora, S. Chand, B.D. Malhotra, Recent developments in bio-molecular
electronic techniques for food pathogens. Analytica Chimica Acta,
Vol:568, issues:1-2,pg:259-274, 2006.

12. F. Reck, N. Barsan, U. Weimar, Electronic Nose: Current Status and
Future Trends. Chem. Rev. Vol:108 (2), pg:705—725,2008.

28

13. F. Villatte, H. Schulze, R. D. Schmid, T. T. Bachmann, A disposable
acetylcholinesterase-based electrode biosensor to detect anatxin-a(s) in
water. Anal Bioanal Chem, Vol:372 pg:322—326, 2002.

14. RT. Monis, S. Giglio, Nucleic acid amplification-based techniques for
pathogen detection and identification. Infection, Genetics and Evolution
Vol:6(1) pg:2-12, 2006.

15.M.G. Morshed, M.K. Lee, D. Jorgensen, J.L. lsaac—Renton, Molecular
methods used in clinical laboratory: prospects and pitfalls. FEMS
IMMUNOLOGY AND MEDICAL MICROBIOLOGY, Vol:49(2) pg:184-191,
2007.

16. Idaho technology Inc. 390 Wakara Way, Salt Lake City, Utah 84108, USA

17.F. Lucarelli, S. Tombelli, M. Minunni, G. Marrazza, M. Mascini,
Electrochemical and piezoelectric DNA biosensors for hybridisation
detection. Analytica Chimica Acta, Vol:609, issues:2,pg:139-159, 2008.

18. N. Fischer,T.M. Tarasow, J.B.H. Tok, Aptasensors for biosecurity
applications. Current Opinion in Chemical Biology Vol:11, Issuez3, pg:316-
328, 2007.

19. D.H.J. Bunka, P.G. Stockley, Aptamers come of age — at Last. Nature
Reviews Microbiology Vol24, pg:588-596, 2006.

20.A. Johansson, G. Blagoi, A. Boisen, Polymeric cantilever-based
biosensors with integrated readout. Applied Physics Letters, Vol:89(17)
pg:3505.1-3505.3 2006

21 . B. llic, D. Czaplewski, M. Zalalutdinov, single cell detection with
micromechanical oscillator. Journal of Vacuum Science & Technology B,
Vol:19(6) pg:2825-2828, 2001.

22. AK. Gupta, P.R. Nair, D. Akin, Anomalous resonance in a
nanomechanical biosensor. PNSA, Vol: 1 03 Issuez36 pg: 1 3362-1 3367,
2006.

23.J. Waswa, J. lrudayaraj, C. DebRoy, Direct detection of E. Coli O157:H7
in selected food systems by a surface plasmon resonance biosensor.
LWT-Food Science and Technology, Vol:40 lssue:2 pg:187-192 2007.

24.S.S. Iqbal, M.W. Mayo, J.G. Bruno, B.V. Bronk, C.A. Batt, J.P. Chambers,
A review of molecular recognition technologies for detection of biological
threat agents. Biosensors & Bioelectronics Vol:15, pg:549—578, 2000.

29

25.8. Krebs, R. Rauchenberger, R. Silke, et al. High-throughput generation
and engineering of recombinant human antibodies. Journal Of
lmmunolgical Methods, Vol:254 lssuez1-2 pg:67-84 2001.

26.C.C. Jung, E.W. Saaski, D.A. McCrae, B.M. Lingerfelt, G.P. Anderson,
RAPTOR: a fluoroimmunoassay-based fiber optic sensor for detection of
biological threats. IEEE Sensors Journal, Vol:3 lssue:4 pg:352-360, 2003.

27. Research intemational,lnc 17161 Beaton Rd SE, WA 98272-1034 360-
805-4930 http://www.resrchintl.com/, (2008)

28.T.H. Rider, M.S. Petrovick, F.E. Nargi, et al.A B Cell-Based Sensor for
Rapid Identification of Pathogens, SCIENCE, Vol:301 lssue25630 pg:213-
215, 2003.

29. M. Magnani, L. Galluzzi, l.J. Bruce, The use of magnetic nanoparticles in
the development of new molecular detection systems. Journal of
Nanoscience and Nanotechnology, Vol:6 lssue:8 pg:2302-2311, 2006,

30. I.M. Hsing, Y. Xu, W.T. Zhao, micro- and Nano- magnetic particles for
applications in biosensing. Electroanalysis Vol:19 lssue:7-8 pg:755-768,
2007.

31 . H.W. Gu, K.M. Xu, C.J. Xu et al. Biofunctional magnetic nanoparticles for
protein separation and pathogen detection. Chemical Communications,
Vol:9 pg:941-949, 2006.

32. D. Horak, M. Babic, H. Mackova, et al. Preparation and properties of
magnetic nano- and micronsized particles for biological and environmental
separations. Journal of Separation Science Vol:30 Issuez11 pg:1751-1772,
2007.

33.D.L. Graham, H.A. Ferreira,P.P. Freitas, Magnetoresistive -based
biosensors and biochips. Trends in Biotechnology, Vol:22 lssue:9 pgz455-
462, 2004.

34. L. Ejsing, M.F. Hansen, A.K. Menon, et al. Planar hall effect sensor for
magnetic micro- and nanobead detection. Applied Physics Letters, Vol:84
lssuez23 pg:4729-4731, 2004.

35. DR. Baselt, G.U. Lee, M. Natesan, et al. A biosensor based on

magnetoresistance technology. Biosensors & Bioelectronics, Vol:13
lssue:7-8 pg:731-739, 1998.

30

36.G.X. Li, V. Joshi, R.L. White, et al. Detection of single micron sized
magnetic beads and magnetic nanoparticles using spin valve sensors for
biological applications. Journal of Applied Physics Vol:93 lssue:10
pg:7557-7559, 2003.

37.T.Q. Yang, M. Abe, K. Horiguchi, K. Enpuku, Detection of magnetic
nanoparticles with ac susceptibility measurements. Physica C:
Superconductivity, Vol:412-414, part 2 pg:1496-1500, 2004.

38.G. Mihajlovic, P. Xiong, S. von Molnar, et al. Submicrometer hall sensor
for Superparamagnetic nanoparticle detection. IEEE Transactions on
Magnetics, Vol:43 lssue:6 pg:2400-2402, 2007.

39.8. Parkin, X. Jiang, C. Kaiser, et al. Magnetically engineered Spintronic
sensors and memory. Proceedings of the IEEE, Vol:91 lssue:5 pg:661-
680, 2003.

40. Neogen Corportion, 620 Lesher Place, Lansing, MI 48912

41 . MagnaBioSciences a Quantum Design Campany 6325 lusk Boulevard,
san Diego, CA 92121-3733

428. Shafazand, R. Doyle, S. Ruoss, et al. Inhalation Anthrax: Epidemiology,
Diagnosis, and management. CHEST, Vol:116 lssue:5 pg:1369-1376,
1999.

43.P. Lasch, H. Natterrnann, M. Erhard, et al. MALDl-TOF Mass
Spectrometry compatible inactivation method for highly pathogenic
microbial cells and spores. Analytical Chemistry, Vol:80 lssue:6 pg:2026-
2034, 2008.

44. MB. Prentice, L. Rahalison, Plague. Lancet vol:369 pg:1196-1207, 2007.

45. R.L. Fagaly, Superconducting Quantum interference device instruments
and applications. Review of Scientific instruments,Vol:77(10) pg:101101-
101101-45, 2006.

46.S.X. Wang, G. Li, Advances in giant magnetoresistance biosensors with
magnetic nanoparticle Tags: review and outlook. IEEE Transactions on
Magnetics, Vol:44 No:7 pg:1687-1702, 2008.

47.Y. Kumagai, Y. lmai, M. Abe, et al. Sensitivity dependence of hall
biosensor arrays with the position of superparamagnetic beads on their
active regions. Journal of Applied Physics Vol:103, lssue:7, pg:07A309
2008.

31

Chapter 2
The Big Idea: Magnetic Modification of Bacteria
Introduction

The overarching goal of this project is to design and fabricate a functional
pathogen detection platform based on basic physical laws that uses simple
electronic, chemical and biological technologies, and is capable of the real time
detection of a single bacterium. We first begin with a single bacterium and build
up our methodology and detection platform around it — a more detailed
discussion on the specific features, concepts, and components of the detection
platform will be presented in the following sections.

Real time detection of a single bacterium is not an easy task. There is very
little in the way of measurable or inducible signals that would allow detection, let
alone be specific enough for species identification. The obvious solution then is
to add a signal to the bacterium. This can be accomplished through numerous
probes of varying design. The difficulty is finding a probe that increases the
signal of a single bacterium sufficiently to allow detection over the background.
Magnetotactic bacteria can produce magnetic fields that are 20 to 30 times that
of the earth’s magnetic field (50 pT) [1], and some with field as high as 0.6 T [2].
There are several types of magnetic sensors with sensitivities far below those
needed for the detection of magnetotactic bacteria ranging from nT to 1T. In our
discussion of ferromagnetic nanoparticles in chapter 4 we will see that cobalt
nanoparticles can be made that are smaller but produce fields equal to those of

magnetotactic bacteria. Putting this all together it becomes evident that detection

32

of a single bacterium is possible with the use of magnetic probes and a magnetic
sensor

Our strategy is to make use of the large magnetic field produced by
ferromagnetic nanoparticles to detect the bacteria. In our strategy, ferromagnetic
nanocomposite particles (FMNCP) (figure 2.1) that contain surface recognition
elements that can detect and bind to specific bacterial cell surface structures are
used even if known to contain many complex surface antigens. The FMNCPs
utilize the bacterium’s surface as a scaffold to coat the bacterium and effectively
make a micro-magnet (figure 2.2) which can be detected by several types of
magnetic sensors. Our method of detection is based on Faradays Law of
Induction: V: N AB/t
Faraday’s Law states that the voltage V induced in a system is proportional to the
number N of turns in the coil, to the area A of the coil, and the magnetic flux B
passing through the coil, while it is inversely proportional to the time t it takes to
pass through the coil [3]. This fundamental physical law makes up the basis of
our detection method. We can imagine the same concept but on a much smaller
scale where we have a micrometer sized magnetized bacterium passing through
a coil and inducing a voltage which we measure, the detector itself is made up of
a simple coil attached to a very sensitive voltmeter. We now have a very
sensitive, fast, and simple detection method. In order to control sensitivity we can
change the number of loops or the speed at which the bacterium passes through
the coils to increase the voltage. Now we must incorporate this detection method

into a platform that will allow the bacterium to come into contact with the

33

 

Figure 2.1 FMNCP architecture. The FMNCP is made of a metal nanoparticle
core (green), coated with a macromolecular polymer system (multicolored
spheres). With antibodies/ surface recognition molecules (pink).

34

Figure 2.2 A) The bacterium is mixed with the FMNCP suspension. As the
bacterium mixes with the FMNCP suspension the FMNCPs coat the cell’s
surface (B). The FMNCPs interact cooperatively as they bind, transforming the
bacterium in to a micromagnet.

35

Bacteria

 

Bacteria

 

FMNCPs and then pass through the coil. Our detection platform consists of
FMNCPs in a suspension medium that is continually circulating in a loop that
passes through the coil (Figure 2.3). When the target bacterium is introduced into
the system the FMNCPs bind and encapsulate it as it circulates. As the FMNCP-
encapsulated bacterium passes through the coil, it induces a voltage per
Faraday’s Law, signaling that the bacterium has been detected.
Design of FMNCP

There are many features that most be incorporated in to the design of the
FMNCP. These include the ability to identify specific cell surface chemical
structures, contain a signal element for detection; forms stable suspensions, and
contain functionality which allows for derivatization. The ability to identify specific
cell surface chemical structures is done by the use of surface recognition
molecules, such as antibodies. The method of detection is magnetic induction so
the signal element must be magnetic. The metal nanoparticle core will be made
up of the ferromagnetic metal cobalt. The ability to form stable suspensions, and
contain functionality which allows for derivatization is a difficult challenge, that
leads us to the design of a macromolecular polymer system.
Macromolecular polymer system

The macromolecular polymer system features are, it must (1) show
chemical properties similar to biomacromolecules such 3 proteins, (2) incorporate
sites of functionality that can bind transition metals, (3) contain functionality that

facilitates attachment of antibodies, and (4) contain

37

Nanocomposite

suspension L_|C0il r_|

..
$7

 

Figure 2.3 The lay out of the general design of the detection platform. The
detection platform consists of an induction coil attached to a voltmeter, through
which a FMNCP suspension (green) passes through which is circulated in a
continuous loop by the pump. Once the target bacterium enters the system in
encounters the FMNCPs becoming coated with them, at the same time it is
swept through the coil, and induces a voltage signaling detection.

38

functionality that prevents aggregation. The chemical structure of the
macromolecular polymer system can be seen in figure 2.4 it has been designed
to incorporate all of the above features. The macromolecular polymer system is a
highly functionalized, polypolar chiral polymer. It has many structural and
chemical similarities with that of biomolecules, with a polyamide backbone with
side chains made up of carboxylic acid and alcohol groups. The large polyamide
chain is a peptide mimetic polymer. The polyamide backbone can also aid in
secondary structure formations. The polymer backbone also contains thioethers,
sulfer atoms have high affinity for transition metals; the numerous thiols in the
backbone of the macromolecular polymer system provide multiple points of
attachment to the metal nanoparticle core, allowing the polymer scaffold material
to envelop the metal nanoparticle core. Once the macromolecular polymer
system has bound to the metal nanoparticle, the polymer is stabilized by its polar
backbone and side chains in aqueous suspension and behaves like a protein.
Stabilization of the nanoparticle suspension is also greatly affected by the charge
of the carboxylic acid group. The overall negative charge of the
polymer/nanoparticle complex is not only favorable in an aqueous suspension,
but it prevents agglomeration of the nanocomplexes through electrostatic
repulsion of like charges. The Carboxylic acids share the same charge as many
bacteria do which reduces nonspecific bonding. In addition to providing stability,
the side chains of the polymer scaffold material also provide an excellent scaffold

for further chemical modification.

39

Amide bond

Carbox lic acid

 

Figure 2.4 Structures of macromolecular polymer system. Top) Chemical
structure of the polymer, middle) ball and stick, bottom) space filling. The
macromolecular polymer system consists of a chiral backbone made up of a
polyamide bonds and thioethers. The polymer scaffold side chains are made up
of carboxylic acid and alcohol groups.

40

Detection platform Design

The overall design of our detection platform (Figure 2.3) is rather simple, it
consists of an induction coil attached to a voltmeter, through which a FMNCP
suspension passes and is circulated in a continuous loop by a pump. Our vision
for this platform is a miniature, simple to use, and portable detector that is small
enough to be incorporated into a cell phone or other personal electronics as
discussed in Chapter 1.

The simplicity of our detection platform allows for its components to be
manufactured individually and then assembled in the platform, which lets us use
off-the-shelf technology. With the many technologies available today, many of the
components are easily miniaturized and readily available from a variety of
manufacturers. There are several companies that specialize in high quality coils,
as well as micro pumps, which fit the size and performance requirements of our
detection platform. This should speed the production and make fabrication more
economical. The only true difficulty lies with the miniaturizing of the voltmeter.
There are many small commercially available voltmeters, but their range of
detection is limited to the millivolt range. This creates a challenge for us as our
system will create a current in the nanovolt range. While many of the features on
larger nanovolt meter are unnecessary and will not need to be incorporated into
the final device, it is still difficult to miniaturize and maintain such a low detection
limit. The use of custom designed integrated circuits (IC) will greatly aid the
miniaturization of the volt meter. For increased functionality, other electronic

components will be incorporated allowing for data storage and analysis. The

41

device will be programmable with a variety of functions. Another solution is to use
a different detector such an atomic magnetometer.

The shrinking of the components of the device to make a smaller
instrument is a key step in making a versatile and functional detector. A device
that is easily held in your hand, clipped to your belt, or slipped into a pocket,
much like a cell phone is the ideal size. The increasing computing and memory
capacity of smaller and smaller electronics and modern microfabrication,
decrease cost, while increasing the capabilities of the detection platform. These
advancements allow for on board data storage and analysis, wireless capabilities
with down loads, updates and alerts.

Theoretical Calculations and Expected values

In this section, theoretical calculations are used to demonstrate that the
mode of operation of the device is physically possible and within the realm of
today’s technological capabilities. Please note many of the calculations are done
as theoretical maximum and some assumptions are made in order to simplify the
calculations. More complicated calculations and simulations could be used but
the number of variables and the broad range of magnetic properties brought
about by small changes in nanoparticles make it an act of futility, unless all of the
variables are known. For this reason it is better in some cases to do
experimentation first and then try and determine which properties are playing the
crucial parts.

In order to determine the voltages induced by a bacterium coated with

ferromagnetic nanoparticles as it passes through a coil, we can first start by

42

looking at what voltage should be induced by a single ferromagnetic nanoparticle
(FMNP). Faraday’s law (equation 1) [4] can be used to calculate the voltage
induced, where N is the number of turns in the loop, CD is the magnetic flux
moving through the coil, and dt is the time it takes to pass through the loop:

V = -N(d<D/dt) 1.
In our calculation we will assume that the FMNP passes through the loop at its
center with its magnetic field perpendicular to the plane of the loop. The particle
will be treated as a point magnetic dipole (Figure 2.5). In order to calculate the
voltage using equation 4 we must determine dd); all of the other variables are

known. The magnetic flux is given by equation 2:
CD =1 8 . dA 2.

where A is the area of the loop and B is the magnetic flux density given by
equation 3 [5]:
B = y/47r(-m/r3 + (mom/,5) 3.

Where I). is the permeability of free space, m is the magnetic moment and r is the
distance from the point magnetic dipole to the loop. Integrating equation 3 over
the area of the loop at radius R in the z direction we get equation 4. This gives us
the magnetic flux through the loop in the z direction at distance 2 away (Figure
2.6).

42(2) = mn/Z[(R2+Z2)’”2 — Z2(R2+Z2)'3/2] 4.
from which we derive

dCD/dz = W2 [(4/3)Z3(R2+22)'5/2 - 3Z(R2+22)'3’2 1 5.

43

 

Figure 2.5 The drawing above shows the FMNP (red) moving down in the z axes
through a single loop (green) with its magnetic moment (m) perpendicular to the
loop, which is in the x-y plane.

44

 

3.005-19 1

2.50549 ‘ }
l

2.00E-19

 

 

 

 

 

1.50E-19

1.00E-19 ‘ i
l l
I l \
l T .

5.00E-20

 

Magnetic Flux
\.
/

 

 

 

 

 

 

 

 

 

 

 

 

Z Distance (m)

Figure 2.6 Graph of equation 4 showing the magnetic flux of a single FMNC as a
function of distance from the loop.

45

The voltage due to the changing flux
V = -N(d<l>/dt) = -N(d<D/dz) - (dz/dt) 6.

may be calculated as

v = -N*,um/2 [(4/3)(v*t)3(R2+(v*t)2)’5/2 — 3(v*t)(R2+(v*t)1/2] * v 7.
Equation 5 gives us the rate of change in the magnetic flux as a function of Z
(Figure 2.7). Now if we substitute equation 5 into equation 6 and let 2 from
equation 4 equal v*t where v is the velocity of the FMNP and t is time, we get
equation 7 which gives us the voltage as a function of time as it passes through
the loop. The graph of equation 10 (Figure 2.8) of the voltage vs. time has a peak
voltage of 1 .36x10 12V for a single FMNP passing though the coil at 53 m/sec.
Assuming that all of the FMNP are bound to the surface of the bacterium and are
aligned in parallel; then the theoretical maximum voltage induced by a single
bacterium coated with 15,000 FMNP would be 2.04x10'8 V. This value is based
on estimated and predicted values of a proposed detection platform in order to
get theoretical maximum values and not actual experimental data (see appendix
III for calculations). The predicted voltages are well within the lower range (1x10‘9
V) of many commercially available voltmeters. We must also remember that if we
want to increase our signal we can always add more turns to the coil, and speed
up the flow of the suspension. While orientation of FMNCPs may be an issue,
one can easily solve this by adding an external field, and in fact it would be

beneficial to impose an external field regardless of the size of the particles.

46

2.75E-14
2.25E-14
1 .75E-14
1 .25E-14
7.50E-1 5
2.50E-1 5
-2.50E-1 5

Change In Flux

-7.50E-1 5
-1 .25E-14
-1 .75E-14
-2.25E-14

 

-2.75E-14
95

e e e e e e e e e e
9 9 9 9 9 9 9 9 9 9
eé" $960 eéo e6" eé’ 991° eé” 9°“ e99 63% eé" 9° eé’ 692'

o be )0 a. be ”a (be 5. 6. 6. $0 9'

2 Distance (m)

be

0)

Figure 2.7 Graph of equation 5 showing the change in the magnetic flux of a
single FMNC as a function of distance from the loop.

47

1 .6E-12
1 .4E-12
1 .2E-1 2
1 .0E-12
8.0E-1 3
6.0E-13
4.0E-13
2.0E-13
0.0E+00
-2.0E-13
-4.0E-13
-6.0E-1 3
-8.0E-1 3
-1 .0E-12
-1 .2E-12
-1 .4E-12
-1 .6E-12

Voltage (V)

 

eeee«e«eeeee
9 9 ”@909 9e>§9©9§9¢99§9

9° 9° 9°
3 6’" 8‘" 6'0 0‘0“}?

123‘
)‘9'9'9’9‘ 9 °

‘3' K. N. I». We 50 5. ‘0
Time (Sec)

0

Figure 2.8 Graph of equation 7 showing the voltage induced by a single FNP
passing through a single loop.

48

Noise

As with most electronic devices, there are both internal and external
sources of noise. Many external sources of noise are mainly electromagnetic
waves/fields caused by modern electronics; luckily, this noise is easy to remove
by proper shielding of the device as well as use of low/high bandpass filters.
There are also biological sources of noise such as magnetotactic bacteria in the
environment, however chances of encountering them in everyday sampling are
very small. When considering noise there are three major sources: chemical,
instrumental, and environmental noise.

Environmental noise in the form of electromagnetic radiation is
everywhere, saturating our world. Cell phone towers, microwave towers,
satellites, radios, high power lines, and much more, all produce electromagnetic
radiation (there are also natural sources such as the earth’s magnetic field and
solar flares). The background of electromagnetic radiation in our environment
can make low level voltage measurements difficult without proper grounding and
shielding. It is no surprise that nearly all electronic devices require some form of
shielding because the 110 V that is used in every home in the US can induce
millivolt voltages. Shielding consists of surrounding the instrument and/or its
components (circuits, wires, other) with a conducting material that is grounded to
the earth, the shielding structure is known as a Faraday cage. Proper grounding
and shielding of instruments can eliminate much of the environmental noise

caused by electromagnetic radiation. In our detection platform using a Faraday

49

cage to house the detector-shielded wires is a simple solution to environmental
noise from electric fields but not magnetic.

Instrumental noise in low level measurements used by our detection
platform can mainly be attributed to thermal (Johnson) noise and thermoelectric
effects (thermocouple). Thermal noise arises from thermal agitation of the charge
carrying elements of the instrument. The thermal agitation causes charge
inhomogeneities that in their turn cause voltage fluctuations. The magnitude of
the thermal noise for the platform can be determined using equation 8 [6].

VRMS = \/4kTRF 8.
Where k is the Boltzmann constant, T is the temperature (K), R is the source
resistance (ohms) and F the noise bandwidth (Hz). The second major contributor
to the instrumental noise is the thermoelectric effects which are small potentials
generated by differences in temperature at junctions of dissimilar metals. These
can arise from junctions of the same metal as well. The thermal EMF that can be
generated is given by equation 9.

E = Qab(T1-T2) 9.
Where Qab (0.2 pV/°C for Cu-Cu junction) is the thermoelectric coefficient with
respect to the two metals in the junction and T1 and T2 is the temperature of the
respective metals. Using equation 9 we can see that a difference in just one
degree in a copper-copper junction can yield an EMF of 0.2 W. While
thermoelectric effects can produce large EMF, they can be minimized by using
clean, same metal junctions that are maintained at the same temperature. In both

cases of instrumental noise for our detection platform, proper calibrations of the

50

device for the temperature it is to be operating at allows for noise to be minimized
and accounted for.

For this paper, chemical noise will refer to the noise induced by the free
nanocomposites in suspension. The noise from the free FMNPs assuming that
they are independent of each other is given by Poisson noise. We assume that
there are only two orientations that the FMNPs can take with their magnetic
moment perpendicular to the loop: either up or down. The number of particles in
the coil at any one time will be around 3. 78x105 FMNPs when using a 1nM
solution. The Poisson noise produced from these particles can be expected to be
about +/- 5.92x10’0 V. This is approximately 2.9% of the voltage that is expected
to be induced by the coated pathogen. We can also expect the noise to be less
than this, due to the tumbling time of the free FMNPs. A 30 nm FNP has a
tumbling time of 21.6 psec and the time for a particle to pass through the loop is
37. 7 usec. This means that, the time that it takes the FMNPs to pass through the
loop, they have tumbled through every possible orientation at least once. As
stated earlier the tumbling time of a bacterium is 0.15 see, which is much slower
than that of the free FMNPs, and then the concern is its orientation to the loop.
This problem may be solved by using a series of loops and not just a single loop

for detection.

51

Reference:

. C. N. Keim, M. Farina, U. Lins,. Magnetoglbus; Magnetic Aggregates in
Anaerobic Environments. Microbe, Vol:2 pg: 437-445.

. D. Schuler,. Eds. Magnetoreception and magnetosomes in bacteria;
Springer-Verlag Berlin Heidelberg 2007 pg 8.

. D. Jiles, Introduction to magnetism and magnetic materials, Chapman and
Hall, New York. 1991

. D. Halliday, R. Resnick, J. Walker, Fundamentals of Physics Fifth Edition.
USA (1997)

. J. Marion, W. Homyak, Physics for science and engineering part 2. New
York (1982)

. D. A. Skoog, F. J. Holler, T. A. Nieman, Principles of instrumental
analysis; Saunder College Publishing: United States, 1998.

52

Chapter 3
Macromolecular polymer system

Introduction

The macromolecular polymer system is a new polymer that has never
been made before. It has been designed to accomplish many features. The
ferromagnetic nanocomposite particles consist of a ferromagnetic nanoparticle
metal core encapsulated by the macromolecular polymer system, through which
the SRMs are covalently linked. In designing the macromolecular polymer
system for synthesizing the nanocomposite particles, we are in search of a
material with many robust attributes. Nanoparticles that are left uncapped tend to
agglomerate and precipitate out of suspension quickly whereas capped particles
remain in suspension much longer [1, 2, 3]. There are a multitude of capping
agents available with a broad range of chemical makeups and sizes, from small
polar molecules to large hydrophobic molecules. In order for a compound to be a
capping agent it must have the proper chemical features that will allow it to do
two things. One chemical feature is that it must have an affinity for the metal
nanoparticle and remain associated with it (bonded to it). The second chemical
feature is that it must be soluble in its surrounding solvent; this favorable
interaction of capping agent with its surrounding solvent slows the precipitation of
nanoparticles out of suspension. There is also a third chemical feature, that while
not required in a capping agent, is an advantageous feature especially when
dealing with magnetic nanoparticles. The feature is one that prevents

agglomeration of nanoparticles such as electrostatic repulsion, which slows the

53

precipitation nanoparticles out of suspension. All of these features help to
maintain and increase the longevity of the suspension.

The polymer scaffold material is an excellent capping agent because, not
only does it embody all of the above features, it also has other features that
make it suitable for our purpose. It has structural and chemical similarities with
that of biomolecules, with a polyamide backbone with side chains made up of
carboxylic acid and alcohol groups. The sulfur atoms in the polymer backbone
have a high affinity for transition metals and provide binding interaction to the
metal core allowing the polymer to encapsulate the nanoparticle. And the
Carboxylic acids have the same charge of bacteria which reduces nonspecific
bonding, prevents aggregation, and are a point for chemical modification. The
macromolecular polymer system is a unique and novel polymer not only by way
of its structure but its synthesis as well [4, 5]. Syntheses of high molecular weight
polyamides is not an easy task, and not viable in step wise methods. The
macromolecular polymer system made using 2(H)—furanone and L-cysteine
methyl ester, both of which are organic compounds that are derived from natural
sources. The synthesis of the polymer uses mild conditions and no organic
solvents, making it an environmentally friendly as well as inexpensive process.

The reaction takes place in a stepwise manor (figure 3.1) first, with the
sulfur of the L-cysteine methyl ester attacking the double bound on the 2(H)-
furanone in a Michael addition, to form a Iactone monomer. The Iactone
monomer then polymerizes in a head to tail manner with the amine group

attacking at the carbonyl, and opening the Iactone ring. After polymerization the

54

H o
Hm] SH 'OH H\NH
+ \ O
o g o
/o /O

HO O OH HO 0 OH
O O

/ N s N s/
H H
O

O OH OH
Figure 3.1 Reaction mechanism of macromolecular polymer system synthesis.
First a Michael addition of the L-cysteine methyl ester (red) to the 2(H)-furanone

to form a Iactone monomer. Next the Iactone monomer polymerizes in a head to
tail manor and then the methyl ester is remove.

macromolecular polymer system is deesterified, and size specific membrane
fractionation using molecular weight cut offs (MWCO) of 3,500 and 13,000 Da is
done.
Experimental

Synthesis of macromolecular polymer system A and B. Sodium
bicarbonate (0.84 g, 0.01 mol) and L-Cysteine-methlyester (1.71 g, 0.01 mol)
were dissolved in water (20 mL). Once dissolved Furanone (0.84 g, 0.01 mol)
was added quantitatively using water (5 mL). The reaction mixture was left at 25
°C for 1 hour with stirring. At the end of one hour the reaction mixture was
rotovapped to remove the water, and then placed in an oil bath for 3 hours at 90
°C. Then the reaction mixture was heated at 90 °C for 18 hours for further
polymerization. The polymer mixture was then dissolved in a water solution and
placed in 3,500 MWCO dialysis membranes and dialyzed for 24 hours, changing
the water every 2 hours for the first 8 hours, then every 4 hours. The solvent was
then removed to yield the macromolecular polymer system A. The same
procedures were fallowed for macromolecular polymer system B expect that after
the polymerization the polymer mixture was dissolved in a water solution of
sodium hydroxide (0.2 M) for one hour with stirring at 25 °C to remove the methyl
ester. And placed in 3,500 MWCO dialysis membranes and dialyzed for 24
hours, changing the water every 2 hours for the first 8 hours then every 4 hours.
The solvent was removed to yield the macromolecular polymer system B. All

NMR spectra were taken using a Varian lnova 300 NMR.

56

Results and discussion

NMR spectroscopy of the macromolecular polymer system was used to
determine the reaction mechanism (figures 3.2, 3.3, 3.4) and characterize the
macromolecular polymer system structure (figure 3.5). The NMR at time 0 of the
reaction mixture show the chemical shifts of both starting materials. After 15 min
the chemical shifts of the furanone have decreased by more then 2/3 and new
chemical shifts for the Iactone monomer have appeared. After 30 min the
furanone shifts are nearly gone. Looking at the NMR for the final polymer the
chemical shifts for the Iactone are gone and the new shifts of the polymer
backbone can be seen, this and the broadening of these chemical shift
confirmers polymerization of the macromolecular system. IR spectroscopy (figure
3.6) of the macromolecular polymer system as well as electrospray mass
spectrometry (figure 3.7) was performed. The IR spectra shows the amide l and
II bands, as well as the starches for the alcohol and carboxylic acid side chain
groups, which further confirms amide bond formation and polymerization of the
polymer. In the mass spectrum we see a typical distribution of fragments for a
polymer, and that the major fragments of are multiples of 205 Da which is the
mass of the monomer unit -1, which confirms the repeat structure of the
macromolecular system. The use of size specific membrane fractionation also
insures polymerization and high molecular weigh materiel greater then 3,500 and
13,000 Da. The end product is a high molecular weight macromolecular polymer
system that is a peptide mimetic with a number of chemical functionality, is

inexpensive, and easy to synthesize.

57

 

 

 

1; iii ii, v .
A 2”,) h...) be: “0......2

-1 H"

' 5 4

 

 

1" 1'
H7" H 811
H
ofic
o
H
d
C
‘0‘ .L E
l' T ' ' '01 i ‘I
3 2 l I!“

 

Figure 3.2 1H NMR of and L-Cysteine-methylester and Furanone reaction at time

0.

58

 

 

' T=15min

 

l

 

 

 

 

r—‘I' I i- —T— T T r r T I ""‘"‘“ "" "“'— ’ - . "-‘ ' v ‘ ['1

7 i 5 I 3 2 l m
Figure 3.3 1H NMR of and L-Cysteine-methylester and Furanone reaction after

15 minutes.

 

 

 

59

 

T=30min ‘ a

   
 

 

. r "2 4 ' ‘ a; 2 ' 2“" “a!
Figure 3.4 1H NMR of L-Cysteine-methylester and Furanone reaction after 30
minutes ‘

 

 

FLT—~T—W-' T" 7 """""° "'

 

6O

 

 

 

 

 

Final polymer .‘ 0 °" .8.
rrc H

“H: s *
I H H "

H a“ H

O\ H d

H
b I
c :4: a"? e
4* Jr iyi‘s/I RM," xvii i/j \Vc... #9.;

 

 

Figure 3.5 1H NMR of final macromolecular polymer system after polymerization
and deesterification.

61

60

 

 

Amide II T
1570-1515 N

1710-1680
0
4 OH 1680—1640

~30

~10

 

 

T I I I

3498 2998 2498 1998 1498 998
Wavenumbers

Figure 3.6 IR spectroscopy of macromolecular polymer system.

62

498

Transmittance

.Ecfigm 8520a .m_:oe_oEoLomE *0 E2603 wmmE Eamoboom Em 959“.

Sun . 8mm Dawn 80..” comm oo_mN . omNN econ Shr . comp omNr .ooow own com omN

  

    

     

     
 
 

       

   

ME 0

 
 

vdov

 

999 W 3.5
P 532.88 «68
«.35 5mm
«98
:o :0 o W
o ,, .
T— : W
m z Hm z
w z w
I
I: H/ o H/ , 38
o o o: :o o o: W

355.: W

.2:

63

REFERENCES

. S. Link , M.A. El-Sayed, Shape and Size dependence of radiative, non-
radiative and photothermal properties of gold nanocrystals. INT. Reviews
in Physical Chemisty, Vol:19 lssue:3 pg:409-453, 2000.

. C.J. Xu, S.H. Sun, Monodisperse magnetic nanoparticles for biomedical
applications. Polymer lntemational, Vol:56 lssue:7 pg:821-826, 2007.

. P. Majewski, B. Thierry, Functionalized magnetite nanoparticlses-
synthesis, properties, and bio-applications. Critical Reviews in Solid State
and Material Sciences, Vol:32 pg:203-215, 2007.

. Cao, P. Development and Synthesis of Novel Poly (B-Amino acid) Drug
Delivery Systems; Thesis Michigan State University, 2004.

. US. Patent # 6153724, Preparation of cross-linked 2-dimensional
polymers with sidedness from a,B-lactones. 2000.

64

Chapter 4
Nanocomposite Material

Introduction

The optical and magnetic characteristics of nanoparticles differ greatly as
compared to their bulk material. These differences are dependent on the size
and shape of the nanoparticles, this dependence allows one to control the optical
and magnetic characteristics of nanoparticles by changing the size and/or shape
of the nanoparticles. The high surface to volume ratio of nanoparticles results in
a larger percentage of the total atoms displayed on the surface. As a particle gets
smaller, the dispersion, or the fraction of atoms on the surface, increases, and
scales with the surface area divided by the volume [1]. The cohesive energy,
which is the average bond energy of a solid, can be seen to decrease as the
particle gets smaller. This is due to the fact that the atoms at the surface have
lower coordination and fewer bonds than the atoms in the interior which lowers
the average bond energy of the particle. The lowering of the cohesive energy is
one factor that gives rise to the unique properties of nanoparticles such as lower
melting points. The optical properties of nanoparticles are due to two separate
phenomena: quantum confinement and surface plasmon. Quantum confinement
effects cause increasing separation of energy levels as a particle decreases in
size similar to a particle in a box [1, 2]. Quantum confinement effects on metal
nanoparticles are typically observed in particles smaller than 2 nm, and are
distinct from surface effects. The surface plasmon of a nanoparticle, however, is

affected by the surface, and is a result of the interactions of the conducting

65

electrons at the particle’s surface with the electric field of a light wave. This leads
to a strong absorption of light, with high molar extinction coefficients in the range
of 1x109 M’1cm'1, which is several times higher than that of strong absorbing
organic dyes [2]. The core of the FMNCP must be made up of a ferromagnetic
material such as iron, cobalt, or nickel. The core of our FMNCP is made up of a
cobalt nanoparticle. Cobalt has a magnetic moment of 161 Am2/kg second only
to iron [3]. The magnetic properties of nanoparticles are greatly influenced by
their shape and size as well, and are discussed in the next section.

The optical and magnetic properties of nanoparticles make them ideal
components to be incorporated into our nanocomposite material. Nanoparticles
have distinctive advantages over traditional materials, one being that they don’t
undergo irreversible reactions when a measurement is made, allowing for
measurements to be repeated and the recycling of materials. All of the different
components of the F MNCP serve specific functions and are integrated to form a
simple composite of complex function.

Ferromagnetic Nanoparticles

In bulk ferromagnetic materials domain wall formation is driven by the
balance between the magnetostatic energy, which increases proportionally to the
volume of the material, and the domain-wall energy, which increases
proportionally to the interfacial area between domains [4]. As a nanoparticle’s
size is decreased the surface-to-volume ratio changes and domain wall formation
becomes unfavorable. The area over which the energy would be most highly

distributed is larger than the particle itself and the result is a single domain

66

ferromagnetic nanoparticle. These single domain ferromagnetic nanoparticles
can be thought of as a giant magnetic dipoles with unique magnetic properties
dependent on size and shape. Ferromagnetic nanoparticles have much higher
spin densities than that of their bulk material. Also asymmetries at the surface of
the nanoparticles cause enhanced anisotropy, increasing the anisotropy energy.
The size at which this happens varies with the material and shape of the
nanoparticle; for spherical cobalt nanoparticles the critical diameter is 80 nm [5].
As the ferromagnetic particles get smaller they reach a second critical diameter.
This second critical diameter is one at which their magnetic properties change
drastically. This is due to the lowering of the anisotropic energy as the particle
gets smaller. At the second critical diameter the thermal energy (at room
temperature) becomes equivalent to that of the anisotropic energy allowing the
magnetization of the particle to rapidly fluctuate from one direction to another.
This results in ferromagnetic nanoparticles that have a net magnetic moment of
zero and are paramagnetic. These nanoparticles are said to be
superparamagnetic; spherical cobalt nanoparticles become superparamagnetic
below 10 nm. The exchange energy still holds strong and maintains the
cooperative alignment of the point magnetic dipole unlike paramagnetic materials
where the thermal energy is proportional to or greater than the exchange energy.
In these particles the reversal of magnetization takes place by coherent rotation,
that is all of the dipoles remain parallel as the direction changes.
Superparamagnetic materials fall inbetween paramagnetic and ferromagnetic

behavior, having a much higher magnetization (M) and saturated magnetization

67

(Ms) than paramagnetic materials and having coercivity and remanence of zero
[6].

The anisotropy energy of a single domain nanoparticle can be determined
using the following equation [7]: 5(0) = Keff*v sin26.
Where Keg is the anisotropy constant, V is the volume of the particle and 9 is the
angle between the magnetization and the easy axis; where Keg V is the energy
barrier that must be overcome to reverse the magnetization. For cobalt the
calculation is much simpler due to its hexagonal (uniaxial) anisotropy. That is
cobalt only has one easy axis due to its hexagonal crystalline structure. In the
case of iron the calculation is not as simple; the cubic anisotropy must be
calculated from the magnetization of the three crystal easy axis. The easy and
hard directions of magnetization along the crystal axis are the result of spin orbit
coupling [8]. The Neel-Brown expression can be used to give the relaxation time
of the magnetic moment of a particle, using the anisotropy energy and the
thermal energy [7]:

r = To 9(Kefi' V/ka) note ro=1093

The Neel-Brown expression is used to determine if a ferromagnetic nanoparticle
is in a paramagnetic state or a blocked state. That is, if the relaxation time is
shorter than the experimental, then it is superparamagnetic. The transition from
ferromagnetic to superparamagnetic for a given particle size is called the
blocking temperature [9]. Superparamagnetic nanoparticles are very useful
because they do not agglomerate like ferromagnetic particles do until placed in a

magnetic field of suitable strength.

68

Interaction of Ferromagnetic Nanoparticles

The function of our detection platform will be dependant upon the
interaction of ferromagnetic/superparamagnetic nanocomposite particles
attached to the bacterial cell surface. In order to better understand the
phenomena one might expect to see, we shall first take a look at some bacteria
with naturally occurring ferromagnetic nanoparticles.
Magnetotactic bacteria are aquatic bacteria that orient and swim along magnetic
field lines, a behavior known as magnetotaxis [10]. Magnetotactic bacteria make
up a diverse collection of prokaryotes but all contain magnetosomes.

Magnetosomes are intracellular structures made up of mineral crystals of
magnetite (Fe304) or greigite (Fe384) surrounded by a lipid bilayer [11].
Magnetoscme crystals range in size from 35 to 120 nm long and are single
domain nanomagnets. The magnetosomes are arranged in a chain longitudinally
across the cell, with the number of magnetosomes varying, but averaging 20
[10]. It has been shown by Frankel et al. [10] that the interaction of the
magnetosomes with the Earth’s magnetic field supplies enough energy to
overcome the thermal agitation and cause an average orientation parallel with
the Earth’s magnetic field. The alignment of the magnetotactic bacteria allows
them to swim along the earth’s magnetic field lines, a phenomenon called
magnetotaxis. The magnetotaxis allows the bacteria to swim down in the water

column to find the oxic—anoxic interface and anoxic regions that they inhabit.

69

In order to determine the orientation of a magnetic dipole in an external
field like those in magnetotactic bacterium or nanoparticles one can use the
Langevin function: cos 9 = L(a)

Where 9 is equivalent to the average angle between the magnetic moment and
the magnetic field and L (a) is the Langevin function:

L(a) = cth(a) - (0)" a = mH/kT
Where L(a) is a function of the ratio of the thermal energy and the Zeeman
energy (Figure 4.1); the Zeeman energy is the energy of the magnetic dipole
moment interacting with an external magnetic field and is given by the equation:

E = chose

The Zeeman energy is a function of the magnetic moment m, the external field H
and the angle 9 between them. The magnetic moment depends on the particle’s
volume and the material it is made of: the number of atoms and the size of their
atomic moments [12]. Figure 4.2 shows the energy of a cobalt nanoparticle
interacting with the Earth’s magnetic field as a function of the particles diameter.
The energy increases exponentially as the particle gets larger, as one would
expect because the number of moments added increases proportionately to the
volume. We can also see that it takes a rather large particle (= 50 nm) to equal
that of the thermal energy of 4.14x1021 J. While one single particle may not
have enough energy to align itself with the earth’s field, one must remember that
the energy is dependent on the number of atomic moments, which can be

increased simply by

70

 

 

 

 

 

 

 

 

 

 

 

 

mHIkT

Figure 4.1 The Langevin equation plotted as a function of the ratio mH/kT.

71

 

ze-zo
l

1.8E-20 le—w 3——__-.l _W.. .._JW_

1.65-20 W 4*- W- ,W W

1.45-20 -W- W *7 ,W .l —.

 

 

1.25.20» ,W fi- 9 4— i. W

15.204 47* =4 — e. W e a.-- _i- - ,W

 

Energy (J)

 

 

8E-21 ~~~

 

 

 

 

 

6E-21 ~- ~ «A 7 -~ W ”—7 J- W_ _--W W

 

 

 

 

 

 

4E-21277 :7 a A -4 A e- a- ”e .7

 

 

 

25.21 I - - —- a» 4——-_ E r- __ .. L. ._s

 

 

 

 

 

 

 

o — 4 4
0.00E+00 1.00E-08 2.00E-08 3.00E-08 4.00E-08 5.00E-08 6.00E-08 7.00E-08 8.00E-08
Particle diameter (m)

Figure 4.2 The energy of a cobalt nanoparticle interacting with the Earth's
magnetic field as a function of its diameter (green) and the thermal energy (pink).

72

adding more. That is, the total Zeeman energy can be increased by adding more
particles. This can be seen in the Magnetotactic bacteria where a single
Magnetosome is not sufficient and is made up of 20 or more magnetite
nanoparticles. The reason for the difference between the size and number of
magnetite and cobalt nanoparticles are that the magnetic moment of a cobalt
nanoparticle is greater then that of magnetite nanoparticle of equal mass.

Another energy that comes into play is the magnetostatic energy which is
dipole-dipole interactions. The magnetostatic interaction energy for two magnetic
dipoles is given by the equation [13]:

Ems = m1m2(1-cosze)/R3

Where m1 and m2 are the magnetic moments of the two dipoles, R is the distance
between them and 6 is the angle between them. The maximum range of values
for the magnetostatic energy is :l: 2| m1llm2|/R3 where the negative sign
represents an energy minimum where the magnetic dipoles are aligned parallel
and in the same directions. The positive sign represents an energy maximum
where the magnetic dipoles are aligned parallel but in opposite directions. The
area between these two maximums is separated by many smaller maximums
and minimums. In the Magnetotactic bacteria the moments of the magnetite
nanoparticles are aligned parallel, minimizing the magnetostatic energy and
maximizing the Zeeman energy. It is known that concentrated assemblies of
nanoparticles have stronger magnetostatic effects then dilute assemblies,
showing ferromagnetic ordering at room temperature [8]. These calculations are

done based on single non-interacting particles, which work well for dilute

73

samples with narrow size distributions. When samples become concentrated
predicting magnetic interactions and length of scale becomes complicated. It is
difficult to discern which interactions are responsible for the observed magnetic
properties, because in many cases several interactions can contribute to the
same properties. The interactions of these assemblies will depend on their
proximity to each other, which is dependent on their polymer coating. The closer
they are to each other the stronger the interaction.

While the magnetic properties of individual non-interacting ferromagnetic
nanoparticles can be well characterized based on size and shape, these
properties change as large assemblies form and begin to interact. That is
individual ferromagnetic nanoparticles that are superparamagnetic at room
temperature can, when concentrated, also form ferromagnetic ordered
assemblies with a net magnetic moment.

FMNCP and Bacterium Binding

The effective binding of the FMNCP to its target bacterium presents three
specific issues: specificity, affinity, and coverage. In order to discuss these issues
we must first discuss how the FMNCPs bind the bacterium. The binding of the
FMNCPs to the bacterium is accomplished by the use of surface recognition
molecules (SRM), such as antibodies that are conjugated to the FMNCP. Surface
recognition molecules bind to unique molecular structures presented on the
surface of the bacterium known commonly as antigens when antibodies are
used. Nearly all bacteria display unique molecular structures on their surfaces

that are strain specific. These unique molecular structures are either polypeptide

74

or polysaccharide in composition depending on the bacterial species, with
polysaccharide being the most common. One can use these unique molecular
structures as antigens to the SRMs to provide specificity for a certain species.
The issue of affinity, or how well the FMNCPs bind the pathogen, is addressed
by the use of more than one SRM. The FMNCP can be formulated with different
SRMs at ratios that reflect the pathogen’s surface makeup. The use of SRMs that
bind the same antigen but in different ways, in the fashion of polyclonal,
antibodies can add to the FMNCPs affinity for a particular pathogen. The affinity
of the individual SRMs can also be optimized using the methods and technology
discussed in Chapter 1, such as antibody recombinant technology. The final
issue is choosing an antigen that is sufficiently distributed throughout the cell
membrane and is found in high enough numbers in order to maximize the signal.
Capsular polysaccharides (CPS), also known as K-antigens, are just such
surface molecules (antigens), and can be found on a wide range of bacterial
species [14]. The CPS with its unique chemistry and structure acts as a physical
barrier protecting the cell against dehydration and is an important virulence factor
in the pathogenicity of many bacteria [15]. Capsular polysaccharides have been
used for serotyping bacterial species for well over fifty years. Serotyping is used
to determine specific strains of bacterial species. Both gram positive and
negative bacteria can have CPS. While the CPS can be used to identify specific
species, it is not always strain specific. For example in our experiments we used
Rhizobium Trifolii as our model organism. Rhizobium Trifolii is a gram negative

soil bacterium capable of fixing nitrogen [16]. The CPS of Rhizobium Trifolii was

75

used as an antigen to generate antibodies/SRMs to the Rhizobium Trifolii to be
used in the detection platform. The structure of the CPS produced by Rhizobium
Trifolii has been determined (figure 4.3) [17, 18, 19]. While the CPS of Rhizobium
Trifolii is species specific, it is not strain specific. That is it can be used to identify
Rhizobium but it can not be used to differentiate between different strains of
Rhizobium.

In cases where the CPS does not have the specificity that is needed or is
not present other polysaccharides are used such as Lipopolysaccharides (LPS)
(O-antigens) of gram negative bacteria, and cell wall glycopolymers (CWG) of
gram positive bacteria [21] as well ,as capsular proteins. The structure of the LPS
of Rhizobium Tn'folii has been solved, along with several other surface
polysaccharides of Rhizobium (figure 4.4) [ 22, 23, 24, 25, 26, 27]. LPS are made
of three parts, the lipid A which acts as the hydrophobic anchor in the outer
leaflet, the core which is made up of nonrepeating oligosaccharide and the 0-
antigen which consists of repeating polysaccharide units [21]. There is some
debate about the use of SRM to identify bacteria, and that the surface molecular
structures can randomly change or be engineered by man, making them a poor
choice. This is not correct; as mentioned the CPS is an important virulence factor
and changes will directly affect the pathogenicity. The cell surface molecules of
pathogenic bacteria are essential to their virulence and are not likely to change.
The truth is that variation in the makeup of living organisms is a result of entropy.
The thermodynamic laws that govern our world favor disorder, which is why you

will never find two identical bacteria. There will always be some difference in

76

OH OH OH OH
O O
o o 0
' — '0 O O O O
0 HO HO HO

n
OH OH OH HO O.
O OH \
HO
OH
0
O
OH
HO
OH
OOH 0

Figure 4.3 Structure of octasaccharide repeating unit of Rhizobium trifolii
capsular polysaccharide [17, 18, 19].

77

 

H
O
HO
0 o 0 OH
O HO O
0 l

 

o 0 B
OH OH WW
0
H O
HO H
OH
H OH 0
H HO
HO H
OW C
0

Figure 4.4 Structures of Rhizobium polysaccharides. Total Rhizobium trifolii 4S
LPS Structure (A) [21, 22, 23, 24], sulfoquinovosyl diacylglycerol (B), digalactosyl
diacyl glycerol (C) [27, 28].

78

DNA, proteins, carbohydrates, etc. This is why questions of random changes or
genetic modification in target surface molecules are trivial. One cannot plan or
anticipate all possible chance happenings, and no one method of detection is
exempt from this, whatever the specific target maybe it is subject to the
consequences of entropy.

Experimental

Synthesis of Gold Nanoparticles - The polymer scaffold materials A and B
(0.335 g) were dissolved in water (6.65 g) to make a 3.35% w/w solution. In a 12
well tissue culture plate, a gold (Ill) chloride (825, 412, and 206 uM) solution was
placed in the first 3 columns decreasing in concentration from top to bottom rows.
The fourth column contained equivalent volumes of water. Polymer scaffold
material B (20 pL) was added to each well in the first column and row one of the
fourth column. Polymer scaffold material A (20 pL) was added to each well in the
second column and row two of the fourth column. Borane in THF (3 drops) was
added to each well of the plate. The experiment was repeated, substituting
sodium borohydride (0.01 M) as the reducing agent.

Synthesis of cobalt nanoparticles - Cobalt nitrate (30 mM) was added to
six vials labeled A 1-6 (0.3, 0.2, 0.1 0.3, 0.2, 0.1 mL) and adjusted to a final
volume of 3 mL. A second set of six vials labeled B 1-6 had sodium hydroxide
(0.2 M, 0.1 mL) added to each vial. To the second set of vials number 1-3
polymer scaffold material B (0.06 mL, 5% w/w)
was added. The second set of vials was adjusted to a final volume of 3 mL.

Sodium borohydride (0.01 M, 0.2 mL) is added to all six of the first set of vials

79

containing the cobalt nitrate solutions. After adding and mixing the sodium
borohydride, the corresponding numbered vial of the second set is then added
and mixed.

Antibody synthesis - Keyhole limpet hemocyanin (10 mg) and Rhizobium
trifolii CPS (2 mg) are dissolved in a sodium acetate buffer solution (pH 4) with
stirring at 25 °C for 1 hour. Sodium cyanoborohydride (2 mg) was added to the
mixture and left to stir at 25 °C for 36 hr. The reaction mixture was dialyzed using
3500 MWCO tubing for 48 hrs, changing the water every 2 hours for the first 8
hours, and then every 4 hours. The solution was then lyophilized and sent to
Antibodies Incorporated. Antiserum produced was made in rabbits and consisted
of one test bleed and one production bleed. Binding of antibodies was checked
using heat treated Rhizobium trifolii and F ITC labeled anti-rabbit antibody.

Nanocomposite Synthesis - Cobalt nanoparticle suspension (60 mL) was
concentrated down to 15 mL. The suspension was then dialyzed against water
(500 mL) using a dialysis chamber with continuous flow (0.5 mL/min). The
suspension was then dialyzed against a solution of sodium phosphate (20 mM,
pH 4, 300 mL) using the same dialysis chamber with continuous flow. Dialyzed
suspension (1 mL) is then added to two micro tubes which each contain 1-ethyl-
3-[3-dimethylaminopropyl] carbodimide (EDC) (30 mg), and the mixture was left
to stir for 1 hour. The mixture is then centrifuged (30 sec, 6000 rpm), the
supernatant is discarded, and the pellet is resuspended in phosphate buffered
saline (PBS) (1 mL). Antibody serum (200 pL) is then added to one tube and an

antibody solution (3.5 mg/mL, 200 pL) added to the other as a control, and was

80

stirred 18 hours at 4 °C. The mixture is then centrifuged (30 sec, 6000 rpm), the
supernatant is discarded, and the pellet is resuspended in phosphate buffered
saline (PBS) (1 mL), and the suspension is stored at 4 °C. SEM samples were
prepared in one of two ways. First Rhizobium trifolii cells were heat fixed to a
carbon coated glass slide, to which one to two drops of the nanocomposite
suspension was added and incubated for 1 hr. Then the samples were washed
first with PBS and then with water, and let dry. In the second method the
nanocomposite suspension was mixed with a suspension of Rhizobium trifolii
cells, and left for 18 hours at 4 °C. The samples were then centrifuged (30 sec,
6000 rpm), the supernatant discarded, and the pellet resuspended with water (1
mL). The samples were then pipetted (5 pL) to carbon coated slides for SEM and
nickel grids for TEM and let dry.
Results and discussion

Once the macromolecular polymer system had been synthesized,
experiments could begin using it to make nanoparticles. The first experiments
were done using gold. As mentioned before, sulfur groups have a high affinity for
transition metals with gold having the highest. The ability of the polymer scaffold
to stabilize metal nanoparticles in suspension was first tested using gold as the
core of the nanoparticle. Gold nanoparticles suspensions also have distinctive
colors depending on the size of the particles, which is a good litmus test as to
whether or not nanoparticles have formed. The first experiment used two forms
of the polymer, one with free carboxylic acid groups, and one in which the

carboxylic acid groups were converted to methyl ester groups. Figure 4.5 shows

81

Gold Polymer ester Polymer
solution + gold + gold

Polymer

Decreasing
Gold
concentration

 

Figure 4.5 Gold nanoparticle suspensions and controls. The 13 column is a rugold
control with no polymer, the 2nd is gold with the methyl ester polymer, the 3 is
gold with the carboxylic acid polymer, and the 4th has both polymers with no gold
as controls. The concentration of gold decreases as you move from the top of
the column to the bottom. The gold solution were reduced using sodium
borohydride.

82

the gold nanoparticle suspensions and their controls. The gold control samples
with no polymer precipitated out, whereas the gold/polymer samples formed gold
nanoparticle suspensions. The methyl ester polymer formed purple colored
suspensions whereas the carboxylic acid polymer formed more gold to straw-
colored gold suspensions. Over a few days the methyl ester polymer did begin to
show some precipitation at the highest concentrated suspension. The
suspensions were dried and formed a film on the bottom of the plate. Pieces of
the film from the dried suspensions were subjected to fluorescent confocal
microscopy (Figure 4.6). The fluorescence of the film pieces at different
wavelengths indicated that there was a distribution of different sized
nanoparticles within the film. Images of the gold nanoparticles tells us that the
macromolecular polymer system stabilizes gold nanoparticles in a aqueous
suspension. The gold color of the gold nanoparticle suspension indicates that
particles smaller then 10 nm were formed.

Once the macromolecular polymer system was shown to be capable of
facilitating the formation and stabilization of gold nanoparticles, it was next used
in experiments with ferromagnetic metals. Experiments using cobalt,
macromolecular polymer system solutions yielded stable nanoparticle
suspension (figure 4.7). The cobalt nanoparticles were also shown to have
ferromagnetic properties; when a magnet was placed next to the sample, the
particles were pulled to, and accumulated on the, side where the pole of the
magnet was located (figure 4.8). Small amounts of precipitation were observed in

some of the samples after 1 to 2 weeks, though all suspensions maintained their

83

H H

:1] I'm 5"] [till i

 

Figure 4.6. Fluorescent confocal microcopy images of a dried piece of the gold
nanoparticle suspension. The top left shows the florescence with 488 nm light;
the top right at 633 nm. The bottom left is at 543 nm, and the bottom right is an
overlay of all three. Images taken with a Zeiss LSM 5 PASCAL confocal
microscope.

84

 

Figure 4.7. Cobalt nanoparticle suspensions (1-3) were made with polymer and
cobalt solutions mixed and reduced using sodium borohydride. The controls (4-6)
contained the same concentrations of cobalt solution and were reduced using
sodium borohydride. The controls did not have any polymer in them.

85

N S

 

Figure 4.8. Cobalt nanoparticle suspension after a bar magnet has been
removed form its side. There is a dark line where the particles have been
concentrated in response to the magnet.

86

gold color. Many of the suspensions were stable after 1 year. Stability seems to
depend on concentration and pH, with higher concentration samples precipitating
faster, as one would expect from a particle stabilized by a protein mimetic
polymer. The suspensions would precipitate at pH 3 and below but were stable at
neutral to high pH. The sizes of the cobalt nanoparticles was determined to be on
average 3 nm using a tunneling electron microscope (T EM) (figure 4.9).

We have seen that the macromolecular polymer system stabilizes both
gold and cobalt nanoparticle suspensions. The stability of the nanoparticles is
dependent on concentration and pH, with suspension stability ranging from a few
weeks to over a year. Suspensions of higher concentration tended to aggregate
and precipitate faster then those of lower concentration. At higher concentrations
the frequency of interaction increased, which increased aggregation. The
suspension’s stabilities also varied with pH; they were stable in basic, neutral,
and weak acidic solutions, but quickly precipitated out at pH 3 and below. This is
due to the decrease of the negative charge on the particles; as the pH is lowered
the number of protonated carboxylic acid groups increases which reduces the
charge on the particles allowing them to aggregate.

An interesting result was that some of the cobalt nanoparticle suspensions
depending on their concentration after some time (more then two weeks), formed
needle-like structures. Using the confocal microscope to look at these newly
formed structures, the needle-like structures were determined to be tubules,
which had formed from sheets of the polymer that had rolled up upon

themselves. Cobalt nanoparticles were incorporated into the polymer sheets and

87

 

. 4 mm-
Figure 4.9 TEM image of dried cobalt nanoparticle suspension with particle and
average particle size of 3 nm. The image was taken with a JEOL 100 CXII.

88

tubules, making them magnetic. At higher magnification, thousands of micron
and submicron sized particles were observed moving (Brownian motion) around
and inside the tubules. The micron and submicron sized particles were mainly
concentrated inside the tubules and at their ends. This concentration of particles
was due to the magnetic field that is made by the tubules. The particles, when
exposed to an external magnetic field, became ordered and lined up with the
magnetic field (figure 4.10). The magnetic particles in and around the tubules are
clusters of cobalt nanoparticles whose magnetic forces had caused aggregation
despite their charged polymer coating.

TEM images showed that the metal core of the cobalt nanoparticle has an
average diameter of 3 nm. From this we can calculate the particle’s magnetic
energies as well as some projected values for comparison with experimental
results. Knowing that the core is less then 3 nm we may assume that it is
superparamagnetic, nevertheless we can still calculate the anisotropy energy.
The anisotropy energy for a 3 nm cobalt particle is 5.8x10'21 J. This is a little
bigger then the thermal energy of4.14x10'21 J. We can use the Neel-Brown
expression to determine if a ferromagnetic nanoparticle is in a paramagnetic
state or a blocked state. Remember that the state is relative to the experimental
time. Our experimental time was 0.1 s and the relaxation time was determined to
be 4.06x10'9 3, which is far less than the experimental, meaning that the cobalt
core particles are superparamagnetic.

In some of the cobalt samples, needle like structures formed. Closer

examination of the needle structures showed them to be hollow with micron size

89

Ne Magnetic Field Magnetic Field

      

_E.'_.
Figure 4.10. The confocal image on the left is of one of the tubules with particles
randomly concentrated inside of it and at the ends. The image on the right is the
same tubule but in a magnetic field, the particles have become aligned with the
magnetic field (bar scale 20 um).

90

and smaller particles moving in and out of them and concentrated at the ends of
the needles. Exposure of the sample to a magnet caused the needles as well as
the particles to align with the magnetic field. The hollow magnetic needles are the A
result of polymer sheets formed from the polymer scaffold material that has rolled
up on its self. As the sheets are formed they incorporate the cobalt nanoparticles
making them ferromagnetic. The earth’s magnetic field causes the magnetization
of the hollow needles, which in turn act like hollow bar magnets. The particles
seen in the sample are aggregates of the cobalt nanoparticles, as the particles
get bigger, so does the net moment, increasing its magnetization in the earth’s
field. The magnetized particles are pulled to the ends or poles of the needles.
Brownian motion prevents the particles from fully aligning and keeps the particles
moving along the field lines of the needles.

One might view this as an interesting side note, but I think that it helps to
demonstrate an important point. That is, the magnetic behavior of the cobalt
nanoparticles in large assembles. We know that the cobalt nanoparticles are
superparamagnetic and the magnetic energy of the earth’s magnetic field and an
individual particle’s field is less than the thermal energy. The magnetic energy
can increase as the particle gets bigger (more and more particles add to the
aggregate) to the point where the magnetic energy is higher then the thermal
energy. We can see this in our results, that the particles as well as the needles
are magnetized by the earth’s magnetic field. The magnetic energy is strong
enough to overcome the thermal agitation of the cobalt nanoparticles. We can

expect to see similar results with the nanocomposite bound to the bacterium. The

91

nanocomposites by themselves will have no net moment in the earth’s field but
as they accumulate on the bacterium the net moment will increase to the point
where the magnetic energy is stronger than the thermal energy causing the
bacterium to align with the earth’s field.

In order to make the final nanocomposite, we must obtain antibodies for
our model organism Rhizobium trifolii. The CPS of Rhizobium trifolii was
conjugated to the carrier protein keyhole limpet hemocyanin (KLH). The CPS
conjugate was used for the production of the antibodies by Antibodies
Incorporated. Antibodies Incorporated produced the antibody serum used. Two
test immunopurifications were done on the serum, one using a salt gradient and
the other using a pH gradient. In order to determine what fractions contained the
antibodies, AUN 843 cells were heat fixed to microscope slides and incubated
with the fractions, then they were incubated with an anti rabbit antibody labeled
with FITC. The slides were then examined under a fluorescent microscope. The
slides showed poor separation of fractions. The pure serum was then tested
along with purified antibodies (figure 4.11). While the antibody serum was shown
to bind the Rhizobium trifolii it did not have the intensity of the purified antibodies.
This is due to differences in concentration between the serum and purified
antibodies. The serum, while binding the CPS, is just not concentrated enough to
show the same intensity.

Once the binding was confirmed, the antibodies could be incorporated into
the final nanocomposite. The antibodies were conjugated to the polymer coated

cobalt nanoparticles using the coupling agent 1-ethyl-3-[3-dimethylaminopropyl]-

92

Purified antibody

 

Figure 4.11 Antibody evaluation using FITC-labeled goat anti-rabbit antibody of
antibodies raised to capsular polysaccharide.

93

-carbodimide (EDC), to make the final nanocomposite. The final treatment and
modification does decrease the stability of the nanocomposite in suspensions,
which tend to form large aggregates which can be resuspended. The binding of
the nanocomposite with the Rhizobium trifolii was tested by mixing the two
together, and samples were placed on nickel grids for TEM and carbon coated
glass slides for SEM. Figure 4.12 shows the Rhizobium trifolii control SEM
image. Figure 4.13 shows the nanocomposite and Rhizobium trifolii samples, the
top image shows a large agglomeration of the nanocomposite with the
Rhizobium trifolii cells. The bottom image is of several Rhizobium trifolii cells
coated with nanocomposite. Figure 4.14 shows the cells coated with the
nanocomposite as well as large nanocomposite aggregates. Figure 4.15 shows
the Rhizobium trifolii control TEM image. Figure 4.16 is a TEM image of the
nanocomposite and Rhizobium trifolii sample, showing four cells coated with
nanocomposite. Figure 4.17 shows the TEM image of the nanocomposite and
Rhizobium trifolii mix, the nanocomposite has large concentrations around the
sides of the cells.

The side chains of the polymer scaffold provided attachment sites for the
conjugation of antibodies to form the nanocomposite. The specificity of our
device relies on polyclonal antibodies as opposed to monoclonal. Polyclonal
antibodies will give the nanocomposite some flexibility in binding. By not requiring
the exact conformation for binding the antigen it will increase the number of

antibodies capable of binding. Conjugation of the antibodies to the nanoparticles

94

 

Figure 4.12 SEM image of heat fixed Rhizobium trifolii cells.

95

Figure 4.13 SEM images of nanocomposite and Rhizobium trifolii mix samples.
The top image shows a large agglomeration of the nanocomposite with the
Rhizobium trifolii cells. The bottom image is of several Rhizobium trifolii cells

coated with nanocomposite.

96

FIGURE 4. 13

 

 

1 [.l m

Figure 4.14 SEM image of nanocomposite and Rhizobium trifolii mix, the cells
are coated with the nanocomposite as well as large nanocomposite aggregates.

98

 

Figure 4.15 TEM image of Rhizobium trifolii control.

99

         

Figure 4.16 TEM image of the nanocomposite and Rhizobium trifolii sample,
showing four cells coated with nanocomposite.

100

 

Figure 4.17 TEM image of nanocomposite Rhizobium trifolii mix. With a large
concentration of nanocomposite around the cell.

101

decreases the nanocomposites electrostatic repulsion, increasing the rate of
agglomeration and precipitation. Pipetting and vortexing is sufficient to resuspend
the samples. The SEM images of the nanocomposite and Rhizobium trifolii cells
clearly show the nanocomposite coating the cells surface. The cells of the control
appear flat looking almost like shadows on the slide, as one would expect to see
little difference between them and the carbon background. In contrast, the
images of the Rhizobium trifolii mixed with the nanocomposite has a 3D surface
and contrast with the background. The large contrast between the control and
experimental samples is due to elemental differences. The large difference in
atomic number of the cobalt to the carbon results in the contrast observed in the
samples. We can see that the concentration and thickness of the nanocomposite
can vary on the cells, with light areas indicating higher concentration. We also
see that there are large aggregates of nanocomposite that are bound to the cells
as well. The TEM images support what is seen in the SEM images. In the control
images we can make out the cells but they are relatively transparent. In the
images of the Rhizobium trifolii mixed with the nanocomposite, the cells are
much more visible, appearing darker than the controls due compared to the

nanocomposite coating.

102

REFERENCES

6. S. Link , M.A. EI-Sayed, Shape and Size dependence of radiative, non-
radiative and photothermal properties of gold nanocrystals. INT. Reviews
in Physical Chemisty, Vol:19 lssue:3 pg:409-453, 2000.

7. E. Roduner, Size matters: why nanomaterials are different. Chemical
Society Reviews, Vol:35 lssue:7 pg:583-592, 2006.

8. American Institute of Physics Hand book, McGraw-Hill book company Inc,
1957, pg:5-208.

9. AH. Lu, E.L. Salabas, F. Schuth, Magnetic Nanoparticles: Synthesis,
Protection, Functionalization, and Application. Angew. Chem. Int.
Ed, Vol:46 lssue:8 pg:1222-1244, 2007.

10.K.M. Krishnan, A.B. Pakhomov, Y. Bao, P. Blomqvist, Y. Chun, M.
Gonzales, K. Griffin, X. Ji, B.K. Roberts, Nanomagnetism and spin

electronics: materials microstructure and novel Properties. J Mater Sci 41
pg:793-815, 2006.

11.6. Li, V. Joshi, R. L. White, S. X. Wang, J. T. Kemp, C, Webb, R. W.
Davis, S. Sun, Detection of single micron sized magnetic beads and
magnetic nanoparticles using spin valve sensors for biological
applications. Journal of Applied Physics, Vol:93, #10 pg:7557-7559, 2003.

12A. Lu, E.L. Salabas, F. Schuth Magnetic nanoparticles: Synthesis,
Protection, Functionalization, and Application. Ange. Chem. Int. Ed. Vol:46
pg:1222-1244, 2007.

13.S.A. Majetich, M. Sachan, Magnetostatic interaction in magnetic
nanoparticle assemblies: energy, time and length scales. J. Phys. D: Appl.
Phys. Vol:39 png407-R422, 2006.

14.Y.W. Jun, J.W. Seo, J. Cheon, Nanoscaling laws of magnetic
nanoparticles and their applicabilities in biomedical sciences. Accounts of
chemical research. Vol:41, No2 pg:179-189, 2008.

15. RB. Frankel, Magnetic Guidance of Organisms. Annual Review of
Biophysics and Bioengineering, Vol:13 pg:85-103, 1984.

16.D.A. Bazylinski, R.D. Frankel, Magnetosome Formation in Prokaryotes.
Nature Reviews Microbiology Vol:2 lssue:3 Pgesz217-230, 2004.

17.C. Kittel, J.K. Galt, Ferromagnetic Domain Theory. Sold state physics
Vol:3, 1956.

103

18. DC. Mattis, The Theory of Magnetism Made Simple. World Scientific,
New Jersey, 2006.

19. IS. Roberts, The biochemistry and genetics of capsular polysaccharide
purduction in bacteria. Annu. Rev. Microbiol. Vol:50 pg:285—315, 1996.

20.Y.S. Ovodov, Bacterial capsular antigens structural patterns of capsular
antigens. Biochemistry (Moscow), Vol:71, N029, pg:937-954, 2006.

21 . F .B. Dazzo, G.L. Truchet, Interactions of Lectins and their saccharide
Receptors in the Rhizobium-Legume Symbiosis. J. Membrane Biol.
Vol:73, pg:1-16, 1983.

22. RI. Hollingsworth, M. Abe, J. Shenlvood and FE. Dazzo. Carbohydrate
Research Vol:134 pg:C7-C11 (1984).

23. RI. Hollingsworth, A. Mort and PB. Dazzo. Journal of Bacteriology
Vol:169 pg:3369-3371, 1987.

24.Hollingsworth, R.I., and Dazzo. F.B., Hallenga, K. Musselman, B.
Carbohydrate Research, Vol:172, pg:97-112, 1988.

25. C. Weidenmaier, A. Peschel, Teichoic acids and related cell-wall
glycopolymers in Gram-positive physiology and host interactions. Nature,
Vol:6 pg:276-287, 2008.

26.C.R.H. Raetz, C. Whitfield, Lipopolysaccharides Endotoxins. Annu. Rev.
Biochem, Vol:71 pg:635-700, 2002.

27. RI. Hollingsworth and R.W. Carlson, J. Biol. Chem Vol:264 pg:9300-9304,
1989.

28. RI. Hollingsworth, R.W. Carlson, F. Garcia and D. Gage. J. Biol. Chem,
Vol:264 pg:9294-9299, 1989.

29.The structure of the O-antigenic chain of the lipopolysaccharide of
Rhizobium trifolii 43. Wang, Y. Hollingsworth, R.l. Carbohydrate Research.
Vol:260, issue:2 pg 305-317, 1994.

30.Y. Wang Ph.D. Thesis, Michigan State University East Lansing, MI, 1995
31.Y. Tang, R.l. Hollingsworth, Digalactosyl diacylglycerols, plant glycolipids
rarely found in bacteria, are major membrane components of bacteroid

forms of Bradyrhizobium japonicum. GlycoBiology Vol:7, no:7, pg:935-
942, 1997.

104

32. RA. Cedergren, R.I. Hollingsworth. Occurrence of sulfoquinovosyl
diacylglycerol in some members of the family Rhizobiaceae. Journal of
lipid research, Vol:35 pg:1452-1461, 1994.

105

Chapter 5
The Detection System
Introduction

The design of the detection system is based on magnetic sensing of a
bacterium that has been magnetically augmented by ferromagnetic
nanocomposite particles. Magnetic sensing is the keystone of our detection
platform. Magnetic sensors are used to control and analyze thousands of
functions, and can be found everywhere from cars to cell phones. There are
many types of magnetic sensors of varying form and function. Here we discuss
some different types of magnetic sensors and their limits of detection.

Giant magnetoresistance sensors are very useful and are used as
read/write heads in computer hard drives. Giant magnetoresistance sensors are
made up of a sandwiched four part electrode. There are two ferromagnetic layers
separated by a nonmagnetic conductor. The fourth part is an antiferromagnetic
layer on one of the ferromagnetic layers. The antiferromagnetic layer pins the
ferromagnetic layer, holding its magnetization in one direction. This
ferromagnetic layer is said to be hard, where the other layer is said to be soft.
The soft layer is susceptible to magnetization by external fields. The function of a
GMR sensor is as follows, if the magnetic fields of the ferromagnetic layers are in
a parallel configuration then the resistance does not change. If an external field
magnetizes the soft ferromagnetic layer antiparallel to the hard layer then the

resistance will increase. GMR sensors are also very sensitive with detection

106

limits in the hundreds of picotesla range [1] as well as having a high on chip
density of 0.1-1 million sensors per square centimeter [2].

Hall Effect sensors are based on the Hall Effect as the name implies. The
basic principle of the Hall Effect is that when a current is passed through the
length of a thin rectangular shaped conductor, and a magnetic field is applied
perpendicular to the plane of the sheet, a potential appears on the opposite
edges of the rectangle. This potential is proportional to the current flowing
through the conductor and the magnetic flux. Hall Effect sensors have been
produced with detection limits in the picotesla range (10 pT) [3]. Hall Effect
sensors are easily integrated on silicon chips making them inexpensive and easy
to use.

Fluxgate magnetometers are widely used in compass based navigation
systems because they can measure the intensity and orientation of the lines of
flux. Fluxgate magnetometers are made up of two parallel ferromagnetic cores
each wound with a primary coil, with one of them wound in reverse. A secondary
coil then surrounds the primary coil and is attached to an amp meter. The
primary coil is attached to an AC current. The induced field of the primary coils is
canceled out and there is no change measured in the secondary coil. When
exposed to an external magnetic field there will be some component in the
direction of one of the cores and a separate component opposing in the other.
This difference will cause a magnetic field that induces a voltage in the

secondary coil, allowing the magnetic field to be measured. Fluxgate

107

 

magnetometers have been made with detection limits in the picotesla range
(0.015 nT) [4].

Atomic magnetometers (also called optically pumped magnetometers) are
made up of alkali metal (Rubidium, or cesium) vapor cells, through which linear
polarized laser light passes and is detected by a photodiode. The linear polarized
laser light causes a polarization of the alkali metal vapor atoms, this polarization
is disrupted when the atoms are exposed to a magnetic field, resulting in a
modulation of the intensity of the transmitted light. The modulation or resonant
response of the atoms gives a measure of the Larmor frequency. The magnetic
field is then determined by the Larmor or precession frequency.

Smaller chip scale atomic magnetometers have been developed with picotesla
detection limits [5]. Atomic magnetometers are some of the smallest sensors (3
mm") with the lowest detection limits.

The search coil is one of the simplest and oldest magnet sensors and has
a broad dynamic range. Search coil sensors are based on Faraday’s Law of
Induction, which states that the voltage induced in a coil is proportional to the
change in the magnetic flux in the coil. Coil based sensors have sensitivity in the
hundreds of femtotesla range [6].

Frequency based detection is also useful; Fourier transform can be used
to pick up frequency dependent signals that are buried in the noise. Many of the
above sensors make use of this to filter out noise. The widespread use of
magnetic sensors has driven the development of improved and smaller sensors,

and one can only expect this trend to continue. We can see that there are many

108

magnetic sensors available with the sensitivities far below that which we would
expect to see from the bacterium coated with the F MNCPs. For this detection
system a search coil is used for its magnetic sensor, as the design was
discussed in chapter 2. The simplicity and sensitivity of search coils make them
an excellent chose.

Experimental

Spinning coil in the Earths field — The coil was mounted on a ruler which
was rotated about an axis parallel to a level plane. The coil was rotated by an
electric motor with a constant speed, located approximately 2 feet away. The
speed of the coil was varied by changing the radius of the rotating coil. The
voltage was recorded as the coil was spinning in the earths field for several
different radius.

Testing Magnetic Beads - A dilution of the stock magnetic beads
suspension was made (1 to 1000) using PBS. Magnetic beads used were 1 pm
BcMag amine-terminated magnetic beads made by BioClone Inc. From the
diluted stock suspension serial dilutions of 1 to 10 were made to make a total of
eight samples. Each sample dilution (5 pL) was loaded in the detection platform,
starting with the lowest and going to the highest. Measurements were taken over
a 5 min period. In between samples, 10 mL of PBS were pumped through the
system to wash it out.

Testing Nanocomposite Material - A turbid suspension of Rhizobium trifolii
was made using a PBS solution. Using the suspension, 1 to 10 serial dilutions

were made for a total of 6 sample. The Rhizobium trifolii were grown up on

109

Modified Bergensen’s medium (Blll) agar plates at 28 °C for 5 days. Aliquots
(100 pL) from each of the last three dilutions were then plated on Blll plates and
were later counted to determine the number of viable cells in the samples. The
nanocomposite suspension was then loaded into the detection platform. The
Rhizobium tn'folii suspension samples (4 pL) were then loaded in to the detection
platform starting from the lowest concentration to the highest. Measurements
were taken for 5 minutes before the next sample was added. The control
experiment was done fallowing the same procedures above with Escherichia coli
used in place of Rhizobium trifolii.

Control experiment one - An experiment was run using different water
samples, tap water, rusty tap water, milliQ water, deionized water, as well as
some sodium chloride solutions of different ionic strengths (0.001, 0.01, 0.1, 0.5,
1 M). For this control the speed of the pump started at 0 then 5 then 10 and then
back down to 5 then 0, the speed was changed every 5 min. The same pump
profile was used when the pump was run in reverse.

Control experiment two - An experiment was performed to measure the
voltage of the nanocomposite suspension at 3 different dilutions. The
nanocomposite suspension was prepared as stated in chapter 4, and 2 dilutions
were made to give three samples (1 , 1/2, 1/10). The voltage was measured for
five minutes with the different nanocomposite suspensions at speed 5 and 10.

Control experiment Three — An experiment was run to measure the

voltage induced by the different concentrations of Rhizobium Trifolii cells.

110

 

Different concentrations of cell suspensions were made the same as in previous
experiment, and samples were analyzes using the same parameters as well.
Results and Discussion

The fabrication of the device consisted mainly of acquiring the individual
components. The pump used was a Buchler instruments peristaltic pump. The
voltmeter used was a Keithley 2182 nanovoltmeter. The meter was controlled
and measurements were recorded using a Labview program. The coil had to be
fabricated in house, consisting of a capillary tube with 3000 turns. The coil was
placed into a Faraday cage, leads were cleaned and connected, as well as all the
tubing, and all was assembled for testing.

Knowing that the earth has a magnetic field of 5x105 T, and that
magnetotactic bacteria produce fields 20 to 30 times grater then this. And that a
bacterium coated with the FMNCPs has fields equal to or greater then those of
the magnetotactic bacteria. We can see that if our coil can detect the earths field
and it moves through it, then the coil can detect the FMNCP coated bacterium. In
order to see what affect the earth’s magnetic field had on the coil, the voltage
was measured as it cut the earth’s magnetic field at different velocities. Figure
5.1 shows the induced voltages of mechanically spinning the coil in the earth’s
field at different speeds, along with theoretical values. Looking at the trend lines
fitted to the data we see that the slope of the mechanically spun data is less then
that of the theoretical data, as well as a large y intercept value of about 800 W.
This large value can be attributed to the noise produced by the electric motor

used to spin the coil. The coil was also spun by hand as well and fitted with

111

8.0E-06

 

7.0E-06 -

 

 

6.0E-06 -

 

5.0E-06 -

 

4.0E-06 -

 

Voltage (V)

 

3.0E-06 —

 

2.0E-06 ~

1.0E-06 -

 

 

 

 

 

 

 

 

 

0.0E+00

 

0 5 10 15 20 25 30 35 40 45 50 55 60
Velocity (cm/s)

Figure 5.1. Graph of Voltage vs. Velocity. The pink squares are the voltages
induced in the coil while spun in the earth’s field at different velocities, and the
orange line is a line of best fit to the data. The green line is the line of best fit for
the hand spun data. The blue line is the expected theoretical induced voltage in
the coil by the earth’s field at different velocities. the vertical light blue line
represents the velocity of the nanocomposite particles through the cell.

112

 

a line. The line for the hand spun data has a slope that is close to that of the
theoretical, which is good considering the conditions (unshielded) with a intercept
of 121 W. While we may think that this is still high, it can be accounted for if we
think about what we know about the earth’s core. The rotation of the inner core is
3 to 2 ° per year faster than that of the earths crust [7, 8]. This works out to be
about half a mile per day or 1.06 cm/s. Using this information we have
determined that a voltage of 125 nV can be induced in our coil as the result of the
earth’s main field moving with the core relative to the crust. Actual measurements
of the average voltage when the coil was stationary were about 122 nV. This puts
our experimental and theoretical data are in agreement.

Once the coil was shown to be capable of detecting the earth’s
field, preliminary testing of the detection system was done using 1 pm
superparamagnetic magnetite beads with from BioClone Inc. The magnetic
beads were made from dilutions of the bead suspension were added to the
platform in increasing concentrations and the voltage was measured. The voltage
increases with the first three samples to about 50 nV, and fluctuates with the last
four from about 40 to 73 W. The data is graphed in figure 5.2 and shows
average corrected voltage induced by the different magnetic bead samples of the
5 minute analysis time.

The BioClone magnetic beads used were 1 micron in diameter, similar in
physical size to our bacteria cells and in magnetic field. The idea was that if the
beads could induce a detectable change in the voltage then we would expect to

be able to

113

 

1 .40E-07

 

1.20E-07 .

WW _ .. ..
|
I

 

1.00E-O7~ . - .. " W. .

 

 

3.002034“ —« »~ «4— _- _ ._. -.

 

 

5.00503 ~-— " 4 . - .. . WWW“ WW

 

4.00503 4e— W -WW 4. ...... WW __

 

VON-90 (V)

2.00503 ~» --~——-—»—— — . ..__. .__--_

 

 

 

 

 

0.00E+00 4 — - - 4~ --;;——.— - .. . W.

 

 

 

 

-2.00E-08 ————— — . «—. -sfi. 9*
{I g
4.00500 ~ 7 .. _ e9 44 . W. W _ -W_ . . _.

l
I
l
I
I

 

 

 

 

 

 

 

 

 

-6.00E-08
1 1 0 1 00 1 000 1 0000 100000 1000000 1 0000000

#OfBeads

Figure 5.2 The blue diamonds represent the average corrected voltage vs. the
number of magnetic beads.

114

detect a bacterium coated with nanocomposite. We can see from the results
(figure 5.2) that as the number of beads in the detection platform increases so
does the voltage. It must also be noted that analysis of the raw data and visual
observations showed that the magnetic beads quickly became dispersed
throughout the circulating solution. This is seen in the data that the voltage
increases with more beads and maintains an average voltage and not a
distinctive periodic voltage jump when the beads pass through the coil. The
dispersion of the beads has an averaging effect on the voltage induced because
the beads become evenly distributed throughout the solution resulting in an
average magnetic flux passing through the coil. The dispersion of particles
throughout the circulation solution seen with the magnetic beads can also be
seen with the nanocomposite suspension and Rhizobium trifolii data. The data
shows an averaging of the induced voltage as well.

With the preliminary tests showing results it was time to test the
nanocomposite and Rhizobium trifolii. The nanocomposite was loaded into the
detection platform to which samples of Rhizobium trifolii were added with
different concentrations. Figure 5.3 shows the voltage measured over the 5
minute analysis time for one of the samples (raw data). Figure 5.4 shows a
graph of all of the samples with some line smoothing applied to the data. In
Figure 5.2 we can see the graph of the average corrected voltage and number of
cells in the system. The voltage appears to increases logarithmically as a
function of the number of cells. As a control the nanocomposite was tested using

Escherichia coli instead of Rhizobium trifolii. Again the nanocomposite was

115

3.0E-07

2.5E-07

2.0E-07

1.5E-O7

1.0E-O7

Voltage (V)

5.0E-08

0.0E+00

-5 .OE-08

 

—1.0E-07

0
Q
~ e

CFUs/mL

Figure 5.3 Graph of Average corrected voltage vs. the number of CFUs/mL. The
orange squares represent the nanocomposite and Rhizobium trifolii data. And the
yellow triangles represent the nanocomposite and Ecoli cell data.

116

 

8.00E-07 ,

 

 

l i ‘
zoos—07 ‘9 a i] l
l I
l l

6.00E-07

 

 

5.00E-07 ~

4.00E-07 -9

 

Voltage (V)

 

 

soars-or 4. ! all. (-

2.00E-07 -

 

1.00E-07 1

 

 

 

 

 

l
o.oors+oo i i .

Time (S)

Figure 5.4 Raw data of voltage induced over the 5 minute analysis time from the
7.89x106 CFU sample (raw data).

117

Voltage (V)

 

CFUs/mL

Figure 5.5 Graph of all of the nanocomposite and Rhizobium trifolii data samples
with line smoothing applied to the data.

118

loaded into the detection platform and then cell suspensions of different
concentrations were added to the detection platform and the voltage measured.
The data for the average corrected voltage for the nanocomposite and
Escherichia coli cells is plotted on Figure 5.3. The voltages are less then those of
the Rhizobium tn'folii and nanocomposite, and fluctuate as the number of cells
increases.

The nanocomposite and Rhizobium trifolii experiments results showed a
measurable induced voltage as the number of Rhizobium trifolii increased. The
increase in voltage is not proportional to the increase in the number of cells, and
the voltage levels off (figure 5.6). This means there is an upper limit of detection.
This is due to the fact that there are a limited number of NC particles in the
suspension to bind to the target. Once all of the NC particles have been used up
(bound to target) addition of more cells will not increase the voltage.

As mentioned before, the core of the nanocomposite is
superparamagnetic. How is it possible that they are not only able to induce a
voltage but a positive voltage in the experiments? In order to explain the
observed phenomena we have to take another look at the magnetic energies
involved, specifically the Zeeman energy. The Zeeman energy is dependent on
the magnetic moment, the external magnetic field and the angle between them.
In our case the external magnetic field is the Earth’s magnetic field. While we
used a Faraday cage to remove noise from electromagnetic radiation, the earth’s

magnetic field is not removed.

119

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2.00E-07 . ‘ W
1.80E-07~~ ~ — ~———- “4 —.W (W W W W W. W. W WW1:
1.soe-o7~— 24¢— fim -4 W_ - ._.W W WWW _W_W_(_W_ W_
/
1.40E-07“ 7777*7——-WWWWWWWW
1.205-07 v a 9 ~ 7 v e fie ~ 4.. a ~- a
E I
O
a - W W W W W W WW W W .W W W W W W W
g 1.00E-07
O
> 3005-03 e t e fi—v ~e+ i——--e—L— Wi—W—4— ——
6.00E-08A4-W *7 WW, . 7 i »— T] 4 4 — —W—-W W W W_ W W W- ..W
4-OOE‘08 *T "A i ) _‘ i 7 H fi fi ‘— _ — a — — r7 —— ——T ——
2005-081— * E» i- a 4~—-e»~-—-~ r;---—- e W W W. W W W W W- -W .W W
0.00E+00 r
Qfi° (0.59 (0445‘ (045‘ «’9‘ .8“ «9" (323‘ 623‘
e° N9 «(,9 25° 049 62° 62° «9 414°

CFUslmL

Figure 5.6 Average corrected Induced Voltages by the nanocomposite and
Rhizobium trifolii.

120

 

Knowing the core particle size we can now calculate the Zeeman energy
of a single particle in the field and use the Langevin function to determine its
average orientation. The Zeeman energy of a single nanocomposite particle in
the earth’s field is 1.125x10'24 J, much less then the themal energies 4.14x10'21
J. While a single nanocomposite particle does not have enough energy to align
with the earth’s field, a large enough assembly of them does. If we use the
Langevin function and look at the average angle as a function of the number of
nanocomposites we see that at about 5000 particles the Zeeman energy is
greater then the thermal (Figure 5.7). With large assemblies of the
nanocomposite on the bacteria surface in the range of 15,000 we can see that
there is sufficient energy for interaction. The interaction of the nanocomposites in
a large assembly and their interaction with the Earth’s magnetic field not only
allows them to overcome the thermal energy and behave as a large
ferromagnetic particle with a large magnetic moment, but also aligns it
with the earth’s magnetic field, which causes them to pass through the coil in the
same orientation. And because of this permanent orientation of the magnetic field
the voltage induced is positive. One question that you may be wondering by now
is that why isn’t the average voltage zero. As we would have expected from
looking at the theoretical calculation where the voltage goes through a positive
and then a negative maxima which would lead to a average voltage of zero, and
yet we can see from the results that there is a net voltage. What this tells us is
that there is difference in the voltage induced by the bacteria when it inters the

coil and when it leaves the coil. So there must be some change in the direction of

121

0.9 -

0.8 ~

0.7 -

0.6 -

<Cos 8>

0.4 -

0.3 -

0.2 -

0.1 —

 

0.5 -

 

l

I

t

.c

I

l

l
I
l
i l
. . L W
l

5.5E-1 3

-~ 4.5E-13

1 3.5543

4 A— 2.55.13

Energy (are)

‘ ‘ 0 1.5E-13

:2 5E-14

 

I
I
l.
l .4
i l

l l l W -55...

l

 

5000 10000 15000 20000 25000 30000 35000 40000 45000 50000

Number of 3nm particles

Figure 5.7 Graph of the average angle of alignment of the nanoparticles in the
earth’s field as a function of the number of 3 nm particles (blue line). The
Zeeman energy of the nanoparticles in the earth’s field as a function of the
number of 3 nm particles (pink line) and the Thermal energy (Yellow line) at room
temperature.

122

the magnetic field as the bacterium moves through the coil. The Lenz’s Law
states that the path of a moving charge particle is changed in any way by its
interaction with a magnetic field, the change is always of such a nature as to
generate a new magnetic field which directly opposes the one which caused the
change [9]. This means that a voltage induced by the magnetic bacterium moving
through the coil generates a magnetic field that apposes the field of the magnetic
bacterium. The apposing Lenz field acts to push against a magnet entering a coil
and pulls it back exiting the coil. If the Lenz field is strong enough one can see
how it can change the direction of the magnetic field. As the magnetic bacterium
enters the coil it induces a voltage that generates an opposing field. This
opposing field in not felt right away for there is some lag. The opposing field
reaches a maximum after the magnetic bacterium has passed half way through
the coil. The Lenz field causes a reorientation or flipping of the magnetic
bacterium as it is heading out of the coil, which maintains the positive sign of the
voltage. Using the voltage induced by the magnetic bacterium, the Lenz field
produced in the coil is equivalent to that of earth’s field. This is strong enough for
the above interactions and explains the net average voltage. The strength of the
Lenz field is equal to that of the field of the FMNCP coated bacterium. The fields
produced by the samples are listed in table 5.1. The magnetic fields are the
same order of magnitude to that of the earth’s field which we know from earlier
calculation and as a result rotates the magnetized bacteria.

Escherichia coli were used to test for nonspecific binding interactions of

nanocomposite with non-target bacteria. The data is graphed in figure 5.2 along

123

 

 

Table 5.1 Voltages and magnetic fields generated by the FMNCP coated
bacterium.

 

 

 

 

 

 

 

 

 

 

Sample Average Voltage (V) Magnetic Field (T)

nc6 3.20615E-07 4.25908E-05
nc5 3.47608E-07 4.61766E-05
nc4 3.71457E-07 4.93448E-05
nc3 3.94712E-07 5.24339E-05
nc2 4.0903E-07 5.43359E-05
nc1 4.47905E-07 5.95001 E-05

 

124

 

with the other data. The graph shows us that the voltage fluctuates up and down
with increasing concentrations of cells, with an overall small change in voltage.
The small change in voltage indicates that there is minimal non-specific binding.

Table 5.2 gives the ionic strengths of the different solutions along with the
average voltage induced over the 5 minute analysis time. The Control experiment
one Results were similar for all speeds tested fonlvard and reverse. Meaning that
running the pump backwards does not change the voltage measured. We can
see that there is no correlation between the Average voltage and the ionic
strength of the solution with the data being scattered. While the voltages are high
when compared to that seen in the nanocomposite and Rhizobium trifolii
experiments it must be noted that this data has no way of correcting for
measurement taken on different days for comparison.

Control experiment two tested to see if the voltage induced changed with
the concentration of the nanocomposite suspensions, as well as with the speed
of the suspension passing through the coil. Figure 5.8 shows the average
voltages of the different nanocomposite suspensions at speed 5 and 10. As you
can see the voltage induced, increases as the speed is increased as one would
expect, and the average voltage increases minimally with increasing
concentration.

In control experiment Three an experiment was run to measure the
voltage induced by the Rhizobium Trifolii cell by then selves to insure that the
voltage induced by the nanocomposite and Rhizobium trifolii experiments was

not just a result of adding Rhizobium trifolii. Figure 5.9, 5.10 shows the graph of

125

 

Table 5.2 Control samples conductivity and average voltage with standard

 

 

 

 

 

 

 

 

 

 

 

 

deviation.

Sample Conductivity (1/0) Average Voltage (V) STDEV

Rusty Tap water 808 1.79093E-07 4.22983E-08
Tap water 898 1.73363 E-07 5.4657E-08
D.I Tap water 5.2 1.784E-07 3.02854E-08
MilliQ water 0.4 1.20468 E-07 3.08291 E-08
PBS 14620 2.68714E-07 5.99821 E-08
NaCl 1M 82100 1.22225E-07 8.52331E-08
NaCl 0.5M 55000 1.85987E-07 8.86639E-08
NaCl 0.1M 17600 1 .89771 E-07 9.89374E-08
NaCl 0.01M 1230 1 .52554E-07 9.90029E-08
NaCl 0.001 M 144 1 .28248E-07 7.54727E-08

 

 

 

 

126

 

 

3005-07 . 1 ,

 

2.50E-07

 

 

} .

 

1.50E-07

Voltage

 

1.00E-07

 

5.00E-08

 

 

 

 

 

 

 

 

0.00E+00 4 4
0 0.2 0.4 0.6 0.8 1

Relative concentration

Figure 5.8. Average voltages of the different nanocomposite suspensions at
speeds 5 (pink) and 10 (Blue).

127

1.2

 

1.00E-07

8.00E-08

6.00E-08

 

4.00E—08

Voltage (V)

2.00E-08

0.00E-r-00

-2.00E-08

 

-4.00E-08
1.00E+03 1 .00E+04 1.00E+05 1.00E+06 1 .00E+07 1.00E+08 1.00E+09

CFUslmL

Figure 5.9 Graph of the average corrected voltage vs. the number of cells for
Rhizobium Trifolii cell. Data from measurements taken from high concentration of
cells to low concentration are in red the from low to high are in blue.

128

 

ZOE-07

 

 

l
1.8E-07 : 4

 

«4— —. .4W— -

1.6E—07 ' e

 

1.4507 4;

 

 

1.2E-07

 

.__WJWWW y—W

 

1 .0E-07

 

 

Voltage (V)

 

 

8.0E-08 l ’

6.0E-08 .4

 

 

 

4.0E-08

 

 

2.0E-08

 

 

 

W W- .W .. W. W. .WW. WW

 

 

 

I’

l

I
0.0E+00 I

 

o o o o o o
No 00 0° 000 oo 090

CFUslmL

Figure 5.10 Graph of all the corrected voltage data; The orange diamonds show
nanocomposite Rhizobium trifolii data, the yellow triangles are the
nanocomposite and Ecoli cells data. The red circles are the High to low data and
the purple squares are the low to High data.

 

129

the average corrected voltage vs. the number of cells. The test was done twice
once from highest to lowest concentration and once from low to high. While the
voltage increases with concentration in the low to high test in the first 3 samples,
it fluctuates in the last 2 as does the high to low over the whole range. While the
voltages induced are comparable to the voltages of the lower nanocomposite and
Rhizobium trifolii experiment, the data is vary scattered. The voltage induced in
the experiments is due to the nanocomposite and Rhizobium trifolii interaction
and not the Rhizobium trifolii.
Conclusion

I have been able to design, fabricate, and test our detection system and
have shown that our detection platform is capable of a real time detection of
bacteria (within five minutes) and low concentrations of cells. And most definitely
has the potential for single cells detection. I have demonstrating that
Ferromagnetic nanocomposite particles can be used for the detection of bacterial
cell by magnetic induction with a search coil. By design our detection platform is
capable of a 3 star rating for all of the detection platform features, speed,
accuracy, sensitivity, portability, robustness, and ease-of-use. This detection
platform has been shown to have the elements of speed, accuracy, and
sensitivity. The other elements of portability, robustness, and ease-of-use are
areas of improvement and testing in future prototypes. As mentioned before the
true challenge for all detection platforms is speed. The time from the moment a
sample is taken to the point of detection is crucial. Accuracy is also a challenge

that must be balanced with speed. We must keep in mind that there will always

130

 

be trade offs between speed and accuracy. We have shown that our detection
platforms can perform at the desired speed, and are quite capable of embodying
all of the features needed. By no means is our system perfect, nor would we
expect to make one. In our system we seek a balance between certainty and
time. With infinite time one could know with absolute certainty but time is not
infinite and decisions must be made immediately and actions taken.

It is because of this urgency that our idea has evolved beyond a detection

 

platform to a global monitoring system of the origin and progression of disease.
In order to create such a system, improvements upon our design must be made.
Work has already begun on designing the next prototype, which we will describe
here. The detection platform will again make use of Faraday’s law of induction,
but will not require the need of a pump to circulate the suspension; this greatly
reduces the size and possible problems. The basic form and function of the
detection platform is as follows. A bacterium enters the detector where it
becomes encapsulated with FMNCP and is captured on a small rotating disc. As
the FMNCP-encapsulated bacterium rotates on the disc it passes a coil that is
located just above the disc. This coil will be able to detect the changes in the
magnetic field as the disc rotates. This design allows us to greatly increase our
sensitivity in two ways. One is that the speed of the disc’s rotation can be
controlled to increase the signal. Second the immobilized bacterium will rotate
with the disc at a set period, allowing for frequency based detection. Fourier
transform can be used to pick up the frequency dependent signals that are buried

in the noise. The disc will be disposable, made for one day use, coming as easy

131

to change cartridges. The used cartridge is removed and discarded and the
platform is flushed with sterile water. The new disc cartridge is inserted, then a
volume of sterile water is added to rehydrate the disc resuspending and
desolving FMNCP, buffer, etc. Circulation of the suspension in the detection
platform will be accomplished by convection generated by the rotating disc (this
may require some type of structures like grooves, holes,or fins on the disc in
order to make it flow accordingly). We have several ideas for the design of the
disc. One idea is to have a uniform film of FMNCPs suspended or immobilized
in/on a membrane. Basically the FMNCPs are stuck on the disc but have enough
freedom and are in large enough concentration that when a bacterium is
captured on the disc by FMNCP it will become bound by all of the surrounding
FMNCPs. The thought is that a uniform thin film of FMNCPs can be scanned by
the coil and give a uniform signal. The binding of the bacterium will cause a
discontinuity in the film with an area of no FMNCPs surrounding the bacterium
and a large concentration of them at the center. This discontinuity in the field is
then picked up by the coil as it passes over the bacterium.

Another design is one that uses SRMs immobilized on the surface of the
disc and the FMNCPs are free in suspension. The bacterium picks up FMNCP as
it circulates and becomes captured on the disc. The captured bacterium picks up
more FMNCPs as it rotates. The build up of FMNCPs on the disc leads to the
induction in the coil and detection. We have thought of ways for several
pathogens to be detected at the same time by the use of multiple rings on the

disc each with different SRMs for their specific target pathogen. While we have

132

chosen to use coils as our detector in our platform, several different magnetic
sensors could be substituted, such as atomic magnetometers, Hall sensors, and
GMR to name a few.

The proliferation of personal electronics such as cell phones is apparent
all over the world. Integration of our detection platform in to these electronic
devices would allow for a broad distribution of detectors throughout a community,
whether it be rural, suburban, metropolis, third world or first. A distribution of
detectors such as this could supply a level of detection coverage unattainable by
any other method.

There are many possible future uses of this technology that could go
beyond pathogen detection. The ability to modify an organism with
nanocomposite materials has great potential. Not only do we have the ability to
genetically modify an organism but, now we can also physically modify it. The
nanocomposite material will impart optical and magnetic properties of the
nanoparticle to the organism. Incorporation of nanocomposite materials can also
go beyond surface modification. Proper coating of nanocomposites and
treatment of organisms would allow for internalization of nanocomposite
materials. The ability to magnetically modify an organism has many possibilities.

The magnetic modification of an organism allows us a great deal of control
over the organism’s positioning, orientation, spacing and interaction with others.
One can imagine magnetic positioning of cells on surfaces in specific patterns.
These cells could produce specific biomolecules such as proteins, lipids,

carbohydrates, etc. The precise control over the positioning of the cells would

133

 

allow for fabrication of biomaterials with ordering and patterns unparallel to
anything found in nature. This would allow for the production of biomaterials that
not only are vastly superior to current materials, but have properties that we have
only dreamed of. Another use could be for cellular reconstruction/deconstruction
in damaged/unhealthy tissue. One could also imagine magnetically modified
organisms used for power generation. Magnetically modified organisms could be
made to swim through small coils to generate an EMF to power small electronics.
The physical modification of cells brings us closer to closing the gap
between inorganic and bioorganic technologies. While our inorganic technology
has advanced much faster then our bioorganic technology, the latter is slowly
gaining momentum. We can envision a future where the bioorganic and inorganic
technologies merge and form a seamless interface. Science can be simple and
easy or complex and difficult, but if we are to make something that will impact the
world and help mankind we must be able to put it in the hands of the people of
the world in a way that is simple and convenient. We have shown that the
modification of a bacterium with ferromagnetic nanocomposite material can be
used to generate a signal allowing for detection of a single cell. We have also
suggested many improvements and ideas for future devices and applications,

which we hope to see come to fruition.

134

REFERENCES

. Low-frequency noise measurements on commercial Magnetoresistive
magnetic field sensors. Stutzke, N.A. Russek, S.E. Pappas, D.P. Journal
of Applied Physics. Vol:97, 2005.

. Advances in giant magnetoresistance biosensors with magnetic
nanoparticle Tags: review and outlook. Wang, S.X. Li, G. IEEE
Transactions on Magnetics, Vol 44, No 7, pg 1687-1702, 2008.

. P. Leroy, C. Voillot, V. Mosser, A. Roux, G. Chanteur, An ac/dc
magnetometer for space missions: Improvement of a hall sensor by the
magnetic flux concentration of magnetic core of a searchcoil. Sensors and
Achtuators A-Physical Vol:142, 2 pg 504-510, 2007.

. M.H. Acuna, D. Curtis, J.L. Scheifele, et al. The STEREO/IMPACT
magnetic Field Experiment. Space Science Reviews Vol:136 lssue:1-
4 pg:203-226, 2008.

. P.D.D. Schwindt, B. Lindseth, S. Knappe, V. Shah, J. Kitching, Chip-scale
atomic magnetometer with improved sensitivity by use of the Mx
technique. Applied Physics Letters, Vol:90, 2007.

. A. Edelstein, Advances in magnetometry. Journal of physics condensed
matter, Vol:19 lssue:16, 2007.

. X. Song and P. Richards, Seismological evidence for differential rotation

of the Earth's inner core. Nature, Vol:382 pg:221-224, 1996.

. G.A. Glatzmaier and RH. Roberts, "Rotation and magnetism of Earth's
inner core," Science, Vol:274 pg:887-1891, 1996.

. H. E. Burke, Handbook of Magnetic Phenomena, Van Nostrand Reinhold
Company Inc, 1986.

135

 

Appendix A

Magnetism
Introduction

Magnetism is a phenomenon that man has tried to understand and
manipulate for thousands of years, with the earliest discovery credited to the
Chinese in 4000 BC. [1]. The naturally occurring iron ore magnetite, also known
as Iodestone, was of great interest because of its tendency to attract steel and
iron. Beginning with the first compasses that changed navigation to electric
dynamos and motors, and today’s advanced materials we can see throughout
history that as our understanding of magnetism has grown so has our technology
which has allowed for greater discoveries.

The earth’s magnetic field can in most cases be thought of as a bar
magnet with the magnetic field being the strongest at the poles with field lines
emerging from the north pole (geographic south pole) and reentering at the south
pole (geographic north pole). The earth’s magnetic field forms a toroid around the
earth creating the magnetosphere. The earth’s magnetosphere protects the earth
against solar winds and radiation. Even so the earth’s magnetic field is relatively
weak at the surface about 5x10'5 Tesla, this is a few hundred times weaker than
an average refrigerator magnet. While our simple bar magnet model works nicely
to visualize the earth’s magnetic field at the surface and in the magnetosphere it
does not lend any insight as to its origin. The earth is made up of 4 layers,
starting with the crust on the outside then the mantle, followed by a liquid outer

core and a solid inner core. The cores are made up mainly of iron. But it is not

136

 

the permanent magnetization of this iron that gives rise to the earth’s magnetic
field, for the temperatures at the core are far above that of the Curie temperature
of iron, making permanent magnetization impossible. So what is the origin of the
earth’s magnetic field? Many theories have been proposed, but all have fallen
short except for the Geodynamo theory which was first suggested by Larmor in

1919 [2]. Geodynamo Theory, simply put, says that the earth’s magnetic field is

 

the result of electrical currents in the liquid outer core. The electrical currents are

generated by movement of core fluid by convection and rotation [3,4].

 

We have all played with bar magnets at one time or another and noticed
their interesting interactions with different materials and other magnets. Magnets
are always made up of a “north and a south pole”, more properly called a
magnetic dipole. The magnetic field of a magnet can be visualized as coming out
of the north pole and bending back down and around to enter at the south pole in
a toroidal shape. This basic picture of a magnetic dipole holds true for all
magnets regardless of size (Figure A1). This is not to say that the magnetic

properties of a nanomagnet scale proportionally to those of a large bar magnet of

 

the same material, only that the resulting magnetic dipoles look and behave like
those of all magnets.

So what is the origin of the resultant magnetic dipole in our picture? The
origin of the resultant magnetic dipole in a magnet is the product of individual
point magnetic dipoles that are aligned parallel with one another. We can now
visualize our magnet as a collection of point magnetic dipoles (Figure A2) that

are the result of unpaired electrons within the material. The cause of magnetism

137

 

—)z

 

 

 

 

Figure A1 The toroid shaped magnetic fields of magnetic dipoles of different
sizes. From left to right a bar magnet then a nanoparticle and last an electron.

138

    

Spin ‘

Point Magnetic Dipole
(unpaired electron)

 

Ferromagnetic Material

Figure A2 The cooperative interaction of the point magnetic dipoles in
ferromagnetic materials results in a net magnetic moment.

139

 

 

is the movement of charged particles; whenever a charged particle moves, it
creates a magnetic field. For example when a current is passed though a wire
(moving electrons) it creates a magnetic field perpendicular to the direction of the
movement of the current by the logic of the right-hand-rule. The relationship
between a moving charge and a magnetic field is stated by Ampere’s law:
dH = ki(dI/r2)sin0

Where dH is the resultant magnetic field of the moving charge, k is a
constant, i is the current that travels along the path dl, r is the distance from the
path to field of interest, and 0 is the angle between direction of the moving
charge and the field. The movement of the charged particle can be rotational as
well as translational; it is the rotation or spinning of the electrons that generates
the magnetic field. Electrons can have one of two possible spins, that is spin up
or spin down which is assigned a value of + or - Z2, respectively. I mention this
because electrons spins pair in opposite directions, which effectively cancels out
their magnetic fields; this is why only unpaired electrons can contribute to the
magnetic field of the magnet with the exception diamagnetic response. While the
orbital motion of the electron and nuclear spin can contribute to the magnetic
field, they are negligible compared to the magnetic moments of the electron spin.
The unpaired electrons act as point magnetic dipoles within a material. One
might assume that all materials with unpaired electrons are ferromagnetic, when
in fact most are not. While our simple model of a magnet as an assembly of
parallel point magnetic dipoles that sum to make one larger magnet dipole is still

true, there are microscopic and atomic interactions and assemblies of magnetic

140

 

 

dipoles that affect the net magnetic moment of the magnet that causes not all
dipoles to be aligned cooperatively.
Magnetic Energies

There are several energies that contribute to the magnetism of a material,
and the degree of their contributions results in the magnetic properties of a
material. These energies are exchange, magnetostatic, anisotropy, and Zeeman
energies [5]. An important energy that we must not neglect is the thermal energy
(Ethem, = kT), which sets the scale that all other energies are compared to. As
the temperature of the material is increased, its effects dominate the others. The
exchange energy is the result of quantum mechanical exchange interactions that
cause neighboring dipoles to align parallel or antiparallel. The exchange
interaction acts to orientate many point magnetic dipoles into regular patterns.
Magnetostatic energy is due to the interactions of the magnetic dipoles within the
material. While the exchange interactions dominate nearest neighbor
interactions, the long range magnetostatic interactions minimize energy by
domain formation. Anisotropy energy arises from the crystal structure and shape
of the material, and causes the point magnetic dipoles to favor alignment along
specific crystallographic axis. The Zeeman effect causes splitting of orbital spin
energy levels when exposed to an magnetic field.

Polycrystalline ferromagnetic materials are made up of several magnetic
domains. These domains can range in size from a few tens of nanometers to
several microns. As mentioned above, domain formation is driven by

magnetostatic interactions as well as exchange and anisotropy. Magnetostatic

141

I.
A

 

d

(dipole—dipole interactions) energy is lowered by the formation of magnetic flux
loops or closures (Figure A3). This lowering of energy is less than the cost to
form domain walls. These domain closures provide a path for the magnetic flux
within the solid. It is the formation of these domains that causes ferromagnetic
materials to have net magnetic moments that are small or zero before
magnetization. The domain walls are transition layers between domains in which
the spins gradually change direction over many atoms (figure A4). This gradual

change is favored over one distinct jump that would leave a discontinuity, as a

 

gradual change of spins lowers the exchange energy by distributing it over many
spins. While the magnetization of a material is the result of the summation of its
point magnetic dipoles, their cooperative interactions (parallel alignment) is due
to an entirely different force. That is, the magnetostatic energy (dipole-dipole
interactions) is too small to be responsible for the cooperative interactions when

compared with the thermal energy.

 

The exchange energy is responsible for the parallel and antiparallel orientation of
neighboring electrons; in ferromagnetic material parallel configurations are lower
in energy than antiparallel.

The size of these interactions results in different magnetic properties. A
material can be classified as diamagnetic, paramagnetic, and ferromagnetic. The
bulk magnetic properties of a material are best visualized using magnetization
curves, also called hysteresis loops (Figure A5). A hysteresis loop measures the
magnetization of a material in response to an external magnetic field. Then how

do point magnetic dipoles within the material react in the presence and absence

142

3*

 

 

a
—> z
(—

 

 

 

 

 

U!
Z
to
2

 

 

ll

(.—
—>

 

 

 

NSNS

 

T

l

 

t

W

Figure A3 Domain structure of a polycrystalline ferromagnetic material

143

 

int/742799.022 l
llllttewllllll

Bloch wa all
classified as Neel

 

 

Figure A5 Here is an example of a hysteresis loop where the x axis is the applied
magnetic field and y axis is the magnetization of the material. A in the figure
marks the saturation magnetization when all of the domains are aligned with the
field. Point B is the remanence or retained magnetization when the applied field
goes to zero. C is the coercivity which is the field that must be applied to reverse
the magnetization to zero. D is the same as A in the opposite direction. E is the
same as B in the opposite direction, as are F and C.

145

of an applied magnetic field. Diamagnetic materials have no net magnetic
moment per atom, and in the presences of an external magnetic field the point
magnetic dipoles align in opposition to the applied field. The orientation of
magnetic dipoles in a paramagnetic material is random in the absence of an
external magnetic field, because thermal energy causes jostling of the dipoles
and because they have no preferred relationship among them selves. The
random orientation of the dipoles results in no net magnetic moment. When an

external field is applied to a paramagnetic material the energy of the applied field

 

is strong enough to overcome the thermal energy and the magnetic dipoles align
themselves in concurrence with the applied external field. Paramagnetic
materials not only align in agreement to the applied magnetic field, they add to
the magnetic field with a magnetization that is proportional to the applied field.

Ferromagnetic materials, once magnetized, retain a net magnetic moment in the

 

absence of an external magnetic field, and in the presence of an external
magnetic field, the degree of magnetization is a factor of the applied field. The
magnetization that is retained by a ferromagnet after the magnetizing field
diminishes to zero is referred to as its remanence. The magnetic field that is
required to return the magnetization of a ferromagnetic material to zero is called
the coercivity. The coercivity is related to the anisotropic energy of a
ferromagnetic material. The anisotropic energy causes the magnetization of
ferromagnetic material to align along crystallographic axis, which are called
directions of easy magnetization. Ferromagnetic materials with low coercivities

are considered soft and those with high coercivities are called hard. The term

146

susceptibility is also used to describe magnetic materials, that is diamagnetic
materials have negative susceptibilities, paramagnetic materials have low
susceptibilities and ferromagnetic materials have high susceptibilities. The
susceptibility is given by the following equation:

x = M/H

Where X in the susceptibility and M is the magnetization and H is the
applied magnetic field. As mentioned above, paramagnetism is the result of
thermal agitation, which causes the random alignment of the magnetic dipoles
resulting in a net magnetic moment of zero. Ferromagnetic materials have a
temperature at which the thermal energy is greater than the exchange and the
material becomes paramagnetic: this temperature is called the Curie
temperature. The susceptibility of a material as a function of temperature is given
by the Curie Weiss Law in the following equation:

x = C/ (T-Tc)

Where x is the susceptibility, T is the absolute temperature, Tc is the Curie
point temperature and C is the Curie constant of the material. As we can see
there are various energies that are at odds with each other in determining the
magnetism of a material, and one of the largest effects is due to thermal

agitation.

147

REFERENCES

. D.C. Mattis, The theory of magnetism made simple. World Scientific
Publishing Company, (2006).

. P. H. Roberts G. A. Glatzmaier, Geodynamo theory and simulations
Reviews of Modern Physics, Vol:72, No. 4, pg:1081-1123 (2000).

. B. A. Buffett, Earth’s Core and the Geodynamo. SCIENCE Vol:288, No.16
pg 2007-2012 (2000).

. J.A. Jacobs, Reversals of the earth’s magnetic field second edition,
Cambridge University Press (1994)

. S. A. Majetich, M. Sachan, Magnetostatic interactions in magnetic

nanoparticle assemblies: energy, time and length scales. J. Phys. D: Appl.
Phys. 39 pg:R407-R422 (2006)

148

Appendix B

Calculations
Anisotropy energy calculation
549) = K.fi*v sin29.
Where Keg is the anisotropy constant, V is the volume of the particle and 0 is the
angle between the magnetization and the easy axis.
For Co, K... = 4.1x105J/m3, v= 1.4137x1026m3, and 9 = 90
So EA(9) = (4.1x105J/m3)(1.4137x1026m3)sin290
EA = 5.8x102’ J
The relaxation time for a moment of a particle r is give by the Neel-Brown
expression. 1 = 10 exp(Kefi V/ kg 77
Where k3 is the Boltzmann constant and To is z 10'9 s, and Ken: is the anisotropy
constant, V is the particle volume and T is the temperature. If the magnetic
moment reverses at a time shorter the experimental time then the system is said
to be superparamagnetic.
So 1 = e(5.8x10-21J/4.14x10-21)x10'9 = 4.06x10’gs
The magnetic moment of a 3 nm Cobalt particle
r = 1.5 nm = 1.5x10'7 cm,
Volume = 4/3m3 = 4/31'r(1.5x107cm)3 = 1.4237x10'20cm3
p = 8.9 g/cc, 8.9 g/cc * 1.4237x10200m3 = 1.38544x10’99
13854010199" 59 glmol = 23483241021 mol
2.3483x10'21mol * 6.022x1023Atoms/mol = 1414 Atoms,

mg: 1.715 Bohr magnetons/Atom

149

 

1414 * 1.715 = 2425 Bohr magnetons 1413: 9.234x10’24 Am2
9. 2342410-24 Am2* 2425 Bohr magnetons = 2. 249x10” Am2
m = 2.249x10'20Am2, 1emu = 10‘3Am2 m = 2.249x10'17emu
Zeeman energy of particle in earths magnetic field. m is the magnetic moment of
the particle, H is the magnetic field, and 0 angle between them.
15. = m*H case and m = 2.25x10’7emu, H = 0.5 G, e = 0
Then E2 = 1.125x10’7erg, 1 J=107erg
E. = 1.1252410"?4 J
Ampere’s law
dH=kI(dI/r2)sin0
Where dH is the resultant magnetic field of the moving charge, k is a constant, I
is the current that travels along the path dl, r is the distance from the path to field
of interest, and 0 is the angle between direction of the moving charge and the
field.
Surface area of circle (loop of coil)
m2 = A, r = 0.5 mm = 0.5x10'3 m, "(0.5241002 = 7.854x107 m2
The magnetic induction 8 of 1 tesla generates a force of 1 newton per meter on a
conductor carrying a current of 1 ampere perpendicular to the direction of the
induction. Earths magnetic at the surface is 0.5 G or 5x105 T
The magnetic moment (m) of a single Cobalt nanoparticle that has a diameter of
3 nm, r =1.5 nm =1.5x10'7 cm
Volume of a sphere V= 4/317r3 = 4/311(1.5x107 em)3 = 1.4137x1020 cm3

Density of cobalt is 8.9 g/cm3, F .W.= 59 g/mole

150

 

 

8.9 g/cm3 / 59 g/mole = 0.1508 moles/cm3, Avogadro’s # 6. 022x10” atoms/mole
(6. 022x10” atoms/mole)(0.1508 moles/cm3) = 9.08x1022 atoms/cm3

The average Bohr magneton per atom of cobalt is r73 = 1.715

One Bohr magneton = 9.2741x10'24 Am2 or (JlT)

So (9.08x1022 atoms/cm3)( 1.4137x1020 cm3) = 1284 atoms per nanoparticle.
1284 atoms per nanoparticle * 1.715 Bohr magneton per atom = 2200 Bohr
magnetons per nanoparticle.

So m = 2200 * 9.2741x1024 J/T = 2.04x10'20 J/T for a single 3 nm nanoparticle.
Note: For paramagnets in a magnetic field, their magnetization, M is proportional
to the field H. Energy of a magnetic moment in a magnetic field is E = -pomoH.
Langevin function assumed that moments are non-interacting so it uses classical
Boltzmann statistics. P(E) = e('E/"B77

Thermal energy = kBT

Magnetization is defined as the magnetic moment per unit volume.

M = NmL (pomH/kBD N is the total number of moments per unit volume and m is
the magnetic moment.

The Langevin equation L (pomH/kB'D = coth(pomH/k37) - (kBT/pomH)

The Langevin equation leads to paramagnetic susceptibility and the curie law
susceptibility. x = M/H where M is the magnetization and H is the applied
magnetic field. Field intensity: 8 = mH, H is the magnetomotive force per
unit-length of circuit 8 the resulting flux density per unit area And m the
permeability of the material

Curie Weiss Law X = C/ (T-Tc)

151

 

X is the susceptibility, T is the absolute temperature, Tc is the Curie pint
temperature and C is the Curie constant of the material
Values and calculations for theoretical maximum.

Surface area of a 1 x 3 pm bacterium

Area of cylinder = 223 + (Zarx L) , r = 500 nm, L = 3,000 nm.

 

Surface area of the bacterium is = 10,995,574 rim2

Volume of bacterium

 

Volume of cylinder nr2 x L, 24*5002*3ooo = 2.356x10'18 nm3
Area covered by a 30 nm nanocomposite
Area of a circle 2212, r = 15 nm,
Area covered by a 30 nm nanocomposite is 707 nm2
So 10, 995,574 nm2/ 707 nm2 = 15,552 nanocomposite per bacterium
Volume coil if L = 2 mm, r = 10 pm
2221’2 x L = ”(0.001 cm)2 x (0.2 em) = 6.3x107 cc, 1 cc = 1 mL
Speed of particles through the coil, pump flow rate = 1 mUmin = 1.666x10i2
cm3/sec
(flow rate)x(area of coil) = (1.666x102 cm3/sec)( 220.0012) = 53.03 m/sec
time to pass through the coil is 37.7 psec.
Magnetic moment (m) of a cobalt nanoparticle with a radius of 10 nm.
Density of cobalt 8.9 g/cc, Cobalt has a mass magnetic moment m of 100
emu/g, So 100 x 8.9 = 890 emu/cc, 4232211003 = 4.188x1018 cc,
(4.188x10'18 cc)( 890 emu/cc) = 4.1x10’15 emu, units (m) 1 Am2 = 103 emu

m= 4.1x1018 Am2

152

 

REFERENCES
. CRC handbook of Chemistry and Physics 81 edition, 2000.

. D. Jiles, Introduction to magnetism and magnetic materials, Chapman and
Hall, New York. 1991 "

. H.E. Burke, Handbook of Magnetic Phenomena. Van Nostrand Reinhold
Company, New York. 1986.

. D.C. Mattis, The Theory of Magnetism Made Simple. World Scientific,
New Jersey. 2006.

. A.H. Lu, E.L. Salabas, F. Schuth, Magnetic Nanoparticles: Synthesis,

Protection, Functionalization, and Application. Angew. Chem. Int.
Ed, Vol:46 lssue:8 pg:1222-1244, 2007.

153

-