<%BANNER%>

Synthesis and Characterization of Functionalized Magnetite Nanocomposite Particles for Targeting and Retrieval Applications


PAGE 1

SYNTHESIS AND CHARACTERIZATION OF FUNCTIONALIZED MAGNETITE NANOCOMPOSITE PARTICLES FOR TARGETING AND RETRIEVAL APPLICATIONS By BARRY WILLIAM MILLER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Barry William Miller

PAGE 3

This thesis is dedicated to my parents, Ca rol and Bill Miller, who actually did the hardest work of all by raising me and alwa ys encouraging me to do my best.

PAGE 4

ACKNOWLEDGMENTS Many people have contributed their time and effort in helping me to accomplish this project. First and foremost, I must thank my advisor and committee chair, Dr. Laurie Gower, for giving me direction in my research and so many helpful suggestions when things just would not seem to work right in the lab. I am truly thankful that I had the opportunity to work with and learn from her. I must also thank the other members of my committee, Dr. Hassan El-Shall and Dr. Rolf Hummel. They have been nothing but courteous, helpful, and accommodating whenever I have sought their assistance. Next, I must thank the Gower research group. In particular, Matt Olszta has spent so much of his time helping me do all sorts of stuff. He taught me how to use the TEM, how to use a bunch of computer programs, and just about how laboratory research is supposed to work in general. I also thank Xingguo Cheng for running all of my XRD samples, and even for all the humorous conversations about American celebrities and culture. The unique personalities of Debra Lush, Fairland Amos, and Lijun Dai always helped make the days in lab much more enjoyable. I also thank Dr. Mark Meisel and his graduate student Ju-Hyun Park of the Physics Department for running all of my SQUID samples (free of charge) and helping me analyze the data. In addition, I acknowledge the Major Analytical Instrumentation Center (MAIC) for the use of the transmission electron microscope and x-ray diffractometer. iv

PAGE 5

I thank all of my friends who have been there over the years to talk to and laugh with. I must especially thank Kevin, Smooth, Keith, Charlie, Nate, Spooner, and Pye for being the reliable drinking crew that I would go out with most weekends. Somewhere between the hundreds of beers and dozens of stories we have all had together, I think we grew up a little bit along the way, or at least learned more about life. All of my years here would have been much more of a chore without them, and I attribute much of my success (and failure) to this group of guys. I also want to acknowledge and thank all those people who made the road bumpy along the way. And not because I enjoyed being subjected to painfully long discussions on irrelevant issues, receiving deceptive and misleading information, or even listening to those who self-proclaim their unequaled greatness, but rather because it made me realize that these same kind of people will still be inevitably present in every arena of my life even after this thesis and university are behind me. I have had to learn how to diplomatically interact with these people, and so it is for this reason that I thank them. Finally, I want to thank my parents, Carol and Bill, and my two sisters, Leigha and Lindsay. They have always been the one bastion of love, support, and advice that I never have to think twice about going to in good times or bad. Anything that I might ever accomplish in my lifetime should be rightfully attributed to them, especially my parents. They have raised me to be the person I am today and they have always encouraged me to cultivate my talents and pursue my interests. I will be forever proud to call them my mom and dad. v

PAGE 6

TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES...........................................................................................................ix ABSTRACT......................................................................................................................xii CHAPTER 1 INTRODUCTION........................................................................................................1 Biological Inspiration...................................................................................................1 Purpose of Research.....................................................................................................2 Specific Aims................................................................................................................3 2 BACKGROUND..........................................................................................................5 Introduction...................................................................................................................5 Magnetic Basics............................................................................................................5 Types of Magnetism.....................................................................................................7 Paramagnetism......................................................................................................8 Ferromagnetism and Ferrimagnetism....................................................................9 Superparamagnetism...........................................................................................12 Antiferromagnetism.............................................................................................13 Magnetic Particle Applications...................................................................................14 Particle Synthesis........................................................................................................17 3 PARTICLE SYNTHESIS WITH POLY(ACRYLIC ACID) ADDITIVE.................22 Introduction.................................................................................................................22 Materials.....................................................................................................................23 Methods......................................................................................................................24 Setup....................................................................................................................24 Synthesis..............................................................................................................25 Separation............................................................................................................26 vi

PAGE 7

Characterization Results and Discussion....................................................................27 Transmission Electron Microscopy (TEM).........................................................27 X-Ray Diffraction (XRD)....................................................................................35 Superconducting Quantum Interference Device Analysis (SQUID)...................37 Fluorescence Labeling.........................................................................................43 Procedure......................................................................................................43 Fluorescence microscopy.............................................................................44 4 PARTICLE SYNTHESIS WITH POLY-L-GLUTAMIC ACID ADDITIVE...........47 Introduction.................................................................................................................47 Materials.....................................................................................................................48 Methods......................................................................................................................49 Characterization Results and Discussion....................................................................50 Transmission Electron Microscopy (TEM).........................................................50 X-Ray Diffraction (XRD)....................................................................................53 Superconducting Quantum Interference Device Analysis (SQUID)...................55 Fluorescence Labeling.........................................................................................58 5 PARTICLE SYNTHESIS WITH POLY-L-LYSINE ADDITIVE............................61 Introduction.................................................................................................................61 Materials.....................................................................................................................62 Methods......................................................................................................................63 Characterization Results and Discussion....................................................................64 Transmission Electron Microscopy (TEM).........................................................64 X-Ray Diffraction (XRD)....................................................................................65 Superconducting Quantum Interference Device Analysis (SQUID)...................67 6 CONCLUSIONS AND FUTURE WORK.................................................................70 LIST OF REFERENCES...................................................................................................75 BIOGRAPHICAL SKETCH.............................................................................................81 vii

PAGE 8

LIST OF TABLES Table Page 2-1 Magnetic units and conversion table..........................................................................7 3-1 Experimental design of trials with poly(acrylic acid)..............................................26 3-2 Magnetic quantities determined from SQUID analysis of control particles............38 3-3 Magnetic quantities determined from SQUID analysis of PAA particles...............40 4-1 Experimental design of trials with poly-L-glutamic acid.........................................50 4-2 Magnetic quantities determined from SQUID analysis of GLU particles...............56 5-1 Experimental trials with poly-L-lysine....................................................................64 5-2 Magnetic quantities determined from SQUID analysis of LYS particles................68 viii

PAGE 9

LIST OF FIGURES Figure Page 2-1 Magnetization curves for the different types of magnetism.......................................8 2-2 Magnetic alignment in paramagnetic materials; (a) is in the absence of a magnetic field, and (b) shows the response in a moderately applied field................................9 2-3 Schematic of a hysteresis loop.................................................................................10 2-4 The magnetic origin of magnetite............................................................................12 2-5 Stages of gel conversion to magnetite. (a) amorphous gel (0 min)(b) formation of hexagonal platelets (15 min)(c) primary particle formation (30 min)(d) primary particles aggregate to larger particles (45 min); (e) uniform spherical particles (120 min)..........................................................................................................................19 2-6 Rod-like crystals of goethite formed as an unwanted product of particle synthesis.21 3-1 Repeat unit structure of poly(acrylic acid)...............................................................22 3-2 Digital picture of setup.............................................................................................25 3-3 TEM micrographs of control particles.....................................................................28 3-4 Electron diffraction pattern of control sample. This pattern was indexed and found to correspond to magnetite. Note that the presence of diffracted rings rather than spots indicates polycrystalline nature.......................................................................29 3-5 TEM micrographs of particles with PAA, MW = 2,100. Both of the pictures above show particles synthesized with a PAA concentration of 130 ug/mL. The sample with a concentration of 220 ug/mL did not convert during the aging process and thus no particles were formed..................................................................................29 3-6 TEM micrographs of particles with PAA, MW = 6,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 200 ug/mL......................................................30 3-7 TEM micrographs of particles with PAA, MW = 15,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 230 ug/mL......................................................30 ix

PAGE 10

3-8 TEM micrographs of particles with PAA, MW = 30,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 200 ug/mL......................................................31 3-9 X-ray diffraction spectrum of control particles. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction..........................................................................36 3-10 X-ray diffraction spectrum of composite particles with PAA additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction.............................................36 3-11 Hysteresis curves for control particles.....................................................................38 3-12 Hysteresis curves for PAA modified composite particles........................................39 3-13 Transmission and fluorescence micrographs of control particles. The light transmission image on the left shows clumps of particles dispersed on the slide. The fluorescence micrograph on the right shows the fluorescence (or lack thereof) of that same area.......................................................................................................44 3-14 Fluorescence images of PAA composite particles. The first two images come from the sample with MW=15,000 and concentration 100 ug/mL. The third image comes from sample MW=6,000 and concentration 100 ug/mL...............................45 4-1 Repeat unit structure of poly-L-glutamic acid.........................................................47 4-2 TEM micrographs of particles with poly-L-glutamic acid, MW = 7,500. The top two images are taken from the sample with a GLU concentration of 80 ug/mL. The bottom two images had a GLU concentration of 100 ug/mL...................................51 4-3 TEM micrographs of particles with poly-L-glutamic acid, MW = 13,600. The top two, middle two, and bottom two images have concentrations of 60 ug/mL, 100 ug/mL, and 150 ug/mL, respectively.......................................................................52 4-4 X-ray diffraction spectrum of composite particles with GLU additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction.............................................54 4-5 Hysteresis curves for GLU modified composite particles.......................................55 4-6 Fluorescence images of GLU composite particles. Both images come from the sample with MW=13,600 and concentration 150 ug/mL.........................................59 5-1 Repeat unit structure of poly-L-lysine.....................................................................61 x

PAGE 11

5-2 TEM micrographs of particles with poly-L-lysine, MW = 27,000. The top two images are taken from the sample with a LYS concentration of 100 ug/mL. The bottom two images had a LYS concentration of 250 ug/mL...................................64 5-3 X-ray diffraction spectrum of composite particles with LYS additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction.............................................66 5-4 Hysteresis curves for particles with LYS additive...................................................67 5-5 Hysteresis curves for both LYS and control particles..............................................68 xi

PAGE 12

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SYNTHESIS AND CHARACTERIZATION OF FUNCTIONALIZED MAGNETITE NANOCOMPOSITE PARTICLES FOR TARGETING AND RETRIEVAL APPLICATIONS By Barry William Miller May 2004 Chair: Laurie Gower Major Department: Materials Science and Engineering There is much interest in the use of biologically functional magnetic particles for biomedical applications. The general idea is that an external magnetic field can guide the particles to a specific area of the body, and then the functionality of the particles will allow them to target specific cells or tissues. These particles must therefore be small enough to flow through blood capillaries. This study presents the synthesis procedure and characterization of a novel magnetic particle system that possesses organic functionality. These particles can achieve high magnetizations due to their high content of magnetite, Fe 3 O 4 and they are generally on the order of 150-350 nanometers. Functionality was afforded to the particles by addition of three separate polymers during the solution synthesis procedure. The effect of the molecular weight and concentration of the selected polymer additive was explored. Poly(acrylic acid) and poly-L-glutamic acid were used to give the particles carboxyl functionality. Attempts were also made to incorporate poly-L-lysine to give them an amino functionality; xii

PAGE 13

however the nature of this polymer in solution prevented the synthesis of composite particles; instead, a non-functional colloid of magnetite particles was formed. The functionality and reactivity of the carboxyl-modified particles were demonstrated by conjugation of a fluorescent probe and subsequent fluorescence microscopy. All particle trials were characterized with transmission electron microscopy (TEM), which contrasted the morphology of the composite particles with the non-functional control particles. A superconducting quantum interference device (SQUID) was employed to measure magnetic properties and determine magnetic behavior. Finally, x-ray diffraction (XRD) was performed to confirm that the crystalline phase of the particles was magnetite. Overall, this research shows that a new door has been opened to a magnetic particle system that can be used in a wide range of applications. xiii

PAGE 14

CHAPTER 1 INTRODUCTION Biological Inspiration Iron-based compounds are ubiquitous in nature. They exist as geological, biological, and even extraterrestrial minerals. The primary interest for this project is in the iron oxides, namely magnetite. However, other iron oxide phases and some iron oxy-hydroxides will be necessarily addressed in the background chapter. The inspiration behind this thesis project stems from the discovery and research surrounding iron biominerals [1]. Iron-containing minerals have been found in a large variety of organisms, ranging from bacteria to humans [2, 3]. Although the function of a mineral is not known for every organism in which it is found, it has been discovered that organisms utilize these minerals for various purposes, including mechanical grinding and magnetotaxis [1, 4]. What makes these mineral occurrences so interesting is that they must be produced in natural conditions and must have some mechanism controlling their synthesis because such a high level of size and shape control is achieved. It has been found that the synthesis may be controlled in some different ways; a protein matrix is sometimes present which allows the mineral to grow epitaxially [5]; in some instances, a mineral phase is synthesized within a vesicle that can control local concentrations of reactants [6, 7]. Whatever the specific case may be, such research shows that these biominerals and their syntheses are closely intertwined with the effects and properties of various proteins or polymers. 1

PAGE 15

2 Of particular interest to this project are the nano-sized particles of magnetite (Fe 3 O 4 ) found in magnetotactic bacteria. These organisms live in marine or fresh water sediments and their magnetic particles respond to geomagnetic fields which in turn orient and guide them to nutrients [4]. Much research has been performed to characterize these magnetic biominerals and understand their synthesis and function [8-16]. What has been found is that these particles are contained in what is called a magnetosome. The magnetosome is essentially a lipid bilayer membrane with various bound proteins. This structure is what regulates the flux of iron for synthesis, however the exact mechanism is not fully known. The resulting particles are very controlled in size and shape, but also vary by species. This suggests that the protein composition of the magnetosome may also vary and thus influence the nucleation and growth of the magnetite crystals. It is this association of a macromolecule with a magnetic particle that raised interest for this project. The question was, what happens if we try to do this synthetically? My advising professor, Dr. Laurie Gower, made some efforts in her own graduate research to control particle morphology using synthetic polymers. Rather than achieving elongated or specifically oriented minerals, she discovered an interesting matrix-like composite morphology that was not researched further. Therefore, the impetus behind this research was the realization that magnetic mineral and polymer can be used in conjunction to create a magnetic nano-composite with organic functionality. Purpose of Research The overall purpose of this research is to create a magnetic nano-composite particle system with organic functionality. Such a system is desired for various targeting and/or retrieval applications, both medical and environmental. The idea is that the particles can be magnetically guided to specific areas of the body, and/or have the capability for

PAGE 16

3 retrieval by the same mechanism. It was desired to try to achieve a large magnetic moment with this system, thus the material selected for the magnetic aspect was magnetite (Fe 3 O 4 ). Magnetite has been well known and characterized for hundreds of years, and the chosen method of synthesis is relatively inexpensive and easy; therefore, there is also a feasible capacity for large-scale production. The organic functionality is afforded to this system by the addition of a polymer. The three polymers used in this research were poly(acrylic acid), poly-L-glutamic acid, and poly-L-lysine. The first two have carboxylic acid functional groups, whereas the latter has a free amine group. These functionalities can be used with bioconjugation techniques to attach proteins, drugs, antibodies, or other desired probes. Homopolymers of poly(amino acids) were investigated because specific sequences of amino acids are known to bind to specific cells or other receptors. The use of an amino acid homopolymer is the first step toward incorporating sequenced polypeptides. Therefore, the ultimate goal for this system would be to incorporate specifically sequenced polypeptides into magnetic particles that would then target a preselected cell moiety. However, this goal was unrealistic for this project because of the difficulty and expense of synthesizing specific protein sequences at moderate molecular weights. Therefore, this research stands to demonstrate the feasibility of producing such a system. Specific Aims There were three primary goals that I wanted to achieve with this project. The first was to optimize the technique of particle synthesis. The background and procedure of the actual method will come in later chapters, but it was a challenge to find the right parameters to make the reaction go as near to completion as possible on each trial.

PAGE 17

4 Clearly, a reliable and repeatable technique was a necessary component for this research because it facilitates the accurate analysis of results. Once the synthesis was optimized, the next specific aim was to characterize the particle system. In order to do this, it was necessary to first confirm that the mineral phase produced was indeed magnetite (Fe 3 O 4 ). This was performed using x-ray diffraction. It was also necessary to show the particle morphology to support that indeed the particle structure has a matrix-like composite morphology. This was done by transmission electron microscopy, where bright field images were taken to show not only the particle morphology, but also particle size. For biomedical applications, it is important that the particles be small enough to flow through blood capillaries, which are about 5-10 microns in inner diameter [17]. The magnetic properties of the composite particles were measured using a superconducting quantum interference device (SQUID). This characterization was important so that this magnetic particle system can be compared with other magnetic systems in related applications. The final aim of this project was to demonstrate the organic functionality of the particles. Transmission electron microscopy shows that the polymer is present, but it is important to prove that the polymer is also functional. To do this, a fluorescent probe was conjugated to the polymer on the particles. The resulting fluorescence shows that the functional groups of the polymer are indeed reactive.

PAGE 18

CHAPTER 2 BACKGROUND Introduction Since this project deals with the characterization of a magnetic particle system, it is important that we first understand the basics of magnetism. The different types of magnetism will also be discussed and corresponding reference will be given to the materials that were encountered in this research, namely magnetite and goethite. Important magnetic quantities and properties will also be explained, especially as they become relevant to this work. In addition, this chapter will present some of the current magnetic particle systems and also how the composite particles of this research might be used in similar applications. Finally, this chapter will conclude with a discussion on the actual method of particle synthesis and how it was modified for use in this project. Magnetic Basics The root of magnetism is based on the response that a material has when exposed to an external magnetic field. The electron spins in the material align in the direction of the applied field, thereby magnetizing the material. It is important that we first define some magnetic properties before we get too far into discussion. An applied magnetic field, H, incites a response from a material called magnetic induction, B. The relationship between B and H can be defined by equation 2-1, B = H + 4M (2-1) where M is the magnetization of the material. Magnetization is the magnetic moment per unit volume, and is a property that depends on the magnetic moments of the constituent 5

PAGE 19

6 atoms, as well as their interactions with each other. It is important to note that equation 2-1 is for cgs units; the relationship in terms of SI units is shown by equation 2-2. B = 0 H + 0 M (2-2) 0 is a constant called the permeability of free space. There are also common terms for ratios between some of these different quantities because the magnetic properties of a material are often defined by how they vary with an applied magnetic field. So, the ratio of M to H is called the susceptibility and is indicative of how responsive a material is to an applied magnetic field. Equation 2-3 defines this. = M/H (2-3) The ratio of B to H is called the permeability and is indicative of how well the magnetic field can permeate the material. Equation 2-4 defines this property. = B/H (2-4) From equations 2-3 and 2-4 above we can derive a relationship between the susceptibility and the permeability, = 1 + 4 (2-5) or, in SI units / 0 = 1 + (2-6) A summary of these quantities, their units, and their conversion factors is included in Table 2-1. The following section breaks down the different types of magnetism and will reference these properties, so it is important to keep them and their relationships with each other in mind.

PAGE 20

7 Table 2-1: Magnetic units and conversion table Quantity Symbol Gaussian Units (CGS Units) Conversion Factor C SI Units Magnetic Flux Density; Magnetic Induction B Gauss (G) 10 -4 Tesla (T) Wb/m 2 (Wb weber) Magnetic Flux Maxwell (Mx), Gilbert cm 2 10 -8 Weber (Wb) Magnetic Field Strength H Oersted (Oe) 10 3 /4 A/m Magnetization M emu/cm 3 10 3 A/m Magnetic moment m emu (electro-magnetic unit) 10 -3 A m 2 Permeability of vacuum o dimensionless 4 10 -7 Wb/(A m) Permeability dimensionless 4 10 -7 o Wb/(A m) Types of Magnetism Now that the basics have been defined, we must now explain how these properties relate to the actual magnetic behavior of a material. Most magnetic materials are classified by how they relate to these quantities. In particular, magnetization curves are often used to describe the nature of a magnetic material. A magnetization curve plots magnetization, M, or magnetic induction, B, as a function of applied magnetic field, H. Examples of these general curves and the types of magnetism they correspond to can be seen in Figure 2-1 [18]. While individual treatment will be given to paramagnetism, ferromagnetism, ferrimagnetism, and antiferromagnetism, a discussion of the properties of diamagnetic materials will not be presented at length because this form of magnetism does not really

PAGE 21

8 relate to this research. Let it suffice to say that diamagnetic materials tend to exclude the magnetic field from their interior, thus they usually have a small and negative susceptibility, with a permeability less than one; this behavior is shown schematically in the left plot of Figure 2-1. Figure 2-1: Magnetization curves for the different types of magnetism. Notice the difference in scales between the two plots. Paramagnetism The phenomenon of paramagnetism occurs due to weakly coupled magnetic moments. With no applied field, thermal energy causes the magnetic moments to randomly align, resulting in a net magnetic moment of zero [19]. When a magnetic field is applied, the individual moments respond by turning in the direction of the field, however the net magnetization is still relatively weak because the moments do not completely align with the applied field. Therefore, values for susceptibility are small, and permeability is just greater than one. A schematic of the alignment of spins in paramagnets is shown in Figure 2-2. Clearly, temperature will always play a role in paramagnetism, as a higher temperature requires a greater magnetic field to overcome the thermal forces working against the ordered alignment of magnetic moments. Typical paramagnetic materials include transition metal salts and rare earth salts. As we will see

PAGE 22

9 in the upcoming sections, paramagnetic behavior is also observed in other types of magnets under certain conditions. Figure 2-2: Magnetic alignment in paramagnetic materials; (a) in the absence of a magnetic field, and (b) the response in a moderately applied field. Ferromagnetism and Ferrimagnetism For all intensive purposes, the main difference between ferromagnetic and ferrimagnetic materials is that the latter materials are ceramic and good insulators rather than conductors. Other than this, both kinds of magnetism are very similar and will be treated the same in this discussion. Both types of materials typically show a hysteresis loop as a magnetization curve. A hysteresis loop is created because after a magnetic field is applied and then removed, the material retains some magnetization. This is called remanent magnetization. In order to bring the magnetization down to zero, a magnetic field must be applied in the opposite direction until there is no net magnetization remaining. The value for the strength of the field necessary to do this is called the coercivity, H c If the field is applied to saturation in the negative direction and then removed, then applied again in the positive direction, a full hysteresis loop is formed. A schematic of such a magnetization curve is shown in Figure 2-3.

PAGE 23

10 Figure 2-3: Schematic of a hysteresis loop. Materials that display such hysteresis behavior can be classified as either hard or soft magnets. Hard magnets have a large remnant magnetization and a large coercivity, therefore they show a large area within the hysteresis loop. They are so-called hard magnets because it is hard to bring them to saturation and consequently coerce them back to zero. These types of magnets are often used where magnetic memory is desired, such as in magnetic recording media, because of their large remnant magnetization. Soft magnets, however, possess a low coercivity and a low remnant magnetization. This also means that a much smaller magnetic field would be necessary to reach a saturation point of magnetization. However, because of their low remnant magnetization, these materials are usually not good for recording media because increases in temperature can disrupt the alignment of moments very easily. Now we must look at why this behavior occurs in ferromagnetic and ferrimagnetic materials. The spins of unfilled d-band electrons spontaneously align and create a magnetic moment [19]. However, this does not occur over the entire crystal, but rather these groups of aligned magnetic moments are contained in sub-structures known as domains [20]. Each of these domains has their own saturation magnetization without the presence of an external field. However, the material as a whole maintains a net

PAGE 24

11 magnetization of zero in the absence of an applied field because each of the domains are aligned in different directions, effectively canceling each other out. When a magnetic field is applied, the domains already aligned in the direction of the field begin to grow at the expense of the unfavorably aligned domains, thus creating a net magnetization. These domains continue to grow until the material effectively contains one single domain that is oriented in the direction of the field. It is at this point that the material reaches its saturation magnetization. Now that we can see how the domains transform to affect magnetic behavior, it should make sense why hysteresis behavior is observed for these types of materials. It is also important to mention that temperature again plays an important role in ferroand ferrimagnetism. As temperature is increased, there comes a point where the thermal fluctuations will overcome the force of the spontaneously aligned spins and cause them to become randomly oriented [21]. Once this occurs, the material exhibits paramagnetic behavior. This threshold temperature is known as the Curie temperature, T c It is very important to keep this value in mind, especially when dealing with a material with a relatively low T c because it can greatly affect the materials efficiency in its intended application. In addition to this, even if the Curie temperature is not reached, an increased temperature will result in a lower magnetization yield because of the increased thermal fluctuations. It should here be noted that the magnetic material utilized in this research, magnetite (Fe 3 O 4 ), is ferrimagnetic in nature. A more representative way to express the composition would be to write Fe 2+ Fe 3+ 2 O 4 because the iron ions exist in two different valence states, and this is actually what gives rise to the magnetic moment. This material

PAGE 25

12 has the inverse spinel crystal structure, with 8 Fe 3+ ions occupying tetrahedral sites, 8 Fe 3+ occupying octahedral sites, and 8 Fe 2+ occupying octahedral sites. The Fe 3+ ions in the octahedral sites have opposing moments to the Fe 3+ ions in the tetrahedral sites, thus they cancel each other out. Therefore, the overall magnetic moment comes from the sum of the magnetic moments of the Fe 2+ ions in the octahedral sites. Figure 2-4 shows a schematic of the origin of magnetism in this material [22]. Figure 2-4: The magnetic origin of magnetite Superparamagnetism Superparamagnetism is an interesting phenomenon that comes into play when ferromagnetic or ferrimagnetic particles become very small. At particle sizes of about 10 nanometers, these materials begin to exhibit paramagnetic behavior, even when they are below their Curie temperature. The reason for this is that thermal effects, while not strong enough to overcome the forces between individual atoms, are strong enough to

PAGE 26

13 change the magnetization direction of the entire particle. The result is a random arrangement of magnetic directions among crystallites, thus giving a net magnetic moment of zero. This phenomenon gives rise to the limitation of how small magnetic recording media can get because superparamagnetism will cause the particles to loose their memory from thermal influences. Superparamagnetic particles are therefore often used in many magnetic systems in the biomedical field because not only are they small, but they also do not retain any magnetic remanence. The latter reason is important because it means that the particles will not aggregate due to magnetic forces, however the trade-off is that the particles are paramagnetic in behavior and therefore it is more difficult to achieve a high magnetization. For these reasons, this research aimed to use particles that were in the size range of a few hundred nanometers, thus allowing them to retain their ferrimagnetic properties yet still be small enough to flow through blood capillaries if necessary. As we will see in the experimental chapters, the particles are very soft magnets and have only a small remnant magnetization. Antiferromagnetism Like ferromagnets, materials exhibiting this behavior have a spontaneous alignment of moments. The difference is that the adjacent atoms in these materials have antiparallel spins, thus there is no net magnetic moment observed. This is really as far into discussion as we need to get for this project because it explains all we need to know. The particle synthesis process often creates some unwanted byproducts because of oxygen getting into the system (details will be explained in a following section). The main byproduct is goethite, -FeOOH, an iron oxyhydroxide that possesses antiferromagnetic

PAGE 27

14 properties. Therefore, this material can be separated relatively easy since it will not really be attracted to a magnetic field. In practice, the separation is a little more difficult, but for all intensive purposes it is important to note that this byproduct will not contribute much to the magnetic properties of the overall particle system. Also of note is that hematite, Fe 2 O 3 is antiferromagnetic, and it is possible that this material may also form in very small quantities during synthesis, however it can be removed in the same manner as the goethite. Magnetic Particle Applications The capabilities of creating nano and micro-sized particles of magnetite have lead to the use of these small magnetic particles in a range of applications [23]. Many studies have been conducted to characterize the magnetic behavior of various sizes of magnetite particles [24-26]. Because of its useful properties, magnetite has been the material of choice for many magnetic particle systems. Ferrofluids are an interesting example of such systems. A ferrofluid is basically a colloidal solution of magnetic particles that are suspended in either a polar or non-polar liquid [27]. Magnetite is commonly used as the magnetic material, but iron and cobalt particles have been used as well. The particles are typically on the order of about 10 nanometers so that they are superparamagnetic. This is desired so that the fluid remains as a stable suspension and the magnetic particles do not aggregate together and form clumps or settle in the absence of a magnetic field. When a magnetic field is applied, however, the particles will respond and are often used as a seal that can be applied or removed with a magnetic field. These fluids have been used in applications such as rotary seals for disk drives and dampers for audio speakers [28].

PAGE 28

15 Recently there has been interest in using ferrofluids for biomedical applications. The important factor in making this application transition is that the fluid must be biocompatible. There has been some success in creating water-based suspensions [29], and such ferrofluid systems have been used for cell sorting [27]. In order to do this, a biological effector is bound to the particle surface so that the particles can target specific cells. These cells can then be sorted by employing a gradient magnetic field to separate them. This system begins to appear very similar to the system characterized in this thesis because it entails that the magnetic particles are biologically functional. However, this one system in particular uses particles of Fe 2 O 3 a material that will not achieve a very high magnetization as compared to a system with Fe 3 O 4 In addition, where magnetite is used as the material of choice, the particles are still in the superparamagnetic size range, thus they will still not achieve a high magnetization as compared to larger particles that exhibit ferrimagnetic behavior. This is a very important factor that will come into play especially for in vivo applications. Magnetic particle systems are also being explored for site-specific tumor therapy and/or drug delivery. One of the most interesting properties of magnetic particles is that they will begin to heat up if an alternating magnetic field is applied to them [30]. Therefore, the theory behind some of these systems is that it should be possible to magnetically guide the particles to a tumor, apply an alternating magnetic field, and consequently treat the tumor via magnetocytolysis. There has been some demonstrated success doing this [31], however many systems are also trying to incorporate a drug delivery mechanism to additionally treat specific areas of the body. For example, there is one system that contains dispersed magnetic particles in a polymer matrix which also

PAGE 29

16 contains a dispersed therapeutic agent. When an oscillating magnetic field is applied, the diffusional release of the drug is increased [32]. Many studies involving magnetic particles have turned their attention to coatings, or, to the use of magnetic microspheres. This is being done in some instances for certain biocompatibility issues [33], but most often it is for the purpose of functionalizing or attaching drugs to the particles for use as targeted drug delivery systems [34-40]. A very common methodology for these delivery systems is to create a crosslinked protein or polysaccharide microsphere with a magnetic core [41-47]. The drugs can then be chemically or physically associated with the particles. A large drawback to these systems is that they contain only a small volume fraction of magnetic material, and this of course degrades overall magnetic capabilities. Functional and site-specific magnetic particles are also of great interest to the field of magnetic resonance imaging (MRI). Magnetic particles work well as negative contrast agents in MRI because they shorten the T 1 and T 2 relaxation times [48-50]. Therefore, for better imaging purposes, it would be useful to have a contrast agent that could target a specific tissue, organ, or tumor. One method of achieving this is through the attachment of antibodies to the magnetic particles [51, 52]. MRI just serves as one more example of a scientific field that can benefit from functional magnetic particles. In conclusion, many of the studies of all these kinds of systems are not adequately focused on the magnetic properties of the particles. For in vivo applications, many of these systems will have difficulty in achieving a large magnetization for guidance purposes, and thus the efficacy of these systems will be deteriorated. The main reason for this is that many of the systems do not have a very high mineral loading due to the

PAGE 30

17 needs of coatings or attachment of other materials. As a typical example where the magnetic properties of a system were characterized, one study used particles that had either a 23 or 29 weight percent mineral loading and their corresponding magnetizations were 20 and 30 emu/gram [35]. The primary argument we are attempting to make with our composite particle system is that the particles developed in this research have a very high content ratio of magnetite to polymer, and so the particles can subsequently achieve a high magnetization. In addition, because our composite particles are not superparamagnetic, it follows that a higher magnetization can be achieved just on the basis of intrinsic magnetic properties. In fact, the primary obstacle to many of the targeted drug therapy systems is that it is difficult to obtain enough magnetic force to retain the superparamagnetic particles in deep body tissue. It should also be noted that the magnetic particles introduced in this thesis do not need to be limited to in vivo applications. There exists the possibility of using this system for targeting and retrieval in environmental applications. For example, the anthrax scares in recent years have led to the need for a way to eliminate spores that may have been spread in a room. A magnetic system such as the one researched in this thesis could target these spores, bind them, and then remove them by magnetically retrieving the particles. This is just yet another example of the utility of these particles. Whether it be drug delivery or targeted MRI contrast, this new magnetic particle system is an improvement on current systems and will be useful for a broad range of applications. Particle Synthesis The method for the particle synthesis used in this research was a modification of the method used by Sugimoto and Matijevic [53]. Matijevic is well known for his work in the synthesis of uniform colloids of many different types of ceramic and metal

PAGE 31

18 particles [23, 54-58]. The synthesis of interest is one where a ferrous hydroxide gel is precipitated from solution, and then this gel is aged to create spherical crystals of magnetite, Fe 3 O 4 Similar types of solution synthesis techniques are very commonly used to produce relatively monodispersed, colloidal particles for many different types of materials [59-68]. It was important to choose a procedure that included the transformation of an amorphous intermediate to the final particle so that the polymer additives used to create the composite particles could have a chance to incorporate during the synthesis. The procedure is composed of two main steps. The first step is to precipitate the amorphous gel. This is done by introducing an iron sulfate solution to a potassium hydroxide solution, thus creating the Fe(OH) 2 gel. The second step is then to place this gel in a 90C oil bath for about four hours, or until the conversion from gel to magnetite is complete. It is important to note that the potassium hydroxide solution is also mixed with a potassium nitrate solution, a component necessary for the conversion to magnetite. What actually occurs during aging is interesting. Figure 2-5 shows the conversion from gel to magnetite as the aging progresses. One of the most interesting variables in this synthesis is the effect that the excess ion concentration of either Fe 2+ or OH has on the resulting particles. It seems that when there is an OH excess, the primary particles are less likely to aggregate together to form particles; instead, the primary particles tend to grow by the addition of individual ions to the crystal surface. This often results in the final particles having a cubo-octahedral morphology. On the other hand, if there is a Fe 2+ excess, it seems that the primary particles do indeed aggregate together and then a surface recrystallization mechanism is

PAGE 32

19 induced to create the final spherical particles of magnetite. As we will see in the experimental chapters to follow, a consistent excess of Fe 2+ was used for all of the trials. Figure 2-5: Stages of gel conversion to magnetite. (a) amorphous gel (0 min); (b) formation of hexagonal platelets (15 min); (c) primary particle formation (30 min); (d) primary particles aggregate to larger particles (45 min); (e) uniform spherical particles (120 min).

PAGE 33

20 This is important because the polymer additive can better be incorporated into the final particle by lodging itself between primary particles. Although this has not been proven to be the exact mechanism, it is a very likely explanation. A further discussion of the formation of the composite particles will be given in the following chapters. Of immediate interest to the synthesis of pure magnetite particles is the oxidation reaction that transforms Fe(OH) 2 into Fe 3 O 4 and then also the oxidation of Fe 2+ to Fe 3 O 4 These are only possible reactions and may not fully describe what is happening [53]. For the Fe(OH) 2 to Fe 3 O 4 reactions: 3Fe(OH) 2 + NO 3 Fe 3 O 4 + NO 2 + 3H 2 O 3Fe(OH) 2 + 2NO 2 Fe 3 O 4 + 2NO + 2H 2 O + 2OH 15Fe(OH) 2 + 2NO 5Fe 3 O 4 + 2NH 3 + 12H 2 O For the oxidation of Fe 2+ to Fe 3 O 4 : 3Fe 2+ + NO 3 + 3H 2 O Fe 3 O 4 + NO 2 + 6H + 3Fe 2+ + 2NO 2 + 2H 2 O Fe 3 O 4 + 2NO + 4H + 15Fe 2+ + 2NO + 18H 2 O 5Fe 3 O 4 + 2NH 3 + 30H + It should also be mentioned that this reaction needs to be performed in an inert atmosphere because the presence of oxygen in the system can cause the precipitation of unwanted crystals, namely goethite, -FeOOH. Since we do live on Earth and oxygen is everywhere, it is inevitable that some oxygen always gets into the system and forms rod-like crystals of goethite. The bulk of the synthesis product, however, is magnetite. Goethite is usually not a huge problem becauseas mentioned earlierit is antiferromagnetic and can therefore be separated from the magnetite by magnetic means. However, some traces of this material still exist in the samples and can even be seen in

PAGE 34

21 some of the images shown in the results of this thesis. Figure 2-6 shows an example of what the goethite crystals look like so that there is no surprise when we examine the results in the coming chapters. Figure 2-6: Rod-like crystals of goethite formed as an unwanted product of particle synthesis.

PAGE 35

CHAPTER 3 PARTICLE SYNTHESIS WITH POLY(ACRYLIC ACID) ADDITIVE Introduction As its name suggests, poly(acrylic acid), or PAA, is an acidic polymer that contains a carboxylic acid group (-COOH) on every repeat unit. The repeat unit structure is shown in the figure below. Figure 3-1: Repeat unit structure of poly(acrylic acid) The carboxylic acid group is very reactive and therefore it will be useful for conjugation purposes when incorporated into the magnetic nanoparticles. PAA is an easily synthesized and common polymer, therefore it was inexpensive to purchase a series of molecular weights. The ability to investigate a range of molecular weights means that possible trends may be uncovered and analyzed. PAA was also chosen because it has the same functionality and acidic nature as poly-L-glutamic acid, the polymer of focus in chapter 4. This chapter will also show the results for the control samples and discuss them in relation to the composite particles that are formed with PAA. The control reaction is simply the synthesis of particles with no polymer added into the solution. The result will be shown to be a colloid of magnetite particles with no organic functionality. Both the control samples and those with PAA were characterized with transmission electron 22

PAGE 36

23 microscopy (TEM) to show morphology and size. Magnetic measurements were made to determine polymeric effects on magnetization, and x-ray diffraction was performed to confirm that the crystalline phase of the particles is indeed magnetite (Fe 3 O 4 ). In addition, fluorescence labeling and microscopy techniques were employed to demonstrate the functionality of the composite particles. Materials Poly(acrylic acid) sodium salt was obtained from Scientific Polymer Products in molecular weights of 2,100 and 6,000; it was obtained from Aldrich Chemical in molecular weights of 15,000, and 30,000. Iron (II) sulfate heptahydrate (Aldrich Chemical) was dissolved in deionized water to make a solution with a final concentration of 0.255 M, which corresponds to an excess ion concentration of [Fe 2+ ] ex = 5 x 10 -3 During the optimization process it was found that this concentration seemed to work best and be the most reliable to get adequate conversion of gel to magnetite. A concentrated potassium hydroxide (KOH) solution was purchased from Acros Organics and diluted to a 0.5 M solution. Potassium nitrate, KNO 3 (Fisher Scientific), was dissolved in deionized water to make a solution with a final concentration of 2.0 M. Although the KOH and KNO 3 solutions must be mixed in equal quantities before the particle synthesis, they were stored separately and kept as stock solutions. The FeSO 4 solution was remade just before each synthesis trial because it begins to precipitate out of solution after sitting for a few days. A deoxygenation solution was also necessary to remove trace amounts of oxygen and carbon dioxide from the nitrogen gas that was used to purge the solutions during synthesis. A pyrogallol (C 6 H 3 (OH) 3 ) and sodium hydroxide (NaOH) solution mixture was used for this purpose. A 5.0 M NaOH solution was prepared, and then solid

PAGE 37

24 pyrogallol was added to create a 1.0 M concentration. This solution was placed into a three-necked flask that was subsequently sealed at all three necks with septa. A fluorescein probe was used for the fluorescence study, specifically 5-(aminoacetamido) fluorescein. Another chemical, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide, or EDAC, was also necessary for the fluorescence study. Both were obtained from Molecular Probes, Inc. Methods Setup In order to perform the synthesis, it was necessary to devise a setup that could separately purge both the FeSO 4 solution and the KOH/KNO 3 solution, and also purge the precipitated Fe(OH) 2 gel with N 2 for ten minutes into the aging process. To do this, the pyrogallol solution was poured into a three-necked flask. The necks were plugged with septa and then the septa were punctured through with 18-gauge septum needles. The needles from the two outside necks were then punctured through other septa that were sealing vialsone containing the FeSO 4 solution and one containing the KOH/KNO 3 solution. The needle in the middle neck of the flask served to deliver the nitrogen from the gas cylinder to the pyrogallol solution. When the gas is turned on, it flows through the needle in the middle neck and bubbles through the pyrogallol solution, thus removing trace amounts of oxygen and carbon dioxide. The gas then escapes through the other two side-neck needles to bubble through the iron sulfate and potassium hydroxide/nitrate solutions. This is what is referred to as the purging period, and this is also what creates an inert atmosphere within the reaction vials. A picture of this setup is shown in the figure below.

PAGE 38

25 Figure 3-2: Digital picture of setup. Synthesis To begin the synthesis, 8 mL of the FeSO 4 solution is added to a vial. 1 mL of the KOH solution and 1 mL of the KNO 3 solution are added to another vial. The poly(acrylic acid) is dissolved into the KOH/KNO 3 solution (Table 3-1 shows the systematic trials performed using PAA). These vials are then plugged with septa and the appropriate needles are inserted. The flow of gas is then turned on and the solutions are in this way purged of oxygen for two hours. After the two-hour purging period, a syringe with a septum needle attached is used to extract the 8 mL of iron sulfate solution from its vial. It is important to note that the syringe was filled and depressed with nitrogen a few times before being filled with solution. The solution was then injected into the vial containing the KOH/KNO 3 solution, thus precipitating the greenish Fe(OH) 2 amorphous gel. The nitrogen is bubbled for an additional minute before the vial is placed into a 90C oil bath. Once placed in the oil bath, nitrogen is bubbled for another ten minutes,

PAGE 39

26 then the needle is removed and the gel is left to age for approximately four hours. It is during this aging process that the magnetite nucleates and grows to form spherical particles. After the aging process is finished, the vial is removed from the oil bath and allowed to cool to room temperature until separation. Table 3-1: Experimental design of trials with poly(acrylic acid) Molecular Weight Quantity Added (mg) Concentration of Polymer (ug/mL) 2,100 1.3 130 2,100 2.2 220 6,000 1.0 100 6,000 2.0 200 15,000 1.0 100 15,000 2.3 230 30,000 1.0 100 30,000 2.0 200 The table above outlines the experimental design of the trials with PAA. Attempts at consistency were made to use concentrations of 100 and 200 ug/mL for each molecular weight, however this was difficult to actually do. The reason is that measuring very small quantities of polymer to exact specifications is hard to doespecially when the polymer is in a powder form. Therefore, I did my best to get as close to these quantities as possible, but as Table 3-1 shows, some of the trials were slightly off this mark. For the purposes of this project, however, these small variances are not a significant concern. Separation A method to separate the particles of magnetite from possible contaminants and unwanted byproducts is necessary because of the nature of the reaction. Poly(acrylic acid) effectively works to inhibit the crystallization process by not allowing primary particles to aggregate freely. By this same notion, the inhibition is what allows the PAA to incorporate into the particles because it traps itself between primary particles during

PAGE 40

27 the aging process. Therefore, after the aging there is likely to be present some unconverted gel. Also, it is so extremely difficult to eliminate all traces of oxygen that some goethite rods will inevitably be formed. So, to separate the magnetite from these unwanted materials, a magnet is held to the vial to attract the particles. Only the particles will be attracted because goethite is antiferromagnetic with no net magnetic moment, and remnant gel is obviously not magnetic because it is amorphous. Therefore, with the particles held in place, the solution is decanted. This process is repeated several times after washing with ethanol. Once the particles have been adequately separated, the solution is poured into a centrifuge tube. The sample is then centrifuged at 6,000 rpm for 3 minutes. After the ethanol is decanted, a kimwipe is placed over the top of the tube and secured with a rubber band. The kimwipe is used to prevent dust or other contaminants from falling into the tube. The sample is then set out in air to dry completely. After drying is complete, the sample is extracted and placed in a small glass storage vial. Characterization Results and Discussion Transmission Electron Microscopy (TEM) Once the particles were synthesized, the next step was to analyze their properties. The first method of analysis was viewing them on the transmission electron microscope (TEM). It was important to do this before any other characterization technique so that it could be confirmed that the desired morphology was obtained. The TEM was used at an accelerating voltage of 200 kilovolts. To prepare the particles for TEM analysis, they had to be mounted onto a grid. Formvar coated copper grids were obtained from Ted Pella, Inc. A small amount of the dry powder of particles was placed into a small (10 mL) beaker. They were then

PAGE 41

28 dispersed in ethanol and sonicated to achieve a better dispersion. Some of the solution with dispersed particles was then micropipetted onto a grid. The grid was viewed under an optical microscope to confirm that an adequate quantity of particles were deposited, and then the grid was placed in an oven at 60C for 3-5 minutes to help the particles set into the Formvar coating. The grid was passed over with a magnet to remove any loose or excess particles. First and foremost, we must look at the control trials so that we have a point of reference for the composite particles. Figure 3-3 shows TEM micrographs of a control sample and Figure 3-4 shows the electron diffraction pattern obtained from these particles. The diffraction pattern was indexed and labeled, showing that the pattern is characteristic of Fe 3 O 4 Figure 3-3: TEM micrographs of control particles. The top two bright field micrographs were color inverted to better show spherical shape. The particles are closely uniform in size, ranging from about 300-500 nanometers in diameter.

PAGE 42

29 Figure 3-4: Electron diffraction pattern of control sample. This pattern was indexed and found to correspond to magnetite. Note that the presence of diffracted rings rather than spots indicates polycrystalline nature. This may or may not mean that individual particles are polycrystalline because the electrons were diffracting off of a group of particles, not just one. Now the granular, composite nature of the particles synthesized with the poly(acrylic acid) additive will be shown. The following figures display the TEM micrographs obtained for each of the samples with PAA. Each separate figure represents the results of a different molecular weight of polymer. However, each figure in itself shows the results for varying concentrations of polymer at its given molecular weight. Figure 3-5: TEM micrographs of particles with PAA, MW = 2,100. Both of the pictures above show particles synthesized with a PAA concentration of 130 ug/mL. The sample with a concentration of 220 ug/mL did not convert during the aging process and thus no particles were formed.

PAGE 43

30 Figure 3-6: TEM micrographs of particles with PAA, MW = 6,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 200 ug/mL. Figure 3-7: TEM micrographs of particles with PAA, MW = 15,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 230 ug/mL.

PAGE 44

31 Figure 3-8: TEM micrographs of particles with PAA, MW = 30,000. The top two images are taken from the sample with a PAA concentration of 100 ug/mL. The bottom two images had a PAA concentration of 200 ug/mL. By looking at any of the TEM images of composite particles with PAA, it can be seen that they have a grainy or almost fuzzy-looking texture. This is indeed the morphology indicative of the nanocomposite particles. But the question arises: what causes this? Although there is no proven or concrete explanation, these results point to the fact that the primary particles of magnetite aggregate during synthesis. Once the primary particles form, they encounter other primary particles and through a surface recrystallization mechanism they grow to a larger particle. This process occurs between many primary particles to generate what becomes the final sphere. When a macromolecular additive such as poly(acrylic acid) is present, it gets in the way of some of these primary particles and partially inhibits their surface recrystallization. Because other particles are able to aggregate, these molecules get trapped and therefore

PAGE 45

32 incorporated into the final particle. This is why a grainy morphology is observed in the TEM images; the primary particles just are not allowed to coalesce like they want to, thus the final particle achieves a composite morphology. The acidic nature of PAA is also a contributing factor to the incorporation of the molecules into the particles. In solution, the COOH groups easily loose their hydrogen, thus giving the molecule a negative charge. The positively charged iron ions will create an attraction and better facilitate the incorporation of the polymer into the iron oxide particle. Examination of the images in Figure 3-3 shows that the control reaction produces relatively uniform, spherical crystals of Fe 3 O 4 without a granular or fuzzy appearance. In contrast, examination of all samples with the PAA additive reflects a grainy, composite morphology. This gives a clear distinction between the control and PAA particles, but now the differences and trends among the PAA samples will be discussed. In examining the two trials that used a molecular weight of 2,100, it makes sense that the sample with a higher concentration of PAA did not convert during aging. The lower concentration sample that did convert appears to be very grainy as it is, suggesting that it had probably approached a maximum limit for PAA incorporation. Therefore, the sample with the greater quantity of PAA was likely overwhelmed with polymer and it was too much for the primary particles to overcome and aggregate together to form final particles. In this instance, the PAA additive worked as a complete inhibitor of the particle synthesis. It makes sense too because there would be so many molecules present in solution as compared to a higher molecular weight polymer at the same concentration. This is why we see later that this concentration, which did not work for a MW of 2,100, did work for

PAGE 46

33 higher molecular weights. If you look at the right-side picture of Figure 3-5, the bottom particle has a string-like structure emanating from it, perhaps showing the best visual evidence from all of the micrographs that the polymer is indeed incorporated into these particles. However, care must be taken in making these statements because the picture alone does not definitively confirm that this structure is the poly(acrylic acid) additive. Now turning attention to Figure 3-6 and the trials with a molecular weight of 6,000, we see that the lower concentration trial yielded much more particle-like structures. The images of the higher concentration sample show much less defined particles. This makes sense from observation of the synthesis because this second trial did not react to completion during the aging, leaving a significant amount of unconverted gel. However, some magnetic attraction was qualitatively observed, so the synthesis was not completely inhibited, and this seems to be what the images are showing. I will here say a word about when the reaction does go to completion. This statement would imply that 100% of the precipitated gel would be converted to particles of magnetite during aging. This is in fact not the case. Although it can be clearly observed when the reaction is significantly inhibited, a complete reaction will still inevitably contain some amounts of unconverted gel in addition to some unwanted byproducts, namely the goethite rods. The presence of the rods can even be seen in some of the TEM images in the figures above. For the purposes of this discussion, and even in the presentation of the results, a trial synthesis will be referred to as complete if most of the gel is converted and the resulting precipitate is black, the color characteristic of the magnetite product. Trials that did not convert very well or even at all had a brownish or greenish color.

PAGE 47

34 Observation of the images in Figure 3-7 shows that both trials with MW=15,000 produced granular composite particles. It is difficult to make any sweeping statements about whether or not there are any significant differences between these samples just based on their TEM micrographs. In contrast, it appears as though there is a difference between the two samples with MW=30,000. It seems that the trial with the higher concentration produced grainier particles. This makes sense because it would be expected that a larger concentration of polymer in solution would allow more of it to be incorporated into the particles, thus creating a more granular composite structure. But are there any observable trends based on changing the molecular weight? It has already been discussed how a large concentration of a low molecular weight polymer can inhibit synthesis, but there will also be a limitation when the molecular weight gets too high. In solution, a polymer chain takes a random coil conformation. As the molecular weight of this chain is increased, the hydrodynamic volume of the random coil becomes larger and larger. Once this gets to a large enough size, it becomes too difficult for the primary particles to aggregate together and pull this large molecule into the final particle. In this way, a higher molecular weight polymer becomes an effective inhibitor to the synthesis. Therefore, it was important in this research to try to find an interplay between two extremes. The higher concentration trial in the 30,000 molecular weight series seems that it may be approaching this upper limit. The particles appear as though they are getting grainy to the point that a definitive spherical particle is becoming hard to establish. Therefore, it seems that the trend in all of these PAA trials based on molecular weight considerations is that MW=2,100 produces somewhat grainy particles, and then the particles become more granular as molecular weight is increased to 6,000 and then

PAGE 48

35 15,000, but then they seem to begin to come to a limit at MW=30,000 where they start to loose their spherical particle morphology. Of course the polymer concentration is an important factor that can affect the synthesis at any of these molecular weights, but this seems to be the general trend for a reasonable concentration of PAA. X-Ray Diffraction (XRD) X-ray diffraction is a characterization technique that utilizes the Bragg Law, n=2dsin, where is the wavelength of the x-rays, d is the spacing in a certain crystalline plane, and is the angle in which the x-ray is diffracted. Depending on the d-spacing of the crystal lattice, incident x-rays are diffracted at different angles and then counted to derive a spectrum of intensity versus angle 2. Standards of data have been compiled and stored in large databases so that an experimental sample can be compared to a standard in order to identify the material and/or phase. For this analysis, a standard for magnetite was obtained and plotted along with the experimentally collected data. Sample preparation for XRD was relatively simple since the particles were already in a powder form. A small piece of double-sided tape was adhered to a glass slide. The particle powder was then deposited on top of the tape and spread out to cover the entire surface of the tape. This step is important in order to ensure that a large enough area will be exposed to the x-rays during data collection. A spatula was then used to press the powder onto the tape so that it would not be blown off. Incidentally, the PAA sample used in this analysis was the one using MW=6,000 and concentration 100 ug/mL. It is not necessary to analyze every particle sample with this technique because as long as the synthesis procedure was kept constant, the product should be the same so long as the reaction went to sufficient completion.

PAGE 49

36 Control (No Polymer Additive)02004006008001000120020253035404550556065702 Theta (degrees)Intensity (counts) Standard Control(311) ( 220 ) ( 222 ) (400)(440)(511)(422) Figure 3-9: X-ray diffraction spectrum of control particles. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction. Poly(Acrylic Acid) Additive010020030040050060070080020253035404550556065702 Theta (degrees)Intensity (counts) Standard PAA MW=6,000(311) ( 220 ) ( 222 ) (400)(440)(511)(422) Figure 3-10: X-ray diffraction spectrum of composite particles with PAA additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction. X-ray diffraction is a very useful technique used to characterize a materials crystal structure. It was used in this instance to confirm that the crystalline phase of magnetite is

PAGE 50

37 indeed present as the product of particle synthesis. Examining Figure 3-9 shows that the controls sample peaks match up very well with the peaks for the magnetite standard in both intensity and angle 2. This confirms that the control sample is Fe 3 O 4 as expected. Upon examination of the spectrum for particles with PAA we can see that the peaks again match up very well and confirm the presence of magnetite. However, it seems that the peaks for this sample are a bit broader than the peaks for the control. The reason for this is that because the polymer is present in the mineral, it causes strains in the crystalline lattice. These strains cause slightly different deflections of the x-rays and thus serve to create broader peaks. So in effect, this result further supports what the TEM images showthat a composite particle morphology has been created with the addition of poly(acrylic acid). X-ray diffraction was not performed on every sample because it would only be superfluous and unnecessary to do so. If the synthesis went to completion and formed particles, then this proves that those particles are indeed magnetite. Superconducting Quantum Interference Device Analysis (SQUID) This technique was used to create a hysteresis curve for each of the samples. The hysteresis curve yields many important magnetic quantities; you can determine saturation magnetization (M sat ), remnant magnetization (M rem ), and coercivity (H c ). It is important to quantify such values if an intended application of the particle system makes use of its magnetic moment. In addition, magnetic properties need to be characterized so that they can be compared to other magnetic systems. A full hysteresis loop was produced for the first few samples until it was realized that the particles show symmetric behavior. Therefore, subsequent samples were only run for half of a loop. The maximum applied field in each trial was 7 Tesla. The data are plotted as magnetization versus applied field.

PAGE 51

38 Sample preparation was a somewhat intensive procedure. Because the magnetometer is so sensitive, very accurate sample weights were necessary to reduce error. However, a 5% error was still given to each trial based on weight. The particle sample had to be carefully put into a gelcap and then tightly compacted with a kimwipe. 02468020406080100 T = 300 K M vs H, Particles with No Additive (Control) M (emu G / gram)B (T) Figure 3-11: Hysteresis curves for control particles. Table 3-2: Magnetic quantities determined from SQUID analysis of control particles Sample Saturation Magnetization, M sat (emu/gram) Remnant Magnetization, M rem (emu/gram) Coercivity, H c (Gauss) No Additive (Control) 78.6 1.5 33 No Additive (Control) 70.4 2.4 35

PAGE 52

39 02468020406080 MW=2,100; 130 ug/mL MW=6,000; 100 ug/mL MW=15,000; 100 ug/mL MW=15,000; 230 ug/mL MW=30,000; 100 ug/mL MW=30,000; 200 ug/mLT = 300 K M vs H, Particles with Poly(Acrylic Acid) M (emu G / gram)B (T) Figure 3-12: Hysteresis curves for PAA modified composite particles. Notice in the figure above that there is no plotted data for the trial that used MW=2,100 and a concentration of 220 ug/mL. This is because that trial did not convert during the aging process; instead, there was only unconverted gel remaining after the aging. There is also no curve for the trial with MW=6,000 and concentration 200 ug/mL. This is because the sample only converted a small amount during the aging process. Much of the final volume after synthesis was unconverted gel and thus a hysteresis curve of this sample would not have been representative of a composite particle system. It is important to stay consistent in presenting the data so only the samples where the reaction went to significant completion are presented here.

PAGE 53

40 Table 3-3: Magnetic quantities determined from SQUID analysis of PAA particles Sample Saturation Magnetization, M sat (emu/gram) Remnant Magnetization, M rem (emu/gram) Coercivity, H c (Gauss) MW=2,100 130 ug/mL 54.2 2.2 40 MW=6,000 100 ug/mL 57.5 4.1 60 MW=15,000 100 ug/mL 67.1 2.9 30 MW=15,000 230 ug/mL 61.3 5.7 62 MW=30,000 100 ug/mL 39.9 2.6 34 MW=30,000 200 ug/mL 46.2 3.6 55 A good amount of information can be obtained about the nature of a magnetic material just by the shape of its hysteresis curve. By examining the hysteresis curves for both the control and PAA samples, we can see that they both have the same general shape. This implies that the formation of composite particles does not significantly change magnetic behavior, but only the magnetic quantities. The particles show soft magnetic behavior, meaning that they are easily magnetized without having to apply a huge magnetic field. This is important for many applications, especially biomedical ones, because many techniques would be easier if a lower field could be applied. In magnetic resonance imaging, for example, current contrast agents are commonly magnetized with a 1.5 Tesla field. Examining the data shows that these particles begin to reach their magnetization limit well before this. The curves also show that the particles have a low coercivity. This means that only a small reverse in the magnetic field is necessary to bring the particles back to a zero magnetization. Coercivity values for the control and PAA particles range from 30 and 60 Gauss. The remnant magnetization

PAGE 54

41 values for both the control and PAA particles are in about the same ballpark as well. Their values range from about 1.0 to 5.5 emu/gram, which is small relative to their saturation magnetizations. (It should here be pointed out that the remnant magnetization is what may cause aggregation problems among the particles. The fact that the particles will retain some magnetic moment implies that they will be attracted to each other when they get in the vicinity of one another. This is an issue that needs to be addressed in future work so that it does not cause a problem in application, particularly one that would use the particle system in vivo.) The hysteresis curves and their corresponding tables show that the M sat and M rem values do not seem to be affected in a definitive manner when a polymer additive is present, but it is hard to say conclusively whether or not there is a significant effect based on this data. In any case, the most interesting values from this analysis and the focus of this discussion are the saturation magnetizations. The figures and their corresponding tables show that the control particles achieve a higher saturation magnetization than any of the PAA composite particles. This makes sense too since defects and impurities in the form of the polymer additive are being introduced to the particles. The polymer molecules have no magnetic properties, thus on a per weight basis they would be expected to degrade the overall magnetization. There has been evidence, however, that defects formed in a magnetic lattice can possess magnetization in and of themselves. Whether or not this is the case with the composite particles has yet to be determined, however if there is a contributing factor from these defects it is not very strong. Magnetite inherently gives a very strong signal because it has many unpaired electrons which create a strong magnetic moment; therefore, its intrinsic signal would likely drown out any small contributions from defects.

PAGE 55

42 Now that it has been established that the control particles achieve a higher M sat we will now explore the differences and trends among the PAA trials. Figure 3-12 shows that the two trials with MW=15,000 have the highest M sat and the two trials with MW=30,000 have the lowest. The other two samples of MW=2,100 and MW=6,000 have values for M sat in between the others. Since the three lowest values are the MW=2,100 and MW=30,000 trials, this could imply that these molecular weights are near the upper and lower limits of allowable molecule sizes for incorporation during synthesis. Anything too far outside these limits may cause complete inhibition and prevent any conversion to magnetite during the synthesis. This would mean that an intermediate molecular weight such as 15,000 may be ideal for incorporation and yield the best particles. This idea is indeed supported by the data because (as just mentioned) the MW=15,000 trials had the highest values for saturation magnetization. Although this may be a reasonable assumption, it is in no way a definitive conclusion. In fact, the basis of this reasoning is somewhat contradicted when trends in concentration are examined. The reasoning is based on what degree the additive inhibits the synthesis. It would therefore be expected that a higher concentration at the same molecular weight would further inhibit a synthesis and result in a lower saturation magnetization. This is indeed what is observed for the MW=15,000 series, but the exact opposite is observed for the MW=30,000 series. The best explanation for this is just simply that the synthesis is either not controlled as much as we would like because of the additive, or that the synthesis yield is not completely uniform. Therefore, it seems the best conclusion that can be made is only to qualitatively say that PAA composite particles will fall in the range of about 40-70 emu/gram. The bright side is that this is sufficient enough to say

PAGE 56

43 that since the particles are in this range they are still better than many of the other magnetic particle systems out there. Fluorescence Labeling Once the particles have been well characterized, it is important to demonstrate their functionality. Fluorescence microscopy combined with bioconjugation techniques is very useful for labeling desired targets. In the case of these composite particles, the labeling target was the poly(acrylic acid). Specifically, the carboxylic acid functional group was utilized to attach a fluorescent probe. The ability to conjugate this probe to the particles and observe its fluorescence demonstrates that the polymer is indeed functional and can therefore be conjugated to other probes. This is an important step toward producing these particles for a specific application. Procedure The protocol for this labeling procedure was taken from a study that labeled the carboxyl groups hanging off of a poly(methyl methacrylate) substrate [69]. This method was actually a modification of another protocol [70], but bioconjugation chemistries for fluorescence analysis in general have been well studied [71]. The fluorescein probe will not conjugate to the carboxylic acid group of the poly(acrylic acid) on its own. Instead, a linker is necessary. This linking molecule, EDAC, is a water-soluble carbodiimide that reacts with the amine group of the fluorescein and the COOH group of the PAA. It is by this reaction that the fluorescein is conjugated to the PAA. The actual procedure to do this is as follows. The EDAC and fluorescein were dissolved in a 100 mM phosphate buffer solution (pH 7) to make a concentration of 0.5 mM each. A few milligrams of PAA composite particles were put into a small vial and covered with the solution. The vial was then gently agitated in darkness for a 15-hour incubation period. When this was

PAGE 57

44 complete, the sample was washed with phosphate buffer four times to remove excess fluorescein molecules that did not conjugate. While suspended in phosphate buffer, the particles were micropippetted onto a glass microscope slide and a cover slip was placed on top. A mercury lamp and burner were used in conjunction with an optical microscope to perform the fluorescence microscopy. Digital images were taken with a microscope-mounted camera. Fluorescence microscopy It was necessary to again run a control with this protocol to ensure that there was no fluorescence associated with particles that did not contain the PAA. This is indeed the result shown in Figure 3-13 below. Figure 3-13: Transmission and fluorescence micrographs of control particles. The light transmission image on the left shows clumps of particles dispersed on the slide. The fluorescence micrograph on the right shows the fluorescence (or lack thereof) of that same area. Fluorescence microscopy is a very effective means to demonstrate the functionality of the poly(acrylic acid) in these composite particles. The specific labeling chemistry used in this research will only target and conjugate to carboxylic acid functional groups. This functionality is only afforded to the particles by the presence of PAA. Therefore, it would not be expected for the control sample to fluoresce in any way because no PAA

PAGE 58

45 Figure 3-14: Fluorescence images of PAA composite particles. The first two images come from the sample with MW=15,000 and concentration 100 ug/mL. The third image comes from sample MW=6,000 and concentration 100 ug/mL.

PAGE 59

46 exists in the particles. The images in Figure 3-13 confirm that this is true. However, the images in Figure 3-14 do show areas of fluorescence because these are images of composite particles with the PAA additive. It seems that only parts of the sample are fluorescing, and this is probably for a few reasons. The first and most likely explanation is that the particles are not completely uniform in their PAA content. There is evidence of this in the TEM images where we can see that some particles are grainier than others within the same image. Also, the PAA may not be completely surface accessible to the fluorescent probe. Some of the chains of PAA will be wedged into the particle too far and thus their functional groups are rendered useless. However, the fact that we do observe fluorescence in the particles at all demonstrates that composite particles are indeed present and their organic functionalities are intact and reactive.

PAGE 60

CHAPTER 4 PARTICLE SYNTHESIS WITH POLY-L-GLUTAMIC ACID ADDITIVE Introduction Poly-L-glutamic acid is a poly amino acid. Like poly(acrylic acid), it is an acidic polymer with a carboxylic acid functional group on every repeat unit. Its structure is shown in the figure below. Figure 4-1: Repeat unit structure of poly-L-glutamic acid. This polymer was chosen for use as an additive for two main reasons. The first is that it has the COOH functional group. This makes it similar to PAA in that it has the same functionality but with a different backbone structure. The second reason for choosing poly-L-glutamic acid (which will also be referred to as GLU) is because it has this backbone that is characteristic of all poly amino acids and proteins. Since a possible future goal with these particles is to incorporate specifically sequenced polypeptides, it is an important step to first try to incorporate a homopolymer of a poly amino acid. Therefore, poly-L-glutamic acid seemed like a fitting choice for this objective. The range of molecular weights explored with this additive was smaller than what was used for the PAA trials. The reason for this is that it is significantly more expensive to produce and consequently purchase poly amino acids. Therefore, only two molecular weights were used in synthesis trials with the GLU additive, however varying concentrations were also explored. 47

PAGE 61

48 This chapter will present the results of the GLU additive trials using the same characterization techniques utilized in the previous chapter. Transmission electron microscopy (TEM) will show the composite particle morphology, x-ray diffraction will confirm the presence of magnetite, and superconducting quantum interference device analysis (SQUID) will show the magnetic properties of the composite particles. Comparisons to control particles or PAA composite particles will be made where necessary. In addition, fluorescence labeling was performed on these particles because they have carboxylic functional groups. The presence of fluorescence demonstrates that the organic functionality of the composite particles remains intact. Materials Poly-L-glutamic acid sodium salt was obtained from Sigma in molecular weights of 7,500 and 13,600. The rest of the materials used for these trials are the same as those used for the PAA trials described in chapter 3. The following is a brief summary of those materials. Iron (II) sulfate heptahydrate (Aldrich Chemical) was dissolved in deionized water to make a solution with a final concentration of 0.255 M, which corresponds to an excess ion concentration of [Fe 2+ ] ex = 5 x 10 -3 A concentrated potassium hydroxide (KOH) solution was purchased from Acros Organics and diluted to a 0.5 M solution. Potassium nitrate, KNO 3 (Fisher Scientific), was dissolved in deionized water to make a solution with a final concentration of 2.0 M. The KOH and KNO 3 solutions were stored separately and kept as stock solutions. The FeSO 4 solution was remade just before each synthesis trial because it begins to precipitate out of solution after sitting for a few days. A pyrogallol (C 6 H 3 (OH) 3 ) and sodium hydroxide (NaOH) solution mixture was used as a deoxygenation solution. A 5.0 M NaOH solution was prepared, and then solid pyrogallol was added to create a 1.0 M concentration. This solution was placed into a three-necked

PAGE 62

49 flask that was subsequently sealed at all three necks with septa. A fluorescein-based probe was used for the fluorescence labeling study, specifically 5-(aminoacetamido) fluorescein. Another chemical, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide, or EDAC, was also necessary for the fluorescence study. Both were obtained from Molecular Probes, Inc. Methods The method for particle synthesisincluding the setup and separation procedureswere the same for the trials with poly-L-glutamic acid as they were for the PAA trials. In addition, the fluorescence labeling technique is also the same. A quick synopsis of these methods follows, but the most complete description can be found in the Methods section of chapter 3. The pyrogallol solution was poured into a three-necked flask. The necks were plugged with septa and then the septa were punctured through with 18-gauge septum needles. The needles from the two outside necks were then punctured through other septa that were sealing vialsone containing the FeSO 4 solution and one containing the KOH/KNO 3 solution. The needle in the middle neck of the flask served to deliver the nitrogen from the gas cylinder to the pyrogallol solution. A picture of this setup can be seen in Figure 3-2 of the previous chapter. For the actual synthesis, the FeSO 4 and KOH/KNO 3 solutions were purged with nitrogen for two hours. 8 mL of the FeSO 4 was then injected into 2 mL KOH/KNO 3 solution, precipitating the Fe(OH) 2 gel. The GLU additive was dissolved in the KOH/KNO 3 solution prior to the two-hour purging with nitrogen. This vial was then aged in a 90C oil bath for four hours to create the Fe 3 O 4 nanoparticles. The particles

PAGE 63

50 were separated out from contaminants and solution using a handheld magnet and then centrifugation. The particles were completely dried and stored as a powder. Table 4-1: Experimental design of trials with poly-L-glutamic acid. Molecular Weight Quantity Added (mg) Concentration of Polymer (ug/mL) 7,500 0.8 80 7,500 1.0 100 13,600 0.6 60 13,600 1.0 100 13,600 1.5 150 The table above shows the experimental trials performed using the GLU additive. Concentrations only ranged from 60 to 150 ug/mL because it seemed the reaction would only go to completion within these limits. It was during the optimization of this synthesis when it was found that the reaction would only seem to produce composite particles using these molecular weights when a concentration in this range was used. Therefore, any possible trends discussed in the results section will be based on these trials. Although only two molecular weights were explored, it is still sufficient enough to see that composite particles are indeed formed using the poly-L-glutamic acid additive. Characterization Results and Discussion Transmission Electron Microscopy (TEM) The first step toward characterizing the composite particles is performing TEM to show their composite morphology. A TEM grid was prepared for each sample by dispersing some particles in ethanol and then depositing them on a formvar coated copper grid using a micropipette. The grids were placed in an oven for a few minutes to help the particles better adhere to the formvar coating. The microscopy was performed at an accelerating voltage of 200 kilovolts.

PAGE 64

51 Figure 4-2: TEM micrographs of particles with poly-L-glutamic acid, MW = 7,500. The top two images are taken from the sample with a GLU concentration of 80 ug/mL. The bottom two images had a GLU concentration of 100 ug/mL. Examination of the images in the figure above reveals that the morphology indicative of composite particles is obtained. We can see the grainy texture of the particles created by the GLU additive disrupting the normal aggregation and recrystallization mechanism of the primary particles. Thus, the TEM analysis gives evidence that the polymer additive has indeed been incorporated into the magnetite particle. The images also show that the particles are in the size range of 150-300 nanometers. This size is small enough so that the particles can flow through blood capillaries and large enough so that they are not superparamagnetic. It is difficult to say whether or not there is a trend with these trials based on the images. The main reason is that there are only two samples at this molecular weight, and more samples would be needed to make a more conclusive statement about trends. However, it appears as though the particles with a concentration of 100 ug/mL are

PAGE 65

52 slightly more granular than those with a concentration of 80 ug/mL. This would make sense because a higher concentration would imply that more GLU additive is present and thus has a better chance of more incorporating into the particles. So even if it is too difficult to make a conclusive statement about the effect of concentration, it seems that these images at least support what would be expected to happen. And most importantly, these images demonstrate that the composite morphology is indeed obtained, which was the primary information desired of the TEM analysis in the first place. Figure 4-3: TEM micrographs of particles with poly-L-glutamic acid, MW = 13,600. The top two, middle two, and bottom two images have concentrations of 60 ug/mL, 100 ug/mL, and 150 ug/mL, respectively.

PAGE 66

53 The images in Figure 4-3 seem to show subsequently grainier particles as the concentration of polymer is increased. This is intuitively what would be expected to happen, as more GLU additive would be available for incorporation into the particles if its concentration were higher. However, it is hard to say whether or not this is the case for sure without quantitative data showing the amount of polymer actually incorporated into the particles. It is important to mention this because if we look at the trial with MW=13,600 and concentration 100 ug/mL, for example, it seems that the resulting particles are not as uniform in morphology as the other two samples in that figure. This could suggest that some particles have incorporated more of the GLU additive than others, thus making it difficult to say on a whole that this trial contains either more granular or less granular particles. Therefore, it is not conclusive that a trend is seen here, however the TEM images give strong support to the affirmative claim. The same should be said about a trend in molecular weight. Actually, it is hard to say even qualitatively that the polymer content is increased or decreased due to molecular weight based on the images in the figures above. And since only two molecular weights were explored, it makes it even harder to say what the effect might be. It is important, however, not to lose track of the most important evidence obtained from the TEM analysis: the nanocomposite particle morphology has indeed been obtained. X-Ray Diffraction (XRD) X-ray diffraction is a very useful technique used to determine the crystal structure of a material. It is used on these composite particles to confirm that the mineral phase is magnetite (Fe 3 O 4 ) and not a different iron oxide or derivative. This technique is appropriate for this application because the polymer additive is not crystalline so it is not expected to contribute anything to diffraction. Sample preparation was a straightforward

PAGE 67

54 process of adhering the sample powder to a glass slide via double-sided tape. XRD was performed on only one of the GLU samples because it is representative of all the GLU samples given that the reaction went to completion; only those samples where the reaction did indeed go to completion are presented in this chapter. Particles with Poly-L-Glutamic Acid010020030040050060070080090020253035404550556065702 Theta (degrees)Intensity (counts) Standard GLU MW=7,500(311)(220)(222)(400)(440)(511)(422) Figure 4-4: X-ray diffraction spectrum of composite particles with GLU additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction. As shown in the figure above, the experimental spectrum was plotted along with a standard for Fe 3 O 4 The peaks for the sample with the poly-L-glutamic acid additive match up with the peaks for the standard in both relative intensity and angle. This confirms that the crystalline phase of the composite particles is magnetite. There seems to be an extra small peak at about 28 degrees. This does not match up with the standard, however it does not match up with any of the high intensity peaks for other phases of iron oxide. There are two likely explanations. Perhaps the standard used for comparison did not express that plane strongly. However, the most probable cause for this peak is from the small amount of steel that may have been deposited onto the sample during

PAGE 68

55 preparation. A steel spatula was used to press the powder firmly onto the tape. The peaks on the spectrum are also somewhat broader than expected for pure magnetite. This is likely due to the lattice strains associated with the GLU additive. These strains will cause the x-rays to slightly diffract differently and thus broaden the peaks. Superconducting Quantum Interference Device Analysis (SQUID) This analysis measures the magnetic properties of the composite particles by creating a hysteresis curve. Since these samples show mirrored behavior at positive and negative magnetic fields, only half of a full hysteresis loop was obtained for all of the samples. The curve is generated by slowly applying a magnetic field to magnetize the sample, and then slowly removing and reversing the magnetic field until the magnetization is zero. The maximum applied field was 7 Tesla. Samples were prepared by packing the particle powder into a gelcap, and the curves were given a 5% error based on weight. 02468020406080100 T = 300 K MW=7,500; 100 ug/mL MW=7,500; 80 ug/mL MW=13,600; 150 ug/mL MW=13,600; 100 ug/mL MW=13,600; 60 ug/mL M vs H, Particles with Poly-L-Glutamic Acid M (emu G / gram)B (T) Figure 4-5: Hysteresis curves for GLU modified composite particles.

PAGE 69

56 The values for coercivity and remnant magnetization are not discernable on the plot above because it is not zoomed in on the origin. However, the zoomed-in plot is not presented here because there are so many data points in that area that it becomes confusing to look at. Instead, the values for all of the critical magnetic quantities for each sample are summarized in the following table. Table 4-2: Magnetic quantities determined from SQUID analysis of GLU particles Sample Saturation Magnetization, M sat (emu/gram) Remnant Magnetization, M rem (emu/gram) Coercivity, H c (Gauss) MW=7,500 80 ug/mL 48.3 2.3 46 MW=7,500 100 ug/mL 60.0 2.1 41 MW=13,600 60 ug/mL 33.1 1.4 50 MW=13,600 100 ug/mL 34.1 2.7 50 MW=13,600 150 ug/mL 36.4 1.0 33 It can be seen from this table that the values for remnant magnetization and coercivity are approximately in the same range of each other for each of the trials. This makes sense from the plot because all five of the curves have the same general shape. The magnetization increases sharply and then begins to level out toward a maximum value as the field strength is increased. Remnant magnetization results from the magnetic crystals maintaining some of their net magnetic moment after the field has been removed. This property may lead to some particle aggregation problems since they will create an effect on each other. The magnetic property that varies most among the trials is the saturation magnetization. We can see from Figure 4-5 that the three trials with MW=13,600 are

PAGE 70

57 within the error range of each other, effectively showing that there is no significant difference between them. This could imply many things. It could mean that the concentration of polymer at this molecular weight has very little or no effect on the magnetic properties of the particles. It could also mean that in each of the trials the particles incorporated the same amount of polymer additive. I think that the latter possibility is on the right track becauseas it was noted in the TEM discussionit is difficult to say whether or not one trial produced grainier particles as a whole than another. It is possible that a limit on polymer concentration was reached early (i.e. 60 ug/mL), where any extra polymer in solution would just not be incorporated during synthesis. But it would seem likely that an excess of additive would better inhibit the synthesis and perhaps prevent significant conversion during the aging process. It is impossible to say what the reason is for sure, so further experimentation would be necessary to elucidate their differencesor lack thereof. It is clear from the plot that there are significant differences between the MW=13,600 trials and the MW=7,500 trials. First of all, if we go back to chapter 3 and look at the control trials, we can see that the controls still have higher values for M sat than any of the GLU trials. However, the MW=7,500 trials both have a higher M sat than the MW=13,600 samples, and the one with concentration 100 ug/mL is higher than the one with concentration 80 ug/mL. This does not make intuitive sense. It would be expected that the higher concentration sample would have more incorporated additive and thus have a reduced overall saturation magnetization. A similar result was seen with the PAA trials in chapter 3, however the more intuitive result was also observed. What this means is that a definitive trend can not be speculated based on this data for the MW=7,500 trials.

PAGE 71

58 However, it does make sense that the MW=13,600 trials all have lower values because the additive is a larger molecule. A larger molecule would be more inhibitory to synthesis, likely creating greater space between primary particles during the aging, thereby creating particles possessing a lower saturation magnetization. This result is indeed observed and is really the most conclusive evidence for particle differences based on this analysis. Fluorescence Labeling This method of analysis is being used in order to demonstrate the reactivity and functionality of the carboxylic acid groups of the poly-L-glutamic acid additive. The procedure for labeling the GLU modified composite particles is the same procedure described in chapter 3 for labeling the PAA composite particles. Therefore, only a brief outline of that protocol will be presented here. The fluorescent probe (5-(aminoacetamido) fluorescein) is conjugated to the poly-L-glutamic acid chains via a linker molecule, EDAC. Both of these materials were dissolved in phosphate buffer to create a 0.5 mM concentrated solution. Particles were then submersed in this solution and gently agitated overnight to allow coupling. After four subsequent washings with phosphate buffer, the particles were mounted onto a glass slide and covered with a cover slip. A mercury lamp and burner were used with an optical microscope to perform fluorescent imaging. Pictures were captured with a scope-mounted digital camera. The labeling protocol used for this analysis is specifically targeted to conjugate to carboxylic acid groups only. Since COOH groups are the only type of functional group that exist on the particles (afforded by the poly-L-glutamic acid), then any fluorescence indicates that these groups are present and reactive. Figure 4-6 supports that this is the case because we can see that there are areas of fluorescence present in the images.

PAGE 72

59 Figure 4-6: Fluorescence images of GLU composite particles. Both images come from the sample with MW=13,600 and concentration 150 ug/mL. The fluorescence in these images is clearly not blinding by any means, nor does it seem to be present on all areas of the particle clumps. The reason for lack of complete ubiquity is that the polymer is likely not sufficiently surface accessible to the fluorescent probe on all particles; therefore the probe would not have been able to get to all or any of the additive and conjugate to it. It also makes sense that the fluorescence is not very strong becauseas can be seen in the TEM imagesthe particles have a very high mineral loading, i.e., the quantity of polymer additive present in each particle is small. If

PAGE 73

60 the particles had a greater quantity of polymer, like a poly-L-glutamic acid coating, for example, then a much brighter and widespread fluorescence would be expected. It is important to note that all observed fluorescence can only be from GLU incorporated in particles because any excess GLU that may have been left over from synthesis would have been washed away during both the separation process and the sample preparation. So, to summarize, the observation of fluorescence does indeed demonstrate the presence and reactivity of the poly-L-glutamic acid within the magnetite particles.

PAGE 74

CHAPTER 5 PARTICLE SYNTHESIS WITH POLY-L-LYSINE ADDITIVE Introduction Poly-L-lysine is an amino acid homopolymer with a primary amine as its functional group. Unlike the polymers used in the previous two chapters, poly-L-lysine is basic in solution. The structure of its repeat unit is shown in the figure below. Figure 5-1: Repeat unit structure of poly-L-lysine. This polymer was chosen for use as an additive primarily because of its functionality. The majority of bioconjugation protocols out there seem to make use of an amino group in order to attach drugs, antibodies, fluorescent molecules, or other probes. Therefore, the reasoning was simply that the particles would become more useful for conjugation purposes if they contained a free amine. In addition, the fact that poly-L-lysine is a poly amino acid lends to the future goal of incorporating sequenced polypeptides. Only one molecular weight of poly-L-lysine (which will be referred to also as LYS) was investigated for reasons that will become apparent in the discussion later in this chapter. However, for the trials that were performed, the samples were characterized with the same techniques used in the previous two chapters. The only exception to this is the fluorescence labeling study because the protocol used for the previous two polymer additives targeted the carboxylic acid functionality, but the LYS additive does not possess this trait. So, the particles were imaged with transmission electron microscopy 61

PAGE 75

62 (TEM), measured for magnetic quantities using a superconducting quantum interference device (SQUID), and the Fe 3 O 4 phase of the particles was confirmed using x-ray diffraction (XRD). Similarities to control and/or composite particle samples will be discussed where necessary. Materials Poly-L-lysine sodium salt was obtained from Sigma in a molecular weight of 27,000. The other materials used for the synthesis were the same as those described in chapters 3 and 4, but for the sake of completeness a brief summary of those materials is again listed here. Iron (II) sulfate heptahydrate (Aldrich Chemical) was dissolved in deionized water to make a solution with a final concentration of 0.255 M, which corresponds to an excess ion concentration of [Fe 2+ ] ex = 5 x 10 -3 A concentrated potassium hydroxide (KOH) solution was purchased from Acros Organics and diluted to a 0.5 M solution. Potassium nitrate, KNO 3 (Fisher Scientific), was dissolved in deionized water to make a solution with a final concentration of 2.0 M. The KOH and KNO 3 solutions were stored separately and kept as stock solutions. The FeSO 4 solution was remade just before each synthesis trial because it begins to precipitate out of solution after sitting for a few days. A pyrogallol (C 6 H 3 (OH) 3 ) and sodium hydroxide (NaOH) solution mixture was used as a deoxygenation solution necessary to remove trace amounts of oxygen and carbon dioxide from the nitrogen gas used for purging. A 5.0 M NaOH solution was prepared, and then pyrogallol powder was added to create a 1.0 M concentration. This solution was placed into a three-necked flask that was subsequently sealed at all three necks with septa. The setup and usage of these materials is described in the following section.

PAGE 76

63 Methods Just as the materials for these trials were primarily the same as those for the previous two chapters, so too were the methods for particle synthesis. Refer to chapter 3 for the most complete description of the setup, synthesis, and separation procedures because only a short description of these methods will be given here. The pyrogallol solution was poured into a three-necked flask that was then plugged with septa. The septa were punctured with 18-gauge septum needles and then the needles from the two outside necks were punctured through other septa that were sealing vialsone containing FeSO 4 solution and one containing an equal mixture of KOH and KNO 3 solution (1 mL of each). The needle in the middle neck of the flask served to deliver the nitrogen from the gas cylinder to the pyrogallol solution. A picture of this setup is in Figure 3-2 of chapter 3. For the actual synthesis, the FeSO 4 and KOH/KNO 3 solutions were purged with nitrogen for two hours. 8 mL of the FeSO 4 was then injected into the 2 mL KOH/KNO 3 solution, precipitating the Fe(OH) 2 gel. The poly-L-lysine additive was dissolved in the KOH/KNO 3 solution prior to the two-hour purging period. This vial was then aged in a 90C oil bath for four hours to create the Fe 3 O 4 nanoparticles. The particles were separated from contaminants using a handheld magnet and then centrifugation, using ethanol as a wash. The particles were completely dried in air and stored as a powder. The following table shows the experimental trials that were performed using the poly-L-lysine. Only two trials were analyzed with this additive becauseas we will see in the resultsparticles did not seem to form with the desired morphology. However, the results of these trials still give useful insight to the synthesis process and are therefore

PAGE 77

64 worthwhile to analyze. A more in-depth discussion will come with the presentation of the results in the following section. Table 5-1: Experimental trials with poly-L-lysine. Molecular Weight Quantity Added (mg) Concentration of Polymer (ug/mL) 27,000 1.0 100 27,000 2.5 250 Characterization Results and Discussion Transmission Electron Microscopy (TEM) TEM was used to obtain images of the particles and show their morphology. To prepare a sample, some particle powder was dispersed in ethanol by sonication, then pippetted onto a formvar coated copper grid. The grids were placed in an oven for a few minutes to help the particles better adhere to the grids. The microscope was operated at an accelerating voltage of 200 kilovolts. Figure 5-2: TEM micrographs of particles with poly-L-lysine, MW = 27,000. The top two images are taken from the sample with a LYS concentration of 100 ug/mL. The bottom two images had a LYS concentration of 250 ug/mL.

PAGE 78

65 The images above seem to indicate that the LYS additive has not been incorporated into the particles because the composite particle morphology is not observed. They are well-defined particles without a grainy or fuzzy-looking texture (for the most part), a morphology that is reminiscent of the control particles. In fact, the two images on the right were the most granular-like particles that could be found on the TEM grid for each of the two trials. The pictures on the left are the most representative of the entire sample. So the question arises: why did this occur? The answer probably lies in the fact that poly-L-lysine is basic in solution. Recall that in the case of the PAA and GLU additives the polymer chains had a negative charge due to the acidic carboxyl group. This negative charge likely helped attract the additive molecules to the positively charged iron ions during the particle synthesis. Therefore, the opposite probably occurred with the LYS additive, which becomes positively charged in solution. The particle constituents would not have attracted the poly-L-lysine and thus made it very difficult for the additive to be incorporated and create composite particles. Instead, the result is the same as (or at least very similar to) the control samples. Further analysis, however, is key in discerning whether or not there actually are differences from the control particles. From this analysis it seems as though they are one in the same. It should also be noted that there is clearly no discernable trend in the effect of concentration because it seems that the additive has no effect on the synthesis at all. X-Ray Diffraction (XRD) This technique was used for the same purpose as in the other two experimental chapters: to confirm that the crystalline phase of the particles is magnetite (Fe 3 O 4 ). Only one of the samples was used for this analysis because it would be superfluous to do both

PAGE 79

66 of them. The sample was prepared by spreading some of the particle powder onto a piece of double-sided tape that had been adhered to a glass slide. Particles with Poly-L-Lysine Additive0100200300400500600700800900100020253035404550556065702 Theta (degrees)Intensity (counts) Standard LYS MW=27,000(311) ( 220 ) ( 222 ) (400)(440)(511)(422) Figure 5-3: X-ray diffraction spectrum of composite particles with LYS additive. The data for the magnetite standard is also plotted for comparative purposes. The peaks are labeled with their corresponding planes of diffraction. The x-ray diffraction spectrum in the figure above shows that the peaks for the particle sample do indeed match up with the peaks for the magnetite standard, thus confirming that the particles are composed of Fe 3 O 4 If we go back and compare this spectrum to that of the PAA and GLU composite particles from the previous chapters, we can see that these peaks do not seem to be quite as broad. This would correlate with the TEM observation that the particles are not really granular. The lack of a polymer additive in the particles would not create an added lattice strain, and thus the x-rays would not be deflected differently and the resulting peaks would be sharper, as observed. This spectrum again seems to show an extra small peak at an angle of about 28 degrees. This can probably be attributed again to the use of a steel spatula in spreading

PAGE 80

67 the particles on the tape during sample preparation. If this is not the root of the cause, then at least it seems to be a constant that all of the XRD spectra are showing. Superconducting Quantum Interference Device Analysis (SQUID) The TEM analysis yielded an interesting result in that the particles synthesized with the LYS additive looked very much like control particles. The SQUID analysis will give further insight to any differences that these particles may have with the controls. This analysis is used to measure magnetic quantities by plotting a hysteresis curve. Magnetization is measured as a function of an applied magnetic field to create the plot. The maximum applied field was 7 Tesla. Samples were prepared by packing the particle powder into a gelcap. A 5% error was allotted to the data based on weight uncertainty. 02468020406080100 MW=27,000; 100 ug/mL MW=27,000; 250 ug/mLT = 300 K M vs H, Particles with Poly-L-Lysine M (emu G / gram)B (T) Figure 5-4: Hysteresis curves for particles with LYS additive. The hysteresis curves for the particles with the poly-L-lysine additive show the same characteristic shape that has been observed for all of the other samples. They show

PAGE 81

68 soft magnetic behavior with a low coercivity and remnant magnetization. The table below shows the actual values for the magnetic quantities. Table 5-2: Magnetic quantities determined from SQUID analysis of LYS particles Sample Saturation Magnetization, M sat (emu/gram) Remnant Magnetization, M rem (emu/gram) Coercivity, H c (Gauss) MW=27,000 100 ug/mL 76.8 1.1 16 MW=27,000 250 ug/mL 71.9 1.5 20 This table also reflects the high saturation magnetization achieved by these samples. These are the most interesting quantities because they correlate so well with the values obtained for the control samples. To demonstrate this, the figure below shows the control samples plotted along with the LYS samples. 02468020406080100 Control Control LYS MW=27,000; 250 ug/mL LYS MW=27,000; 100 ug/mLT = 300 K M (emu G / gram)B (T) Figure 5-5: Hysteresis curves for both LYS and control particles

PAGE 82

69 All four of the samples shown in Figure 5-5 are within the error of the other three, effectively showing that there is no significant difference between any of them. This result further supports what was suspected by the TEM analysis. The particles with the poly-L-lysine additive did not form composite particles, but rather formed just like a control sample. Again, this is likely due to the nature of the poly-L-lysine in solution. Its charge electrostatically repels it from the magnetite particles during synthesis, thus preventing it from inhibiting the reaction or incorporating into the particles. It should now be clear why only two samples with the LYS additive were analyzed. Since the additive does not incorporate and form composite particles, it would have been ultimately useless to try to explore a multitude of polymer concentrations and molecular weights; clearly, no trends would be present since every trial would likely form control-like particles. It should also be clear now why fluorescence microscopy was not performed. It would have required the purchase of a different probe that would target the amino functionality, but since the polymer is not present in the particles, there would be nothing for the probe to conjugate to and consequently there would be no observable particle fluorescence.

PAGE 83

CHAPTER 6 CONCLUSIONS AND FUTURE WORK This study of the effect of various polymer additives on the solution synthesis of magnetite nanoparticles has demonstrated many important findings. First and foremost, it has been shown that composite particles with organic functionality can indeed be formed when an appropriate polymer additive is used at a reasonable concentration and molecular weight. It was found that the two additives with the acidic carboxyl groups worked well to achieve the composite morphology. During the aging process of the synthesis, these polymer chains are attracted to the iron ions that are precipitating the crystalline magnetite particles from the amorphous intermediate. When the primary particles aggregate and undergo a surface recrystallization phenomenon, the polymer additive becomes permanently incorporated into the final particle. The composite particles were characterized to confirm many things. First, transmission electron microscopy has shown that the composite particles have a grainy morphology, indicating the presence of a polymer additive. It was further shown that this additive is not only present, but also functional. This was demonstrated by the conjugation of a fluorescent probe to the functional carboxyl groups of the incorporated polymer. In addition, it can be concluded that the composite particles are indeed the same material as the control, Fe 3 O 4 thereby showing that the additive did not have an effect on what iron oxide phase was formed. Furthermore, it seems that some general trends were established based on the usage of varying molecular weights and polymer concentrations. Care must be taken in making the following statements because they may 70

PAGE 84

71 not hold universally true. In fact, even this study found some disagreements with these general trends. What seemed to be observed was that as the additive concentration was increased, so did the degree of granular morphology, and consequently there was a loss in saturation magnetization. This all makes intuitive sense too because the greater the concentration, the more polymer there is available for incorporation, thus creating grainer particles with more organic content, thereby lowering the magnetization per unit weight. A similar line of reasoning would hold true for increases in molecular weight, however the results make it much harder to speculate this trend. This study has also shown evidence that there are upper and lower limits to both the concentration and molecular weight of the additive. Too high a concentration or molecular weight can completely inhibit the synthesis and prevent particles from forming at all. Too low a concentration or molecular weight and the majority of particles will simply form as a control with no incorporation of additive. Future work on an increased number of different molecular weights and concentrations would help solidify the observation of the general trends and hopefully delineate some upper and lower limits for the different polymer additives. It has also been demonstrated that the composite particles will not form with just any polymer additive. The trials with poly-L-lysine have shown this. Although the molecular weight and concentration was on par with the other additives, none of the polymer was incorporated into the particles. The characterization methods confirm that this is true. The TEM images look like those of control particles, and the SQUID measurements show no significant difference from the control sample data. This all points to the fact that no polymer was incorporated, likely due to the positive charge obtained by the poly-L-lysine chains in solution. This works to repel the additive from

PAGE 85

72 the nuclei during synthesis, therefore a composite particle is not created. A future experiment could be performed where the pH is adjusted during synthesis so that the charge can be flipped and hopefully create an attractive force that will allow the additive to be incorporated. Universally, a few things can be concluded. The first is that all trials produce magnetite as the crystalline phase. This has been confirmed by x-ray diffraction. This research has also shown that the shape of the hysteresis curves remains the same for each sample. The magnetic properties are characterized by a low coercivity and a high saturation magnetization. Although the composite particles demonstrate a lower overall M sat as compared to the controls, they are still stronger than many other magnetic particle systems. Finally, this study has shown that the particles created are indeed small enough for biomedical applications. Although the size distribution is not as monodisperse as desired, even the largest particles get up to only about 500 nanometers; this is still plenty small considering that the inner diameter of capillaries are typically about 5-10 microns. However, it is still important to work at creating a more monodisperse colloid of composite particles. In addition, the size distribution needs to be better characterized. This would be best achieved by the use of a light scattering technique. So where do we go from here? It would be very useful to quantify the amount of polymer actually present in the composite particles. This is especially important if the particle system should ever need to be FDA approved. Thermogravimetric analysis (TGA) should be a convenient method to obtain this data, however a complete and thorough separation process must be established. If the remaining unconverted gel is not

PAGE 86

73 completely removed, it would be difficult to distinguish in a TGA plot which weight loss can be attributed to which material. The presence of magnetic remanence may also cause a problem that needs more work to correct. The particles maintain a small magnetization even when there is no applied field. This will cause the particles to be attracted to each other and therefore aggregate. Use of various deflocculates need to be explored to help prevent or eliminate this problem. Too much aggregation could have disastrous effects if the particles are used in the bloodstream because they could potentially get to the point where they block a blood vessel. This is clearly a danger that needs to be remedied by further work. The ultimate goal of this work is to eventually incorporate specifically sequenced polypeptides that can target specific cell receptors or other molecules. If a composite morphology can be successfully obtained using such an additive, then the particles would then of course need to be fully characterized. Of the greatest interest would be to demonstrate their specific targeting capability. This may be difficult because amino acid sequences often have both spatial and chemical functionality. Therefore, the sequence would have to be sufficiently accessible on the surface of the particles in order for its targeting capabilities to work. If this did not work out, then it is very conceivable that the same sequence could just be conjugated to the functional particles that have been created in this study. In this way, the specifically sequenced peptide would not feel the effects of being incorporated in the crystalline matrix. All in all, this research has shown some very interesting results. A new particle system has been found in which a highly magnetic particle possesses organic functionality. This functionality affords the particles many possible applications,

PAGE 87

74 including biomedical and environmental targeting and retrieval capabilities. There is a myriad of directions that could be taken with this system, but this research serves to lay the foundation for these magnetic nanocomposite particles.

PAGE 88

LIST OF REFERENCES 1. Frankel, R.B. and R.P. Blakemore, Iron Biominerals. 1991, New York: Plenum Press. 2. Kirschvink, J.L., Magnetite Biomineralization and Geomagnetic Sensitivity in Higher Animals-an Update and Recommendations for Future Study. Bioelectromagnetics, 1989. 10(3): p. 239-259. 3. Hautot, D., Q. Pankhurst, N. Khan, and J. Dobson, Preliminary Evaluation of Nanoscale Biogenic Magnetite in Alzheimer's Disease Brain Tissue. Proceedings of the Royal Society of London Series B-Biological Sciences, 2003. 270: p. S62-S64. 4. Frankel, R.B. and D.A. Bazylinski, Magnetotaxis and Magnetic Particles in Bacteria. Hyperfine Interactions, 1994. 90(1-4): p. 135-142. 5. Lowenstam, H.A. and S. Weiner, On Biomineralization. 1989, New York: Oxford University Press. 6. Mann, S. and J.P. Hannington, Formation of Iron Oxides in Unilamellar Vesicles. J. Colloid & Interface Science, 1988. 122:2: p. 326-335. 7. Gorby, Y.A., T.J. Beveridge, and R.P. Blakemore, Characterization of the Bacterial Magnetosome Membrane. Journal of Bacteriology, 1988. 170(2): p. 834-841. 8. Bazylinski, D.A., Structure and Function of the Bacterial Magnetosome. Asm News, 1995. 61(7): p. 337-343. 9. Bazylinski, D.A., Controlled Biomineralization of Magnetic Minerals by Magnetotactic Bacteria. Chemical Geology, 1996. 132(1-4): p. 191-198. 10. Bazylinski, D.A., B.R. Heywood, S. Mann, and R.B. Frankel, Fe3O4 and Fe3S4 in a Bacterium. Nature, 1993. 366(6452): p. 218-218. 11. Devouard, B., M. Posfai, X. Hua, D.A. Bazylinski, R.B. Frankel, and P.R. Buseck, Magnetite From Magnetotactic Bacteria: Size Distributions and Twinning. American Mineralogist, 1998. 83(11-12): p. 1387-1398. 12. Frankel, R.B., D.A. Bazylinski, and D. Schuler, Biomineralization of Magnetic Iron Minerals in Bacteria. Supramolecular Science, 1998. 5(3-4): p. 383-390. 75

PAGE 89

76 13. Frankel, R.B., J.P. Zhang, and D.A. Bazylinski, Single Magnetic Domains in Magnetotactic Bacteria. Journal of Geophysical Research-Solid Earth, 1998. 103(B12): p. 30601-30604. 14. Majhi, P., B. Devouard, M. Posfai, X. Hua, D.A. Bazylinski, R.B. Frankel, and P.R. Buseck, Structural Features of Magnetite From Magnetotactic Bacteria. Meteoritics & Planetary Science, 1997. 32(4): p. A83-A84. 15. Meldrum, F.C., S. Mann, B.R. Heywood, R.B. Frankel, and D.A. Bazylinski, Electron Microscopy Study of Magnetosomes in a Cultured Coccoid Magnetotactic Bacterium. Proceedings of the Royal Society of London Series B-Biological Sciences, 1993. 251(1332): p. 231-236. 16. Proksch, R.B., T.E. Schaffer, B.M. Moskowitz, E.D. Dahlberg, D.A. Bazylinski, and R.B. Frankel, Magnetic Force Microscopy of the Submicron Magnetic Assembly in a Magnetotactic Bacterium. Applied Physics Letters, 1995. 66(19): p. 2582-2584. 17. Germann, W.J. and C.L. Stanfield, Principles of Human Physiology. 2002: Pearson Education, Inc. 18. Spaldin, N., Magnetic Materials: Fundamentals and Device Applications. 2003, Cambridge: Cambridge University Press. 19. Hummel, R.E., Electronic Properties of Materials. Third ed. 2001, New York: Springer-Verlag. 20. Kronmuller, H. and M. Fahnle, Micromagnetism and the Microstructure of Ferromagnetic Solids. 2003, Cambridge: Cambridge University Press. 21. Mohn, P., Magnetism in the Solid State: An Introduction. 2003, New York: Springer-Verlag. 22. Richerson, D.W., Modern Ceramic Engineering. Second ed. 1992, New York: Marcel Dekker, Inc. 23. Matijevic, E., Useful Applications of Monodispersed Colloids Have Come of Age. Abstracts of Papers of the American Chemical Society, 1997. 214: p. 9-24. 24. Afremov, L.L. and A.V. Panov, Magnetic States and Hysteresis Properties of Small Magnetite Particles. Fizika Metallov I Metallovedenie, 1998. 86(3): p. 65-73. 25. Muxworthy, A.R. and W. Williams, Micromagnetic Calculation of Hysteresis as a Function of Temperature in Pseudo-Single Domain Magnetite. Geophysical Research Letters, 1999. 26(8): p. 1065-1068. 26. Ozdemir, O., Coercive Force of Single Crystals of Magnetite at Low Temperatures. Geophysical Journal International, 2000. 141(2): p. 351-356.

PAGE 90

77 27. Roger, J., J.N. Pons, R. Massart, A. Halbreich, and J.C. Bacri, Some Biomedical Applications of Ferrofluids. European Physical Journal-Applied Physics, 1999. 5(3): p. 321-325. 28. Montagne, F., O. Mondain-Monval, C. Pichot, H. Mozzanega, and A. Elaissari, Preparation and Characterization of Narrow Sized (o/w) Magnetic Emulsion. Journal of Magnetism and Magnetic Materials, 2002. 250(1-3): p. 302-312. 29. Fu, L., V.P. Dravid, and D.L. Johnson, Self-assembled (SA) Bilayer Molecular Coating on Magnetic Nanoparticles. Applied Surface Science, 2001. 181(1-2): p. 173-178. 30. Ratner, B.D., A.S. Hoffman, F.J. Schoen, and J.E. Lemons, Biomaterials Science: An Introduction to Materials in Medicine. 1996, San Diego: Academic Press. 31. Jordan, A., R. Scholz, P. Wust, H. Fahling, and R. Felix, Magnetic Fluid Hyperthermia (MFH): Cancer Treatment with AC Magnetic Field Induced Excitation of Biocompatible Superparamagnetic Nanoparticles. Journal of Magnetism and Magnetic Materials, 1999. 201: p. 413-419. 32. Hsieh, D.S., R. Langer, and J. Folkman, Magnetic Modulation of Release of Macromolecules From Polymers. Proceedings of the National Academy of Science, 1989. 78: p. 1863-1867. 33. Mornet, S., F. Grasset, J. Portier, and E. Duguet, Maghemite@Silica Nanoparticles for Biological Applications. European Cells and Materials, 2002. 3(2): p. 110-113. 34. Gruttner, C. and J. Teller, New Types of Silica-Fortified Magnetic Nanoparticles as Tools for Molecular Biology Applications. Journal of Magnetism and Magnetic Materials, 1999. 194(1-3): p. 8-15. 35. Imshennik, V.K., I.P. Suzdalev, O.N. Stavinskaya, N.I. Shklovskaya, V. Schunemann, A.X. Trautwein, and H. Winkler, Preparation and Characterization of Porous Carbon Loaded with Iron Particles: A Possible Magnetic Carrier of Medical Drugs. Microporous Materials, 1997. 10(4-6): p. 225-230. 36. Rettenmaier, M., J. Stratton, P. Disaia, M. Berman, A. Senyei, and K. Widder, Treatment of a Syngeneic Rat-Tumor with Magnetically Responsive Albumin Microspheres Labeled with Adriamycin or Protein-A. Gynecologic Oncology, 1985. 20(2): p. 268-268. 37. Rettenmaier, M.A., J. Stratton, M. Berman, A. Senyei, and K. Widder, D. White, and P. Disaia, Treatment of a Syngeneic Rat-Tumor with Magnetically Responsive Albumin Microspheres Labeled with Doxorubicin or Protein-A. Gynecologic Oncology, 1987. 27(1): p. 34-43. 38. Senyei, A., K. Widder, and G. Czerlinski, Magnetic Guidance of Drug-Carrying Microspheres. Journal of Applied Physics, 1978. 49(6): p. 3578-3583.

PAGE 91

78 39. Widder, K., G. Flouret, and A. Senyei, Magnetic MicrospheresSynthesis of a Novel Parenteral Drug Carrier. Journal of Pharmaceutical Sciences, 1979. 68(1): p. 79-82. 40. Arshady, R., Microspheres for Biomedical ApplicationsPreparation of Reactive and Labeled Microspheres. Biomaterials, 1993. 14(1): p. 5-15. 41. Driscoll, C.F., R.M. Morris, A. Senyei, K. Widder, and G.S. Heller, Magnetic Targeting of Microshperes in Blood-Flow. Microvascular Research, 1984. 27(3): p. 353-369. 42. Fruitwala, M.A. and S. N.M., Site-Specific Drug Targeting with Fluorouracil Microspheres. Drug Delivery, 1996. 3: p. 5-8. 43. Gupta, P.K. and C.T. Hung, Albumin Microspheres. V. Evaluation of Parameters Controlling the Efficacy of Magnetic Microspheres in the Targeted Delivery of Adriamycin in Rats. International Journal of Pharmaceutics, 1990. 59: p. 57-67. 44. Hassan, E.E., R.C. Parish, and J.M. Gallo, Optimized Formulation of Magnetic Chitosan Microspheres Containing the Anticancer Agent, Oxantrazole. Pharmaceutical Research, 1992. 9(3): p. 390-397. 45. Mosbach, K. and U. Schroder, Preparation and Application of Magnetic Polymers For Targeting of Drugs. Febs Letters, 1979. 102(1): p. 112-116. 46. Pulfer, S.K. and J.M. Gallo, Enhanced Brain Tumor Selectivity of Cationic Magnetic Polysaccharide Microspheres. Journal of Drug Targeting, 1998. 6(3): p. 215-227. 47. Widder, K. and A. Senyei, Magnetic Microspheres: A Vehicle for Selective Drug Targeting. Pharmaceutical Therapeutics, 1983. 20: p. 377-395. 48. Weishaupt, D., D.Kochli, V, Marincek, B, How does MRI work? An Introduction to the Physics and Function of Magnetic Resonance Imaging. 2003, New York: Springer-Verlag. 49. Jung, C.W. and P. Jacobs, Physical and Chemical Properties of Superparamagnetic Iron-Oxide MR Contrast AgentsFerumoxides, Ferumoxtran, Ferumoxsil. Magnetic Resonance Imaging, 1995. 13(5): p. 661-674. 50. Bachgansmo, T., Ferrimagnetic Susceptibility Contrast Agents. Acta Radiologica, 1993. 34: p. 1-13. 51. Tiefenauer, L.X., G. Kuhne, and R.Y. Andres, Antibody Magnetite Nanoparticles: In-Vitro Characterization of a Potential Tumor-Specific Contrast Agent for Magnetic Resonance Imaging. Bioconjugate Chemistry, 1993. 4(5): p. 347-352.

PAGE 92

79 52. Suwa, T., S. Ozawa, M. Ueda, N. Ando, and M. Kitajima, Magnetic Resonance Imaging of Esophageal Squamous Cell Carcinoma Using Magnetite Particles Coated with Anti-Epidermal Growth Factor Receptor Antibody. International Journal of Cancer, 1998. 75(4): p. 626-634. 53. Sugimoto, T. and E. Matijevic, Formation of Uniform Spherical Magnetite Particles by Crystallization from Ferrous Hydroxide Gels. Journal of Colloid and Interface Science, 1980. 74(1): p. 227-243. 54. Matijevic, E., Monodispersed Metal (Hydrous) Oxidesa Fascinating Field of Colloid Science. Accounts of Chemical Research, 1981. 14(1): p. 22-29. 55. Matijevic, E., Interactions with Monodispersed Colloids. American Ceramic Society Bulletin, 1984. 63(8): p. 994-994. 56. Matijevic, E., Production of Monodispersed Colloidal Particles. Annual Review of Materials Science, 1985. 15: p. 483-516. 57. Matijevic, E., Monodispersed ColloidsArt and Science. Langmuir, 1986. 2(1): p. 12-20. 58. Matijevic, E., Monodispersed ColloidsAchievements and Problems. Abstracts of Papers of the American Chemical Society, 1990. 200: p. 51-59. 59. Atkinson, R.J., A.M. Posner, and J.P. Quirk, Crystal Nucleation in Fe(3) Solutions and Hydrxide Gels. Journal of Inorganic & Nuclear Chemistry, 1968. 30(9): p. 2371-&. 60. Atkinson, R.J., A.M. Posner, and J.P. Quirk, Crystal Nucleation and Growth in Hydrolyzing Iron(III) Chloride Solutions. Clays and Clay Minerals, 1977. 25(1): p. 49-&. 61. Her, Y.S., S.H. Lee, and E. Matijevic, Continuous Precipitation of Monodispersed Colloidal Particles: SiO2, Al(OH)(3), and BaTiO3. Journal of Materials Research, 1996. 11(1): p. 156-161. 62. Park, J., V. Privman, and E. Matijevic, Model of Formation of Monodispersed Colloids. Journal of Physical Chemistry B, 2001. 105(47): p. 11630-11635. 63. Privman, V., D.V. Goia, J. Park, and E. Matijevic, Mechanism of Formation of Monodispersed Colloids by Aggregation of Nanosize Precursors. Journal of Colloid and Interface Science, 1999. 213(1): p. 36-45. 64. Sada, E., H. Kumazawa, and H.M. Cho, Formation of Magnetite Fine Particles by Chemical Absorption into Aqueous Suspensions of Ferrous Hydroxide. Chemical Engineering Communications, 1989. 75: p. 89-99.

PAGE 93

80 65. Sun, S.H. and H. Zeng, Size-Controlled Synthesis of Magnetite Nanoparticles. Journal of the American Chemical Society, 2002. 124(28): p. 8204-8205. 66. Taylor, R.M., Influence of Chloride on the Formation of Iron-Oxides from Fe(II) Chloride: Effect of [Cl]/[Fe] on the Formation of Magnetite. Clays and Clay Minerals, 1984. 32(3): p. 167-174. 67. Taylor, R.M., Influence of Chloride on the Formation of Iron-Oxides from Fe(II) Chloride: Effect of [Cl] on the Formation of Lepidocrocite and Its Crystallinity. Clays and Clay Minerals, 1984. 32(3): p. 175-180. 68. Lee, S.H., Y.S. Her, and E. Matijevic, Preparation and Growth Mechanism of Uniform Colloidal Copper Oxide by the Controlled Double-Jet Precipitation. Journal of Colloid and Interface Science, 1997. 186(1): p. 193-202. 69. Johnson, T.J., E.A. Waddell, G.W. Kramer, and L.E. Locascio, Chemical Mapping of Hot-Embossed and UV-Laser Ablated Microchannels in Poly(Methyl Methacrylate) Using Carboxylate Specific Fluorescent Probes. Applied Surface Science, 2001. 181(1-2): p. 149-159. 70. Hermanson, G., Bioconjugate Techniques. 1996, San Diego: Academic Press. 71. Kapanidis, A.N. and S. Weiss, Fluorescent Probes and Bioconjugation Chemistries for Single Molecule Fluorescence Analysis of Biomolecules. Journal of Chemical Physics, 2002. 117(24): p. 10953-10964.

PAGE 94

BIOGRAPHICAL SKETCH Barry William Miller was born on July 12, 1981, in Akron, Ohio, to William and Carol Miller. He is the only son and youngest of three children. At age two, his family moved to Florida, and after a short stint in a rental home where he climbed his nostalgic first tree and learned to ride his first banana-seat bicycle, he moved to New Port Richey, Florida, where he grew up and attended school. During this period of his life he had to share a bathroom with his two elder sisters, Leigha and Lindsay, and he consequently learned at an early age the importance of putting the toilet seat down after each usea quality that will assuredly make him a great man, at least in the eyes of most women. Barry was very active as a young man and participated in many sports. After multiple head and body collisions with errant baseballs in Little League, he decided it was time for a change in sport. He made a few pursuits in basketball; however he later discovered his prowess in the swimming pool at age 14. Although the leg-shaving escapades were met with widespread peer ridicule, the form-fitting design of a Speedo never created the need for embarrassment. Everything else aside, he was very successful with swimming, garnering many district and conference championships as well as many commendable performances at state. After graduating from River Ridge High School near the top of his class in 1999, he matriculated at the University of Florida later that fall. A few years earlier, after being inspired by a discovery of how to bake Thirty-Minute Brownies in only twenty minutes, he decided that he would major in the field of engineering. He chose to pursue materials 81

PAGE 95

82 engineering with specialties in polymers and biomaterials. Never satisfied with the pace of his learning, he entered an accelerated program that allowed him to earn his masters degree in coincidence with his bachelors. These degrees will be awarded, with honors, in May 2004. After graduation he plans to pursue an engineering career in industry.


Permanent Link: http://ufdc.ufl.edu/UFE0004820/00001

Material Information

Title: Synthesis and Characterization of Functionalized Magnetite Nanocomposite Particles for Targeting and Retrieval Applications
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004820:00001

Permanent Link: http://ufdc.ufl.edu/UFE0004820/00001

Material Information

Title: Synthesis and Characterization of Functionalized Magnetite Nanocomposite Particles for Targeting and Retrieval Applications
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0004820:00001


This item has the following downloads:


Full Text











SYNTHESIS AND CHARACTERIZATION OF FUNCTIONALIZED MAGNETITE
NANOCOMPOSITE PARTICLES FOR TARGETING AND RETRIEVAL
APPLICATIONS














By

BARRY WILLIAM MILLER


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA


2004





































Copyright 2004

by

Barry William Miller

































This thesis is dedicated to my parents, Carol and Bill Miller, who actually did the hardest
work of all by raising me and always encouraging me to do my best.
















ACKNOWLEDGMENTS

Many people have contributed their time and effort in helping me to accomplish

this project. First and foremost, I must thank my advisor and committee chair, Dr. Laurie

Gower, for giving me direction in my research and so many helpful suggestions when

things just would not seem to work right in the lab. I am truly thankful that I had the

opportunity to work with and learn from her.

I must also thank the other members of my committee, Dr. Hassan El-Shall and Dr.

Rolf Hummel. They have been nothing but courteous, helpful, and accommodating

whenever I have sought their assistance.

Next, I must thank the Gower research group. In particular, Matt Olszta has spent

so much of his time helping me do all sorts of stuff. He taught me how to use the TEM,

how to use a bunch of computer programs, and just about how laboratory research is

supposed to work in general. I also thank Xingguo Cheng for running all of my XRD

samples, and even for all the humorous conversations about American celebrities and

culture. The unique personalities of Debra Lush, Fairland Amos, and Lijun Dai always

helped make the days in lab much more enj oyable.

I also thank Dr. Mark Meisel and his graduate student Ju-Hyun Park of the Physics

Department for running all of my SQUID samples (free of charge) and helping me

analyze the data. In addition, I acknowledge the Major Analytical Instrumentation Center

(MAIC) for the use of the transmission electron microscope and x-ray diffractometer.









I thank all of my friends who have been there over the years to talk to and laugh

with. I must especially thank Kevin, Smooth, Keith, Charlie, Nate, Spooner, and Pye for

being the reliable drinking crew that I would go out with most weekends. Somewhere

between the hundreds of beers and dozens of stories we have all had together, I think we

grew up a little bit along the way, or at least learned more about life. All of my years

here would have been much more of a chore without them, and I attribute much of my

success (and failure) to this group of guys.

I also want to acknowledge and thank all those people who made the road bumpy

along the way. And not because I enjoyed being subj ected to painfully long discussions

on irrelevant issues, receiving deceptive and misleading information, or even listening to

those who self-proclaim their unequaled greatness, but rather because it made me realize

that these same kind of people will still be inevitably present in every arena of my life

even after this thesis and university are behind me. I have had to learn how to

diplomatically interact with these people, and so it is for this reason that I thank them.

Finally, I want to thank my parents, Carol and Bill, and my two sisters, Leigha and

Lindsay. They have always been the one bastion of love, support, and advice that I never

have to think twice about going to in good times or bad. Anything that I might ever

accomplish in my lifetime should be rightfully attributed to them, especially my parents.

They have raised me to be the person I am today and they have always encouraged me to

cultivate my talents and pursue my interests. I will be forever proud to call them my

mom and dad.





















TABLE OF CONTENTS


Page


ACKNOWLEDGMENT S .............. .................... iv


LI ST OF T ABLE S ................. .............._ viii...._......


LIST OF FIGURES .............. .................... ix


AB STRAC T ................ .............. xii


CHAPTER


1 INTRODUCTION ................. ...............1.......... ......


Biological Inspiration .............. ...............1.....
Purpose of Research .............. ...............2.....
Specific Aim s............... ...............3..



2 BACKGROUND ................. ...............5.......... ......


Introducti on ................. ...............5.................

Magnetic Basics............... ...............5.
Types of Magnetism ................. ...............7................
Param agnetism .............. ... .. ...............
Ferromagnetism and Ferrimagnetism ......___ ....... ......__............
Superparamagnetism .............. ...............12....
Antiferromagneti sm ........._ ....... .__ ...............13....
Magnetic Particle Applications............... ..............1
Particle Synthesis............... ...............1



3 PARTICLE SYNTHESIS WITH POLY(ACRYLIC ACID) ADDITIVE.................22


Introducti on ................. ...............22.................
M material s .............. ...............23....
M ethod s .............. ...............24....

Setup ................ ...............24.................
Synthesis............... ...............2
Separation ................. ...............26.................












Characterization Results and Discussion............... ...............2
Transmission Electron Microscopy (TEM) ................. ................. ..........27
X-Ray Diffraction (XRD)........... .. .................. ...............3
Superconducting Quantum Interference Device Analysi s (SQUID) ...................3 7
Fluorescence Labeling ................. ...............43.................
Procedure ................. ...............43.................
Fluorescence microscopy .............. ...............44....



4 PARTICLE SYNTHESIS WITH POLY-L-GLUTAMIC ACID ADDITIVE........... 47


Introducti on ................. ...............47.................
M material s .............. ...............48....
M ethods ................ ..... .. .... ...........4
Characterization Results and Discussion............... ...............5
Transmission Electron Microscopy (TEM) ................. ............................50
X-Ray Diffraction (XRD)........... .. .................. ...............5
Superconducting Quantum Interference Device Analysi s (SQUID) ...................5 5
Fluorescence Labeling ................. ...............58.................



5 PARTICLE SYNTHESIS WITH POLY-L-LYSINE ADDITIVE ............................61


Introducti on ................. ...............61.................
M material s .............. ...............62....
M ethods ................ ..... .. .... ...........6
Characterization Results and Discussion............... ...............6
Transmission Electron Microscopy (TEM) ................. ............................64
X-Ray Diffraction (XRD)........... .. .................. ...............6
Superconducting Quantum Interference Device Analysi s (SQUID) ................... 67



6 CONCLUSIONS AND FUTURE WORK ................. ...............70........... ...


LIST OF REFERENCES ................. ...............75........... ....


BIOGRAPHICAL SKETCH .............. ...............8 1....

















LIST OF TABLES


Table Page

2-1 Magnetic units and conversion table ......___ ... ......___...... ..........7

3-1 Experimental design of trials with poly(acrylic acid) ..........._.._.. .........__ ........26

3-2 Magnetic quantities determined from SQUID analysis of control particles ............38

3-3 Magnetic quantities determined from SQUID analysis of PAA particles ...............40

4-1 Experimental design of trials with poly-L-glutamic acid ................. ................ ..50

4-2 Magnetic quantities determined from SQUID analysis of GLU particles ........._....56

5-1 Experimental trials with poly-L-lysine. ............. ...............64.....

5-2 Magnetic quantities determined from SQUID analysis of LYS particles ................68

















LIST OF FIGURES


Figure Page

2-1 Magnetization curves for the different types of magnetism ................. ................. .8

2-2 Magnetic alignment in paramagnetic materials; (a) is in the absence of a magnetic
field, and (b) shows the response in a moderately applied field. ............. ................9

2-3 Schematic of a hysteresis loop. ............. ...............10.....

2-4 The magnetic origin of magnetite ................ ............... ........ ......... ...12

2-5 Stages of gel conversion to magnetite. (a) amorphous gel (0 min)(b) formation of
hexagonal platelets (15 min)(c) primary particle formation (30 min)(d) primary
particles aggregate to larger particles (45 min); (e) uniform spherical particles (120
m in). ............. ...............19.....

2-6 Rod-like crystals of goethite formed as an unwanted product of particle synthesis.21

3-1 Repeat unit structure of poly(acrylic acid) ................. ............ ..... 22.... ..

3-2 Digital picture of setup. ............. ...............25.....

3-3 TEM micrographs of control particles .............. ...............28....

3-4 Electron diffraction pattern of control sample. This pattern was indexed and found
to correspond to magnetite. Note that the presence of diffracted rings rather than
spots indicates polycrystalline nature ......__.. .........._._ ....._. ............2

3-5 TEM micrographs of particles with PAA, MW = 2,100. Both of the pictures above
show particles synthesized with a PAA concentration of 130 ug/mL. The sample
with a concentration of 220 ug/mL did not convert during the aging process and
thus no particles were formed. ............. ...............29.....

3-6 TEM micrographs of particles with PAA, MW = 6,000. The top two images are
taken from the sample with a PAA concentration of 100 ug/mL. The bottom two
images had a PAA concentration of 200 ug/mL. ........._.._.. ...._.._ ...............30

3-7 TEM micrographs of particles with PAA, MW = 15,000. The top two images are
taken from the sample with a PAA concentration of 100 ug/mL. The bottom two
images had a PAA concentration of 230 ug/mL. ........._.._.. ...._.._ ...............30










3-8 TEM micrographs of particles with PAA, MW = 30,000. The top two images are
taken from the sample with a PAA concentration of 100 ug/mL. The bottom two
images had a PAA concentration of 200 ug/mL. ........._.._.. ...._.._ ...............3 1

3-9 X-ray diffraction spectrum of control particles. The data for the magnetite standard
is also plotted for comparative purposes. The peaks are labeled with their
corresponding planes of diffraction ................. ...............36......_.__....

3-10 X-ray diffraction spectrum of composite particles with PAA additive. The data for
the magnetite standard is also plotted for comparative purposes. The peaks are
labeled with their corresponding planes of diffraction............._._ ........._._ ......36

3-11 Hysteresis curves for control particles. ............. ...............38.....

3-12 Hysteresis curves for PAA modified composite particles ................. ................. .39

3-13 Transmission and fluorescence micrographs of control particles. The light
transmission image on the left shows clumps of particles dispersed on the slide.
The fluorescence micrograph on the right shows the fluorescence (or lack thereof)
of that same area ................. ...............44................

3-14 Fluorescence images of PAA composite particles. The first two images come from
the sample with MW=15,000 and concentration 100 ug/mL. The third image
comes from sample MW=6,000 and concentration 100 ug/mL. ........._.._... .............45

4-1 Repeat unit structure of poly-L-glutamic acid. ........._... ...._.. ........._.......47

4-2 TEM micrographs of particles with poly-L-glutamic acid, MW = 7,500. The top
two images are taken from the sample with a GLU concentration of 80 ug/mL. The
bottom two images had a GLU concentration of 100 ug/mL. ............. ...............51

4-3 TEM micrographs of particles with poly-L-glutamic acid, MW = 13,600. The top
two, middle two, and bottom two images have concentrations of 60 ug/mL, 100
ug/mL, and 150 ug/mL, respectively. ............. ...............52.....

4-4 X-ray diffraction spectrum of composite particles with GLU additive. The data for
the magnetite standard is also plotted for comparative purposes. The peaks are
labeled with their corresponding planes of diffraction............._._ ........._._ ......54

4-5 Hysteresis curves for GLU modified composite particles. ............. ....................55

4-6 Fluorescence images of GLU composite particles. Both images come from the
sample with MW=13,600 and concentration 150 ug/mL..........._.._.._ ........_.._.. ...59

5-1 Repeat unit structure of poly-L-lysine. ................ ................ ................61










5-2 TEM micrographs of particles with poly-L-lysine, MW = 27,000. The top two
images are taken from the sample with a LYS concentration of 100 ug/mL. The
bottom two images had a LYS concentration of 250 ug/mL. ............. ..................64

5-3 X-ray diffraction spectrum of composite particles with LYS additive. The data for
the magnetite standard is also plotted for comparative purposes. The peaks are
labeled with their corresponding planes of diffraction ................. .....................66

5-4 Hysteresis curves for particles with LYS additive. ................. .................6

5-5 Hysteresis curves for both LYS and control particles. .............. ....................6
















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

SYNTHESIS AND CHARACTERIZATION OF FUNCTIONALIZED MAGNETITE
NANOCOMPOSITE PARTICLES FOR TARGETING AND RETRIEVAL
APPLICATIONS

By

Barry William Miller

May 2004

Chair: Laurie Gower
Major Department: Materials Science and Engineering

There is much interest in the use of biologically functional magnetic particles for

biomedical applications. The general idea is that an external magnetic field can guide the

particles to a specific area of the body, and then the functionality of the particles will

allow them to target specific cells or tissues. These particles must therefore be small

enough to flow through blood capillaries. This study presents the synthesis procedure

and characterization of a novel magnetic particle system that possesses organic

functionality. These particles can achieve high magnetizations due to their high content

of magnetite, Fe304, and they are generally on the order of 150-3 50 nanometers.

Functionality was afforded to the particles by addition of three separate polymers

during the solution synthesis procedure. The effect of the molecular weight and

concentration of the selected polymer additive was explored. Poly(acrylic acid) and

poly-L-glutamic acid were used to give the particles carboxyl functionality. Attempts

were also made to incorporate poly-L-lysine to give them an amino functionality;









however the nature of this polymer in solution prevented the synthesis of composite

particles; instead, a non-functional colloid of magnetite particles was formed. The

functionality and reactivity of the carboxyl-modified particles were demonstrated by

conjugation of a fluorescent probe and subsequent fluorescence microscopy. All particle

trials were characterized with transmission electron microscopy (TEM), which contrasted

the morphology of the composite particles with the non-functional control particles. A

superconducting quantum interference device (SQUID) was employed to measure

magnetic properties and determine magnetic behavior. Finally, x-ray diffraction (XRD)

was performed to confirm that the crystalline phase of the particles was magnetite.

Overall, this research shows that a new door has been opened to a magnetic particle

system that can be used in a wide range of applications.















CHAPTER 1
INTTRODUCTION

Biological Inspiration

Iron-based compounds are ubiquitous in nature. They exist as geological,

biological, and even extraterrestrial minerals. The primary interest for this project is in

the iron oxides, namely magnetite. However, other iron oxide phases and some iron oxy-

hydroxides will be necessarily addressed in the background chapter.

The inspiration behind this thesis proj ect stems from the discovery and research

surrounding iron biominerals [1]. Iron-containing minerals have been found in a large

variety of organisms, ranging from bacteria to humans [2, 3]. Although the function of a

mineral is not known for every organism in which it is found, it has been discovered that

organisms utilize these minerals for various purposes, including mechanical grinding and

magnetotaxis [1, 4]. What makes these mineral occurrences so interesting is that they

must be produced in natural conditions and must have some mechanism controlling their

synthesis because such a high level of size and shape control is achieved. It has been

found that the synthesis may be controlled in some different ways; a protein matrix is

sometimes present which allows the mineral to grow epitaxially [5]; in some instances, a

mineral phase is synthesized within a vesicle that can control local concentrations of

reactants [6, 7]. Whatever the specific case may be, such research shows that these

biominerals and their syntheses are closely intertwined with the effects and properties of

various proteins or polymers.










Of particular interest to this proj ect are the nano-sized particles of magnetite

(Fe304) found in magnetotactic bacteria. These organisms live in marine or fresh water

sediments and their magnetic particles respond to geomagnetic Hields which in turn orient

and guide them to nutrients [4]. Much research has been performed to characterize these

magnetic biominerals and understand their synthesis and function [8-16]. What has been

found is that these particles are contained in what is called a magnetosome. The

magnetosome is essentially a lipid bilayer membrane with various bound proteins. This

structure is what regulates the flux of iron for synthesis, however the exact mechanism is

not fully known. The resulting particles are very controlled in size and shape, but also

vary by species. This suggests that the protein composition of the magnetosome may

also vary and thus influence the nucleation and growth of the magnetite crystals. It is this

association of a macromolecule with a magnetic particle that raised interest for this

project. The question was, what happens if we try to do this synthetically?

My advising professor, Dr. Laurie Gower, made some efforts in her own graduate

research to control particle morphology using synthetic polymers. Rather than achieving

elongated or specifically oriented minerals, she discovered an interesting matrix-like

composite morphology that was not researched further. Therefore, the impetus behind

this research was the realization that magnetic mineral and polymer can be used in

conjunction to create a magnetic nano-composite with organic functionality.

Purpose of Research

The overall purpose of this research is to create a magnetic nano-composite particle

system with organic functionality. Such a system is desired for various targeting and/or

retrieval applications, both medical and environmental. The idea is that the particles can

be magnetically guided to specific areas of the body, and/or have the capability for









retrieval by the same mechanism. It was desired to try to achieve a large magnetic

moment with this system, thus the material selected for the magnetic aspect was

magnetite (Fe304). Magnetite has been well known and characterized for hundreds of

years, and the chosen method of synthesis is relatively inexpensive and easy; therefore,

there is also a feasible capacity for large-scale production. The organic functionality is

afforded to this system by the addition of a polymer. The three polymers used in this

research were poly(acrylic acid), poly-L-glutamic acid, and poly-L-lysine. The first two

have carboxylic acid functional groups, whereas the latter has a free amine group. These

functionalities can be used with bioconjugation techniques to attach proteins, drugs,

antibodies, or other desired probes. Homopolymers of poly(amino acids) were

investigated because specific sequences of amino acids are known to bind to specific

cells or other receptors. The use of an amino acid homopolymer is the first step toward

incorporating sequenced polypeptides. Therefore, the ultimate goal for this system would

be to incorporate specifically sequenced polypeptides into magnetic particles that would

then target a preselected cell moiety. However, this goal was unrealistic for this project

because of the difficulty and expense of synthesizing specific protein sequences at

moderate molecular weights. Therefore, this research stands to demonstrate the

feasibility of producing such a system.

Specific Aims

There were three primary goals that I wanted to achieve with this proj ect. The first

was to optimize the technique of particle synthesis. The background and procedure of the

actual method will come in later chapters, but it was a challenge to find the right

parameters to make the reaction go as near to completion as possible on each trial.









Clearly, a reliable and repeatable technique was a necessary component for this research

because it facilitates the accurate analysis of results.

Once the synthesis was optimized, the next specific aim was to characterize the

particle system. In order to do this, it was necessary to first confirm that the mineral

phase produced was indeed magnetite (Fe304). This was performed using x-ray

diffraction. It was also necessary to show the particle morphology to support that indeed

the particle structure has a matrix-like composite morphology. This was done by

transmission electron microscopy, where bright field images were taken to show not only

the particle morphology, but also particle size. For biomedical applications, it is

important that the particles be small enough to flow through blood capillaries, which are

about 5-10 microns in inner diameter [17]. The magnetic properties of the composite

particles were measured using a superconducting quantum interference device (SQUID).

This characterization was important so that this magnetic particle system can be

compared with other magnetic systems in related applications.

The final aim of this proj ect was to demonstrate the organic functionality of the

particles. Transmission electron microscopy shows that the polymer is present, but it is

important to prove that the polymer is also functional. To do this, a fluorescent probe

was conjugated to the polymer on the particles. The resulting fluorescence shows that the

functional groups of the polymer are indeed reactive.















CHAPTER 2
BACKGROUND

Introduction

Since this proj ect deals with the characterization of a magnetic particle system, it is

important that we first understand the basics of magnetism. The different types of

magnetism will also be discussed and corresponding reference will be given to the

materials that were encountered in this research, namely magnetite and goethite.

Important magnetic quantities and properties will also be explained, especially as they

become relevant to this work. In addition, this chapter will present some of the current

magnetic particle systems and also how the composite particles of this research might be

used in similar applications. Finally, this chapter will conclude with a discussion on the

actual method of particle synthesis and how it was modified for use in this proj ect.

Magnetic Basics

The root of magnetism is based on the response that a material has when exposed to

an external magnetic Hield. The electron spins in the material align in the direction of the

applied Hield, thereby magnetizing the material. It is important that we first define some

magnetic properties before we get too far into discussion. An applied magnetic Hield, H,

incites a response from a material called magnetic induction, B. The relationship

between B and H can be defined by equation 2-1,

B = H + 4xnM (2-1)

where M is the magnetization of the material. Magnetization is the magnetic moment per

unit volume, and is a property that depends on the magnetic moments of the constituent









atoms, as well as their interactions with each other. It is important to note that equation

2-1 is for cgs units; the relationship in terms of SI units is shown by equation 2-2.

B = CloH + CloM (2-2)

Cl0 is a constant called the permeability of free space. There are also common terms for

ratios between some of these different quantities because the magnetic properties of a

material are often defined by how they vary with an applied magnetic Hield. So, the ratio

of M to H is called the susceptibility and is indicative of how responsive a material is to

an applied magnetic Hield. Equation 2-3 defines this.

X = M/H (2-3)

The ratio of B to H is called the permeability and is indicative of how well the magnetic

Hield can permeate the material. Equation 2-4 defines this property.

C1 = B/H (2-4)

From equations 2-3 and 2-4 above we can derive a relationship between the susceptibility

and the permeability,

C1 = 1 + 4x: X (2-5)

or, in SI units

1/Cl0 = 1 + X (2-6)

A summary of these quantities, their units, and their conversion factors is included in

Table 2-1. The following section breaks down the different types of magnetism and will

reference these properties, so it is important to keep them and their relationships with

each other in mind.































dimensionless


dimensionless


4n: 107 CLo Wb/(A m)


Types of Magnetism

Now that the basics have been defined, we must now explain how these properties

relate to the actual magnetic behavior of a material. Most magnetic materials are

classified by how they relate to these quantities. In particular, magnetization curves are

often used to describe the nature of a magnetic material. A magnetization curve plots

magnetization, M, or magnetic induction, B, as a function of applied magnetic field, H.

Examples of these general curves and the types of magnetism they correspond to can be

seen in Figure 2-1 [18].

While individual treatment will be given to paramagnetism, ferromagnetism,

ferrimagnetism, and antiferromagnetism, a discussion of the properties of diamagnetic

materials will not be presented at length because this form of magnetism does not really


Table 2-1: Magnetic units and conversion table
Gaussian Units (CGS
Quantity Symbol
Units)
Magnetic Flux
Density; Magnetic B Gauss (G)
Induction
Maxwell (Mx)
Magnetic Flux (])et c2 '


Conversion
Factor C

10-4


108


SI Units

Tesla (T)
W /m2
(Wb -weber)

Weber (Wb)


Magnetic Field
Strength


Magnetization


Magnetic moment

Permeability of
vacuum


Permeability


103/4n


103


H Oersted (Oe)


M emu/cm3


A/m


eniu
m
(electro-magnetic unit)


4x*10


A m


Wb/(A m)










relate to this research. Let it suffice to say that diamagnetic materials tend to exclude the

magnetic field from their interior, thus they usually have a small and negative

susceptibility, with a permeability less than one; this behavior is shown schematically in

the left plot of Figure 2-1.


iM(emu/crn3 )
M(entu/cmj )

2000 -





0.5 H- (Oe)
o) 1oo
Figure 2-1: Magnetization curves for the different types of magnetism. Notice the
difference in scales between the two plots.


Paramagnetism

The phenomenon of paramagnetism occurs due to weakly coupled magnetic

moments. With no applied field, thermal energy causes the magnetic moments to

randomly align, resulting in a net magnetic moment of zero [19]. When a magnetic field

is applied, the individual moments respond by turning in the direction of the field,

however the net magnetization is still relatively weak because the moments do not

completely align with the applied field. Therefore, values for susceptibility are small,

and permeability is just greater than one. A schematic of the alignment of spins in

paramagnets is shown in Figure 2-2. Clearly, temperature will always play a role in

paramagnetism, as a higher temperature requires a greater magnetic field to overcome the

thermal forces working against the ordered alignment of magnetic moments. Typical

paramagnetic materials include transition metal salts and rare earth salts. As we will see









in the upcoming sections, paramagnetic behavior is also observed in other types of

magnets under certain conditions.














Figure 2-2: Magnetic alignment in paramagnetic materials; (a) in the absence of a
magnetic field, and (b) the response in a moderately applied field.


Ferromagnetism and Ferrimagnetism

For all intensive purposes, the main difference between ferromagnetic and

ferrimagnetic materials is that the latter materials are ceramic and good insulators rather

than conductors. Other than this, both kinds of magnetism are very similar and will be

treated the same in this discussion. Both types of materials typically show a hysteresis

loop as a magnetization curve. A hysteresis loop is created because after a magnetic field

is applied and then removed, the material retains some magnetization. This is called

remanent magnetization. In order to bring the magnetization down to zero, a magnetic

field must be applied in the opposite direction until there is no net magnetization

remaining. The value for the strength of the field necessary to do this is called the

coercivity, H,. If the field is applied to saturation in the negative direction and then

removed, then applied again in the positive direction, a full hysteresis loop is formed. A

schematic of such a magnetization curve is shown in Figure 2-3.





















Figure 2-3: Schematic of a hysteresis loop.


Materials that display such hysteresis behavior can be classified as either hard or

soft magnets. Hard magnets have a large remnant magnetization and a large coercivity,

therefore they show a large area within the hysteresis loop. They are so-called hard

magnets because it is hard to bring them to saturation and consequently coerce them back

to zero. These types of magnets are often used where magnetic memory is desired, such

as in magnetic recording media, because of their large remnant magnetization. Soft

magnets, however, possess a low coercivity and a low remnant magnetization. This also

means that a much smaller magnetic field would be necessary to reach a saturation point

of magnetization. However, because of their low remnant magnetization, these materials

are usually not good for recording media because increases in temperature can disrupt the

alignment of moments very easily.

Now we must look at why this behavior occurs in ferromagnetic and ferrimagnetic

materials. The spins of unfilled d-band electrons spontaneously align and create a

magnetic moment [19]. However, this does not occur over the entire crystal, but rather

these groups of aligned magnetic moments are contained in sub-structures known as

domains [20]. Each of these domains has their own saturation magnetization without the

presence of an external field. However, the material as a whole maintains a net










magnetization of zero in the absence of an applied field because each of the domains are

aligned in different directions, effectively canceling each other out. When a magnetic

field is applied, the domains already aligned in the direction of the field begin to grow at

the expense of the unfavorably aligned domains, thus creating a net magnetization.

These domains continue to grow until the material effectively contains one single domain

that is oriented in the direction of the field. It is at this point that the material reaches its

saturation magnetization. Now that we can see how the domains transform to affect

magnetic behavior, it should make sense why hysteresis behavior is observed for these

types of materials.

It is also important to mention that temperature again plays an important role in

ferro- and ferrimagnetism. As temperature is increased, there comes a point where the

thermal fluctuations will overcome the force of the spontaneously aligned spins and cause

them to become randomly oriented [21]. Once this occurs, the material exhibits

paramagnetic behavior. This threshold temperature is known as the Curie temperature,

T,. It is very important to keep this value in mind, especially when dealing with a

material with a relatively low To because it can greatly affect the material's efficiency in

its intended application. In addition to this, even if the Curie temperature is not reached,

an increased temperature will result in a lower magnetization yield because of the

increased thermal fluctuations.

It should here be noted that the magnetic material utilized in this research,

magnetite (Fe304), iS ferrimagnetic in nature. A more representative way to express the

composition would be to write Fe2 Fe3+204, because the iron ions exist in two different

valence states, and this is actually what gives rise to the magnetic moment. This material










has the inverse spinel crystal structure, with 8 Fe3+ ions occupying tetrahedral sites, 8

Fe3+ Occupying octahedral sites, and 8 Fe2+ Occupying octahedral sites. The Fe3+ ionS in

the octahedral sites have opposing moments to the Fe3+ ions in the tetrahedral sites, thus

they cancel each other out. Therefore, the overall magnetic moment comes from the sum

of the magnetic moments of the Fe2+ ions in the octahedral sites. Figure 2-4 shows a

schematic of the origin of magnetism in this material [22].










SCalion in octahedral slte 3 caer ieste
D? cation in teltahedral site 7- '' (32 per an t cell)



Tetrahedral interstice
(64 per unit cell)

Octahedral Telrahedral
Fe OOOOOOOO OOOOOOOO

Fe2 OOOOOOOO

Figure 2-4: The magnetic origin of magnetite


Superparamagnetism

Superparamagnetism is an interesting phenomenon that comes into play when

ferromagnetic or ferrimagnetic particles become very small. At particle sizes of about 10

nanometers, these materials begin to exhibit paramagnetic behavior, even when they are

below their Curie temperature. The reason for this is that thermal effects, while not

strong enough to overcome the forces between individual atoms, are strong enough to










change the magnetization direction of the entire particle. The result is a random

arrangement of magnetic directions among crystallites, thus giving a net magnetic

moment of zero.

This phenomenon gives rise to the limitation of how small magnetic recording

media can get because superparamagnetism will cause the particles to loose their memory

from thermal influences. Superparamagnetic particles are therefore often used in many

magnetic systems in the biomedical field because not only are they small, but they also

do not retain any magnetic remanence. The latter reason is important because it means

that the particles will not aggregate due to magnetic forces, however the trade-off is that

the particles are paramagnetic in behavior and therefore it is more difficult to achieve a

high magnetization. For these reasons, this research aimed to use particles that were in

the size range of a few hundred nanometers, thus allowing them to retain their

ferrimagnetic properties yet still be small enough to flow through blood capillaries if

necessary. As we will see in the experimental chapters, the particles are very soft

magnets and have only a small remnant magnetization.

Antiferromagnetism

Like ferromagnets, materials exhibiting this behavior have a spontaneous alignment

of moments. The difference is that the adj acent atoms in these materials have antiparallel

spins, thus there is no net magnetic moment observed. This is really as far into

discussion as we need to get for this proj ect because it explains all we need to know. The

particle synthesis process often creates some unwanted byproducts because of oxygen

getting into the system (details will be explained in a following section). The main

byproduct is goethite, ot-FeOOH, an iron oxyhydroxide that possesses antiferromagnetic










properties. Therefore, this material can be separated relatively easy since it will not

really be attracted to a magnetic field. In practice, the separation is a little more difficult,

but for all intensive purposes it is important to note that this byproduct will not contribute

much to the magnetic properties of the overall particle system. Also of note is that

hematite, Fe203, iS antiferromagnetic, and it is possible that this material may also form

in very small quantities during synthesis, however it can be removed in the same manner

as the goethite.

Magnetic Particle Applications

The capabilities of creating nano and micro-sized particles of magnetite have lead

to the use of these small magnetic particles in a range of applications [23]. Many studies

have been conducted to characterize the magnetic behavior of various sizes of magnetite

particles [24-26]. Because of its useful properties, magnetite has been the material of

choice for many magnetic particle systems.

Ferrofluids are an interesting example of such systems. A ferrofluid is basically a

colloidal solution of magnetic particles that are suspended in either a polar or non-polar

liquid [27]. Magnetite is commonly used as the magnetic material, but iron and cobalt

particles have been used as well. The particles are typically on the order of about 10

nanometers so that they are superparamagnetic. This is desired so that the fluid remains

as a stable suspension and the magnetic particles do not aggregate together and form

clumps or settle in the absence of a magnetic field. When a magnetic field is applied,

however, the particles will respond and are often used as a seal that can be applied or

removed with a magnetic field. These fluids have been used in applications such as

rotary seals for disk drives and dampers for audio speakers [28].









Recently there has been interest in using ferrofluids for biomedical applications.

The important factor in making this application transition is that the fluid must be

biocompatible. There has been some success in creating water-based suspensions [29],

and such ferrofluid systems have been used for cell sorting [27]. In order to do this, a

biological effector is bound to the particle surface so that the particles can target specific

cells. These cells can then be sorted by employing a gradient magnetic field to separate

them. This system begins to appear very similar to the system characterized in this thesis

because it entails that the magnetic particles are biologically functional. However, this

one system in particular uses particles of Fe203, a material that will not achieve a very

high magnetization as compared to a system with Fe304. In addition, where magnetite is

used as the material of choice, the particles are still in the superparamagnetic size range,

thus they will still not achieve a high magnetization as compared to larger particles that

exhibit ferrimagnetic behavior. This is a very important factor that will come into play

especially for in vivo applications.

Magnetic particle systems are also being explored for site-specific tumor therapy

and/or drug delivery. One of the most interesting properties of magnetic particles is that

they will begin to heat up if an alternating magnetic field is applied to them [30].

Therefore, the theory behind some of these systems is that it should be possible to

magnetically guide the particles to a tumor, apply an alternating magnetic field, and

consequently treat the tumor via magnetocytolysis. There has been some demonstrated

success doing this [31], however many systems are also trying to incorporate a drug

delivery mechanism to additionally treat specific areas of the body. For example, there is

one system that contains dispersed magnetic particles in a polymer matrix which also









contains a dispersed therapeutic agent. When an oscillating magnetic field is applied, the

diffusional release of the drug is increased [32].

Many studies involving magnetic particles have turned their attention to coatings,

or, to the use of magnetic microspheres. This is being done in some instances for certain

biocompatibility issues [33], but most often it is for the purpose of functionalizing or

attaching drugs to the particles for use as targeted drug delivery systems [34-40]. A very

common methodology for these delivery systems is to create a crosslinked protein or

polysaccharide microsphere with a magnetic core [41-47]. The drugs can then be

chemically or physically associated with the particles. A large drawback to these systems

is that they contain only a small volume fraction of magnetic material, and this of course

degrades overall magnetic capabilities.

Functional and site-specific magnetic particles are also of great interest to the field

of magnetic resonance imaging (MRI). Magnetic particles work well as negative contrast

agents in MRI because they shorten the T1 and T2 relaxation times [48-50]. Therefore,

for better imaging purposes, it would be useful to have a contrast agent that could target a

specific tissue, organ, or tumor. One method of achieving this is through the attachment

of antibodies to the magnetic particles [51, 52]. MRI just serves as one more example of

a scientific field that can benefit from functional magnetic particles.

In conclusion, many of the studies of all these kinds of systems are not adequately

focused on the magnetic properties of the particles. For in vivo applications, many of

these systems will have difficulty in achieving a large magnetization for guidance

purposes, and thus the efficacy of these systems will be deteriorated. The main reason

for this is that many of the systems do not have a very high mineral loading due to the









needs of coatings or attachment of other materials. As a typical example where the

magnetic properties of a system were characterized, one study used particles that had

either a 23 or 29 weight percent mineral loading and their corresponding magnetizations

were 20 and 30 emu/gram [35]. The primary argument we are attempting to make with

our composite particle system is that the particles developed in this research have a very

high content ratio of magnetite to polymer, and so the particles can subsequently achieve

a high magnetization. In addition, because our composite particles are not

superparamagnetic, it follows that a higher magnetization can be achieved just on the

basis of intrinsic magnetic properties. In fact, the primary obstacle to many of the

targeted drug therapy systems is that it is difficult to obtain enough magnetic force to

retain the superparamagnetic particles in deep body tissue.

It should also be noted that the magnetic particles introduced in this thesis do not

need to be limited to in vivo applications. There exists the possibility of using this system

for targeting and retrieval in environmental applications. For example, the anthrax scares

in recent years have led to the need for a way to eliminate spores that may have been

spread in a room. A magnetic system such as the one researched in this thesis could

target these spores, bind them, and then remove them by magnetically retrieving the

particles. This is just yet another example of the utility of these particles. Whether it be

drug delivery or targeted MRI contrast, this new magnetic particle system is an

improvement on current systems and will be useful for a broad range of applications.

Particle Synthesis

The method for the particle synthesis used in this research was a modification of

the method used by Sugimoto and Matijevic [53]. Matij evic is well known for his work

in the synthesis of uniform colloids of many different types of ceramic and metal










particles [23, 54-58]. The synthesis of interest is one where a ferrous hydroxide gel is

precipitated from solution, and then this gel is aged to create spherical crystals of

magnetite, Fe304. Similar types of solution synthesis techniques are very commonly used

to produce relatively monodispersed, colloidal particles for many different types of

materials [59-68]. It was important to choose a procedure that included the

transformation of an amorphous intermediate to the Einal particle so that the polymer

additives used to create the composite particles could have a chance to incorporate during

the synthesis.

The procedure is composed of two main steps. The first step is to precipitate the

amorphous gel. This is done by introducing an iron sulfate solution to a potassium

hydroxide solution, thus creating the Fe(OH)2 gel. The second step is then to place this

gel in a 900C oil bath for about four hours, or until the conversion from gel to magnetite

is complete. It is important to note that the potassium hydroxide solution is also mixed

with a potassium nitrate solution, a component necessary for the conversion to magnetite.

What actually occurs during aging is interesting. Figure 2-5 shows the conversion from

gel to magnetite as the aging progresses.

One of the most interesting variables in this synthesis is the effect that the excess

ion concentration of either Fe2+ Or OH- has on the resulting particles. It seems that when

there is an OH- excess, the primary particles are less likely to aggregate together to form

particles; instead, the primary particles tend to grow by the addition of individual ions to

the crystal surface. This often results in the Einal particles having a cubo-octahedral

morphology. On the other hand, if there is a Fe2+ eXCOSs, it seems that the primary

particles do indeed aggregate together and then a surface recrystallization mechanism is






19

induced to create the final spherical particles of magnetite. As we will see in the

experimental chapters to follow, a consistent excess of Fe2+ was used for all of the trials.


Figure 2-5: Stages of gel conversion to magnetite. (a) amorphous gel (0 min); (b)
formation of hexagonal platelets (15 min); (c) primary particle formation (30
min); (d) primary particles aggregate to larger particles (45 min); (e) uniform
spherical particles (120 min).


6);









This is important because the polymer additive can better be incorporated into the final

particle by lodging itself between primary particles. Although this has not been proven to

be the exact mechanism, it is a very likely explanation. A further discussion of the

formation of the composite particles will be given in the following chapters. Of

immediate interest to the synthesis of pure magnetite particles is the oxidation reaction

that transforms Fe(OH)2 into Fe304, and then also the oxidation of Fe2+ to Fe304. These

are only possible reactions and may not fully describe what is happening [53]. For the

Fe(OH)2 to Fe304 reactions:

3Fe(OH)2 + NO3- 4 Fe304 + NO2- + 3H20

3Fe(OH)2 + 2NO2- 4 Fe304 + 2NO + 2H20 + 20H-

15Fe(OH)2 + 2NO 4 5Fe304 + 2NH3 + 12H20

For the oxidation of Fe2+ to Fe304:

3Fe2+ + NO3- + 3H20 4 Fe304 + NO2- + 6H+

3Fe2+ + 2NO2- + 2H20 4 Fe304 + 2NO + 4H

15Fe2+ + 2NO + 18H20 4 5Fe304 + 2NH3 + 30H

It should also be mentioned that this reaction needs to be performed in an inert

atmosphere because the presence of oxygen in the system can cause the precipitation of

unwanted crystals, namely goethite, ot-FeOOH. Since we do live on Earth and oxygen is

everywhere, it is inevitable that some oxygen always gets into the system and forms rod-

like crystals of goethite. The bulk of the synthesis product, however, is magnetite.

Goethite is usually not a huge problem because--as mentioned earlier--it is

antiferromagnetic and can therefore be separated from the magnetite by magnetic means.

However, some traces of this material still exist in the samples and can even be seen in










some of the images shown in the results of this thesis. Figure 2-6 shows an example of

what the goethite crystals look like so that there is no surprise when we examine the

results in the coming chapters.


Figure 2-6: Rod-like crystals of goethite formed as an unwanted product of particle
synthesis.















CHAPTER 3
PARTICLE SYNTHESIS WITH POLY(ACRYLIC ACID) ADDITIVE

Introduction

As its name suggests, poly(acrylic acid), or PAA, is an acidic polymer that contains

a carboxylic acid group (-COOH) on every repeat unit. The repeat unit structure is

shown in the figure below.







Figure 3-1: Repeat unit structure of poly(acrylic acid)

The carboxylic acid group is very reactive and therefore it will be useful for conjugation

purposes when incorporated into the magnetic nanoparticles. PAA is an easily

synthesized and common polymer, therefore it was inexpensive to purchase a series of

molecular weights. The ability to investigate a range of molecular weights means that

possible trends may be uncovered and analyzed. PAA was also chosen because it has the

same functionality and acidic nature as poly-L-glutamic acid, the polymer of focus in

chapter 4.

This chapter will also show the results for the control samples and discuss them in

relation to the composite particles that are formed with PAA. The control reaction is

simply the synthesis of particles with no polymer added into the solution. The result will

be shown to be a colloid of magnetite particles with no organic functionality. Both the

control samples and those with PAA were characterized with transmission electron









microscopy (TEM) to show morphology and size. Magnetic measurements were made to

determine polymeric effects on magnetization, and x-ray diffraction was performed to

confirm that the crystalline phase of the particles is indeed magnetite (Fe304). In

addition, fluorescence labeling and microscopy techniques were employed to demonstrate

the functionality of the composite particles.

Materials

Poly(acrylic acid) sodium salt was obtained from Scientific Polymer Products in

molecular weights of 2,100 and 6,000; it was obtained from Aldrich Chemical in

molecular weights of 15,000, and 30,000. Iron (II) sulfate heptahydrate (Aldrich

Chemical) was dissolved in deionized water to make a solution with a final concentration

of 0.255 M, which corresponds to an excess ion concentration of [Fe2+ ex = 5 x 10-3

During the optimization process it was found that this concentration seemed to work best

and be the most reliable to get adequate conversion of gel to magnetite. A concentrated

potassium hydroxide (KOH) solution was purchased from Acros Organics and diluted to

a 0.5 M solution. Potassium nitrate, KNO3 (Fisher Scientific), was dissolved in deionized

water to make a solution with a final concentration of2.0 M. Although the KOH and

KNO3 Solutions must be mixed in equal quantities before the particle synthesis, they were

stored separately and kept as stock solutions. The FeSO4 Solution was remade just before

each synthesis trial because it begins to precipitate out of solution after sitting for a few

days.

A deoxygenation solution was also necessary to remove trace amounts of oxygen

and carbon dioxide from the nitrogen gas that was used to purge the solutions during

synthesis. A pyrogallol (C6H3(OH)3) and sodium hydroxide (NaOH) solution mixture

was used for this purpose. A 5.0 M NaOH solution was prepared, and then solid










pyrogallol was added to create a 1.0 M concentration. This solution was placed into a

three-necked flask that was subsequently sealed at all three necks with septa.

A fluorescein probe was used for the fluorescence study, specifically 5-

(aminoacetamido) fluorescein. Another chemical, 1 -ethyl-3-(3-dimethylaminopropyl)

carbodiimide, or EDAC, was also necessary for the fluorescence study. Both were

obtained from Molecular Probes, Inc.

Methods

Setup

In order to perform the synthesis, it was necessary to devise a setup that could

separately purge both the FeSO4 Solution and the KOH/KNO3 Solution, and also purge

the precipitated Fe(OH)2 gel with N2 for ten minutes into the aging process. To do this,

the pyrogallol solution was poured into a three-necked flask. The necks were plugged

with septa and then the septa were punctured through with 18-gauge septum needles.

The needles from the two outside necks were then punctured through other septa that

were sealing vials--one containing the FeSO4 Solution and one containing the

KOH/KNO3 Solution. The needle in the middle neck of the flask served to deliver the

nitrogen from the gas cylinder to the pyrogallol solution. When the gas is turned on, it

flows through the needle in the middle neck and bubbles through the pyrogallol solution,

thus removing trace amounts of oxygen and carbon dioxide. The gas then escapes

through the other two side-neck needles to bubble through the iron sulfate and potassium

hydroxide/nitrate solutions. This is what is referred to as the purging period, and this is

also what creates an inert atmosphere within the reaction vials. A picture of this setup is

shown in the figure below.
































Figure 3-2: Digital picture of setup.

Synthesis

To begin the synthesis, 8 mL of the FeSO4 Solution is added to a vial. 1 mL of the

KOH solution and 1 mL of the KNO3 Solution are added to another vial. The

poly(acrylic acid) is dissolved into the KOH/KNO3 Solution (Table 3-1 shows the

systematic trials performed using PAA). These vials are then plugged with septa and the

appropriate needles are inserted. The flow of gas is then turned on and the solutions are

in this way purged of oxygen for two hours. After the two-hour purging period, a syringe

with a septum needle attached is used to extract the 8 mL of iron sulfate solution from its

vial. It is important to note that the syringe was filled and depressed with nitrogen a few

times before being filled with solution. The solution was then injected into the vial

containing the KOH/KNO3 Solution, thus precipitating the greenish Fe(OH)2 amOrphous

gel. The nitrogen is bubbled for an additional minute before the vial is placed into a

900C oil bath. Once placed in the oil bath, nitrogen is bubbled for another ten minutes,









then the needle is removed and the gel is left to age for approximately four hours. It is

during this aging process that the magnetite nucleates and grows to form spherical

particles. After the aging process is Einished, the vial is removed from the oil bath and

allowed to cool to room temperature until separation.

Table 3-1: Experimental design of trials with poly(acrylic acid)
Molecular Quantity Added Concentration of
Weight (mg) Polymer (ug/mL)
2,100 1.3 130
2,100 2.2 220
6,000 1.0 100
6,000 2.0 200
15,000 1.0 100
15,000 2.3 230
30,000 1.0 100
30,000 2.0 200

The table above outlines the experimental design of the trials with PAA. Attempts

at consistency were made to use concentrations of 100 and 200 ug/mL for each molecular

weight, however this was difficult to actually do. The reason is that measuring very

small quantities of polymer to exact specifications is hard to do--especially when the

polymer is in a powder form. Therefore, I did my best to get as close to these quantities

as possible, but as Table 3-1 shows, some of the trials were slightly off this mark. For the

purposes of this proj ect, however, these small variances are not a significant concern.

Separation

A method to separate the particles of magnetite from possible contaminants and

unwanted byproducts is necessary because of the nature of the reaction. Poly(acrylic

acid) effectively works to inhibit the crystallization process by not allowing primary

particles to aggregate freely. By this same notion, the inhibition is what allows the PAA

to incorporate into the particles because it traps itself between primary particles during









the aging process. Therefore, after the aging there is likely to be present some

unconverted gel. Also, it is so extremely difficult to eliminate all traces of oxygen that

some goethite rods will inevitably be formed.

So, to separate the magnetite from these unwanted materials, a magnet is held to

the vial to attract the particles. Only the particles will be attracted because goethite is

antiferromagnetic with no net magnetic moment, and remnant gel is obviously not

magnetic because it is amorphous. Therefore, with the particles held in place, the

solution is decanted. This process is repeated several times after washing with ethanol.

Once the particles have been adequately separated, the solution is poured into a

centrifuge tube. The sample is then centrifuged at 6,000 rpm for 3 minutes. After the

ethanol is decanted, a kimwipe is placed over the top of the tube and secured with a

rubber band. The kimwipe is used to prevent dust or other contaminants from falling into

the tube. The sample is then set out in air to dry completely. After drying is complete,

the sample is extracted and placed in a small glass storage vial.

Characterization Results and Discussion

Transmission Electron Microscopy (TEM)

Once the particles were synthesized, the next step was to analyze their properties.

The first method of analysis was viewing them on the transmission electron microscope

(TEM). It was important to do this before any other characterization technique so that it

could be confirmed that the desired morphology was obtained. The TEM was used at an

accelerating voltage of 200 kilovolts.

To prepare the particles for TEM analysis, they had to be mounted onto a grid.

Formvar coated copper grids were obtained from Ted Pella, Inc. A small amount of the

dry powder of particles was placed into a small (10 mL) beaker. They were then










dispersed in ethanol and sonicated to achieve a better dispersion. Some of the solution

with dispersed particles was then micropipetted onto a grid. The grid was viewed under

an optical microscope to confirm that an adequate quantity of particles were deposited,

and then the grid was placed in an oven at 600C for 3-5 minutes to help the particles set

into the Formvar coating. The grid was passed over with a magnet to remove any loose

or excess particles.

First and foremost, we must look at the control trials so that we have a point of

reference for the composite particles. Figure 3-3 shows TEM micrographs of a control

sample and Figure 3-4 shows the electron diffraction pattern obtained from these

particles. The diffraction pattern was indexed and labeled, showing that the pattern is

characteristic of Fe304.


2 microns


Figure 3-3: TEM micrographs of control particles. The top two bright field micrographs
were color inverted to better show spherical shape. The particles are closely
uniform in size, ranging from about 300-500 nanometers in diameter.
























Figure 3-4: Electron diffraction pattern of control sample. This pattern was indexed and
found to correspond to magnetite. Note that the presence of diffracted rings
rather than spots indicates polycrystalline nature. This may or may not mean
that individual particles are polycrystalline because the electrons were
diffracting off of a group of particles, not just one.

Now the granular, composite nature of the particles synthesized with the

poly(acrylic acid) additive will be shown. The following figures display the TEM

micrographs obtained for each of the samples with PAA. Each separate figure represents

the results of a different molecular weight of polymer. However, each figure in itself

shows the results for varying concentrations of polymer at its given molecular weight.


Figure 3-5: TEM micrographs of particles with PAA, MW = 2, 100. Both of the pictures
above show particles synthesized with a PAA concentration of 130 ug/mL.
The sample with a concentration of 220 ug/mL did not convert during the
aging process and thus no particles were formed.
































Figure 3-6: TEM micrographs of particles with PAA, MW = 6,000. The top two images
are taken from the sample with a PAA concentration of 100 ug/mL. The
bottom two images had a PAA concentration of 200 ug/mL.


Figure 3-7: TEM micrographs of particles with PAA, MW = 15,000. The top two images
are taken from the sample with a PAA concentration of 100 ug/mL. The
bottom two images had a PAA concentration of 230 ug/mL.









SCi97


Figure 3-8: TEM micrographs of particles with PAA, MW = 30,000. The top two images
are taken from the sample with a PAA concentration of 100 ug/mL. The
bottom two images had a PAA concentration of 200 ug/mL.

By looking at any of the TEM images of composite particles with PAA, it can be

seen that they have a grainy or almost fuzzy-looking texture. This is indeed the

morphology indicative of the nanocomposite particles. But the question arises: what

causes this? Although there is no proven or concrete explanation, these results point to

the fact that the primary particles of magnetite aggregate during synthesis. Once the

primary particles form, they encounter other primary particles and through a surface

recrystallization mechanism they grow to a larger particle. This process occurs between

many primary particles to generate what becomes the final sphere. When a

macromolecular additive such as poly(acrylic acid) is present, it gets in the way of some

of these primary particles and partially inhibits their surface recrystallization. Because

other particles are able to aggregate, these molecules get trapped and therefore









incorporated into the final particle. This is why a grainy morphology is observed in the

TEM images; the primary particles just are not allowed to coalesce like they want to, thus

the final particle achieves a composite morphology.

The acidic nature of PAA is also a contributing factor to the incorporation of the

molecules into the particles. In solution, the -COOH groups easily loose their hydrogen,

thus giving the molecule a negative charge. The positively charged iron ions will create

an attraction and better facilitate the incorporation of the polymer into the iron oxide

particle.

Examination of the images in Figure 3-3 shows that the control reaction produces

relatively uniform, spherical crystals of Fe304 without a granular or fuzzy appearance. In

contrast, examination of all samples with the PAA additive reflects a grainy, composite

morphology. This gives a clear distinction between the control and PAA particles, but

now the differences and trends among the PAA samples will be discussed. In examining

the two trials that used a molecular weight of 2, 100, it makes sense that the sample with a

higher concentration of PAA did not convert during aging. The lower concentration

sample that did convert appears to be very grainy as it is, suggesting that it had probably

approached a maximum limit for PAA incorporation. Therefore, the sample with the

greater quantity of PAA was likely overwhelmed with polymer and it was too much for

the primary particles to overcome and aggregate together to form final particles. In this

instance, the PAA additive worked as a complete inhibitor of the particle synthesis. It

makes sense too because there would be so many molecules present in solution as

compared to a higher molecular weight polymer at the same concentration. This is why

we see later that this concentration, which did not work for a MW of 2,100, did work for










higher molecular weights. If you look at the right-side picture of Figure 3-5, the bottom

particle has a string-like structure emanating from it, perhaps showing the best visual

evidence from all of the micrographs that the polymer is indeed incorporated into these

particles. However, care must be taken in making these statements because the picture

alone does not definitively confirm that this structure is the poly(acrylic acid) additive.

Now turning attention to Figure 3-6 and the trials with a molecular weight of 6,000,

we see that the lower concentration trial yielded much more particle-like structures. The

images of the higher concentration sample show much less defined particles. This makes

sense from observation of the synthesis because this second trial did not react to

completion during the aging, leaving a significant amount of unconverted gel. However,

some magnetic attraction was qualitatively observed, so the synthesis was not completely

inhibited, and this seems to be what the images are showing. I will here say a word about

when the reaction does go to completion. This statement would imply that 100% of the

precipitated gel would be converted to particles of magnetite during aging. This is in fact

not the case. Although it can be clearly observed when the reaction is significantly

inhibited, a 'complete' reaction will still inevitably contain some amounts of unconverted

gel in addition to some unwanted byproducts, namely the goethite rods. The presence of

the rods can even be seen in some of the TEM images in the figures above. For the

purposes of this discussion, and even in the presentation of the results, a trial synthesis

will be referred to as complete if most of the gel is converted and the resulting precipitate

is black, the color characteristic of the magnetite product. Trials that did not convert very

well or even at all had a brownish or greenish color.









Observation of the images in Figure 3-7 shows that both trials with MW=15,000

produced granular composite particles. It is difficult to make any sweeping statements

about whether or not there are any significant differences between these samples just

based on their TEM micrographs. In contrast, it appears as though there is a difference

between the two samples with MW=30,000. It seems that the trial with the higher

concentration produced grainier particles. This makes sense because it would be

expected that a larger concentration of polymer in solution would allow more of it to be

incorporated into the particles, thus creating a more granular composite structure.

But are there any observable trends based on changing the molecular weight? It

has already been discussed how a large concentration of a low molecular weight polymer

can inhibit synthesis, but there will also be a limitation when the molecular weight gets

too high. In solution, a polymer chain takes a random coil conformation. As the

molecular weight of this chain is increased, the hydrodynamic volume of the random coil

becomes larger and larger. Once this gets to a large enough size, it becomes too difficult

for the primary particles to aggregate together and pull this large molecule into the final

particle. In this way, a higher molecular weight polymer becomes an effective inhibitor

to the synthesis. Therefore, it was important in this research to try to find an interplay

between two extremes. The higher concentration trial in the 30,000 molecular weight

series seems that it may be approaching this upper limit. The particles appear as though

they are getting grainy to the point that a definitive spherical particle is becoming hard to

establish. Therefore, it seems that the trend in all of these PAA trials based on molecular

weight considerations is that MW=2, 100 produces somewhat grainy particles, and then

the particles become more granular as molecular weight is increased to 6,000 and then










15,000, but then they seem to begin to come to a limit at MW=30,000 where they start to

loose their spherical particle morphology. Of course the polymer concentration is an

important factor that can affect the synthesis at any of these molecular weights, but this

seems to be the general trend for a reasonable concentration of PAA.

X-Ray Diffraction (XRD)

X-ray diffraction is a characterization technique that utilizes the Bragg Law,

nh=2dsin6, where h is the wavelength of the x-rays, d is the spacing in a certain

crystalline plane, and 6 is the angle in which the x-ray is diffracted. Depending on the d-

spacing of the crystal lattice, incident x-rays are diffracted at different angles and then

counted to derive a spectrum of intensity versus angle 26. Standards of data have been

compiled and stored in large databases so that an experimental sample can be compared

to a standard in order to identify the material and/or phase. For this analysis, a standard

for magnetite was obtained and plotted along with the experimentally collected data.

Sample preparation for XRD was relatively simple since the particles were already

in a powder form. A small piece of double-sided tape was adhered to a glass slide. The

particle powder was then deposited on top of the tape and spread out to cover the entire

surface of the tape. This step is important in order to ensure that a large enough area will

be exposed to the x-rays during data collection. A spatula was then used to press the

powder onto the tape so that it would not be blown off. Incidentally, the PAA sample

used in this analysis was the one using MW=6,000 and concentration 100 ug/mL. It is

not necessary to analyze every particle sample with this technique because as long as the

synthesis procedure was kept constant, the product should be the same so long as the

reaction went to sufficient completion.

















Control (No Polymer Additive)


I~III


1001*



801*




401

201


1~111
11~~1


11??1
11111


20 25 30 35 40 45 50
2 Theta (degrees)
-Standard -Control


55 60 65 70


Figure 3-9: X-ray diffraction spectrum of control particles. The data for the magnetite

standard is also plotted for comparative purposes. The peaks are labeled with

their corresponding planes of diffraction.





Poly(Acrylic Acid) Additive


1)111


111'11


11111


20 25 30 35 40 45 50
2 Theta (degrees)
-Standard PAAMW=6.000


55 60 65 70


Figure 3-10: X-ray diffraction spectrum of composite particles with PAA additive. The

data for the magnetite standard is also plotted for comparative purposes. The

peaks are labeled with their corresponding planes of diffraction.


X-ray diffraction is a very useful technique used to characterize a material's crystal



structure. It was used in this instance to confirm that the crystalline phase of magnetite is


~01


1111111

I:??I









indeed present as the product of particle synthesis. Examining Figure 3-9 shows that the

control's sample peaks match up very well with the peaks for the magnetite standard in

both intensity and angle 26. This confirms that the control sample is Fe304, aS expected.

Upon examination of the spectrum for particles with PAA we can see that the peaks

again match up very well and confirm the presence of magnetite. However, it seems that

the peaks for this sample are a bit broader than the peaks for the control. The reason for

this is that because the polymer is present in the mineral, it causes strains in the

crystalline lattice. These strains cause slightly different deflections of the x-rays and thus

serve to create broader peaks. So in effect, this result further supports what the TEM

images show--that a composite particle morphology has been created with the addition

of poly(acrylic acid). X-ray diffraction was not performed on every sample because it

would only be superfluous and unnecessary to do so. If the synthesis went to completion

and formed particles, then this proves that those particles are indeed magnetite.

Superconducting Quantum Interference Device Analysis (SQUID)

This technique was used to create a hysteresis curve for each of the samples. The

hysteresis curve yields many important magnetic quantities; you can determine saturation

magnetization (Msat), remnant magnetization (Mrem), and coercivity (He). It is important

to quantify such values if an intended application of the particle system makes use of its

magnetic moment. In addition, magnetic properties need to be characterized so that they

can be compared to other magnetic systems. A full hysteresis loop was produced for the

first few samples until it was realized that the particles show symmetric behavior.

Therefore, subsequent samples were only run for half of a loop. The maximum applied

field in each trial was 7 Tesla. The data are plotted as magnetization versus applied field.









Sample preparation was a somewhat intensive procedure. Because the

magnetometer is so sensitive, very accurate sample weights were necessary to reduce

error. However, a 5% error was still given to each trial based on weight. The particle

sample had to be carefully put into a gelcap and then tightly compacted with a kimwipe.


|M vs H, Particles with No Additive (Control) |


T = 300 K


100


80


0 2 4 6 8
B (T)


Figure 3-11: Hysteresis curves for control particles.


Table 3-2: Magnetic quantities determined from SQUID analysis of control particles
Saturation Remnant
Coercivity, He
Sample Magnetization, Magnetization,
(Gauss)
Msat (emu/gram) Mr, em 0 ELSE)
No Additive
78.6 1.5 33
(Control)
No Additive
70.4 2.4 35
(Control)












M vs H, Particles with Poly(Acrylic Acid) T = 300 K
80










a) -v-3 MW=2,100; 130 ug/mL
-6- MW=6,000; 100 ug/mL
20 -o-l MW=15,000; 100 ug/mL -
-0- MW=15,000; 230 ug/mL
-0 MW=30,000; 100 ug/mL
-<1 MW=30,000; 200 ug/mL




0 2 4 6 8
B (T)


Figure 3-12: Hysteresis curves for PAA modified composite particles.


Notice in the figure above that there is no plotted data for the trial that used

MW=2,100 and a concentration of 220 ug/mL. This is because that trial did not convert

during the aging process; instead, there was only unconverted gel remaining after the

aging. There is also no curve for the trial with MW=6,000 and concentration 200 ug/mL.

This is because the sample only converted a small amount during the aging process.

Much of the final volume after synthesis was unconverted gel and thus a hysteresis curve

of this sample would not have been representative of a composite particle system. It is

important to stay consistent in presenting the data so only the samples where the reaction

went to significant completion are presented here.









Table 3-3: Magnetic quantities determined from SQUID analysis of PAA particles
Saturation Remnant
Coercivity, He
Sample Magnetization, Magnetization,
(Gauss)
Msat (emu/gram) Mr, em 0 ELSE)

MW=2, 100
54.2 2.2 40
130 ug/mL
MW=6,000
57.5 4.1 60
100 ug/mL
MW=1 5,000
67.1 2.9 30
100 ug/mL
MW=1 5,000
61.3 5.7 62
230 ug/mL
MW=30,000
39.9 2.6 34
100 ug/mL
MW=30,000
46.2 3.6 55
200 ug/mL

A good amount of information can be obtained about the nature of a magnetic

material just by the shape of its hysteresis curve. By examining the hysteresis curves for

both the control and PAA samples, we can see that they both have the same general

shape. This implies that the formation of composite particles does not significantly

change magnetic behavior, but only the magnetic quantities. The particles show soft

magnetic behavior, meaning that they are easily magnetized without having to apply a

huge magnetic field. This is important for many applications, especially biomedical

ones, because many techniques would be easier if a lower Hield could be applied. In

magnetic resonance imaging, for example, current contrast agents are commonly

magnetized with a 1.5 Tesla Hield. Examining the data shows that these particles begin to

reach their magnetization limit well before this. The curves also show that the particles

have a low coercivity. This means that only a small reverse in the magnetic Hield is

necessary to bring the particles back to a zero magnetization. Coercivity values for the

control and PAA particles range from 30 and 60 Gauss. The remnant magnetization









values for both the control and PAA particles are in about the same ballpark as well.

Their values range from about 1.0 to 5.5 emu/gram, which is small relative to their

saturation magnetizations. (It should here be pointed out that the remnant magnetization

is what may cause aggregation problems among the particles. The fact that the particles

will retain some magnetic moment implies that they will be attracted to each other when

they get in the vicinity of one another. This is an issue that needs to be addressed in

future work so that it does not cause a problem in application, particularly one that would

use the particle system in vivo.) The hysteresis curves and their corresponding tables

show that the Msat and Mrem ValUeS do not seem to be affected in a definitive manner

when a polymer additive is present, but it is hard to say conclusively whether or not there

is a significant effect based on this data. In any case, the most interesting values from

this analysis and the focus of this discussion are the saturation magnetizations.

The figures and their corresponding tables show that the control particles achieve a

higher saturation magnetization than any of the PAA composite particles. This makes

sense too since defects and impurities in the form of the polymer additive are being

introduced to the particles. The polymer molecules have no magnetic properties, thus on

a per weight basis they would be expected to degrade the overall magnetization. There

has been evidence, however, that defects formed in a magnetic lattice can possess

magnetization in and of themselves. Whether or not this is the case with the composite

particles has yet to be determined, however if there is a contributing factor from these

defects it is not very strong. Magnetite inherently gives a very strong signal because it

has many unpaired electrons which create a strong magnetic moment; therefore, its

intrinsic signal would likely drown out any small contributions from defects.









Now that it has been established that the control particles achieve a higher Msat, we

will now explore the differences and trends among the PAA trials. Figure 3-12 shows

that the two trials with MW=15,000 have the highest Msat, and the two trials with

MW=30,000 have the lowest. The other two samples of MW=2,100 and MW=6,000

have values for Ms,, in between the others. Since the three lowest values are the

MW=2,100 and MW=30,000 trials, this could imply that these molecular weights are

near the upper and lower limits of allowable molecule sizes for incorporation during

synthesis. Anything too far outside these limits may cause complete inhibition and

prevent any conversion to magnetite during the synthesis. This would mean that an

intermediate molecular weight such as 15,000 may be ideal for incorporation and yield

the best particles. This idea is indeed supported by the data because (as just mentioned)

the MW=15,000 trials had the highest values for saturation magnetization. Although this

may be a reasonable assumption, it is in no way a definitive conclusion. In fact, the basis

of this reasoning is somewhat contradicted when trends in concentration are examined.

The reasoning is based on what degree the additive inhibits the synthesis. It would

therefore be expected that a higher concentration at the same molecular weight would

further inhibit a synthesis and result in a lower saturation magnetization. This is indeed

what is observed for the MW=15,000 series, but the exact opposite is observed for the

MW=30,000 series. The best explanation for this is just simply that the synthesis is

either not controlled as much as we would like because of the additive, or that the

synthesis yield is not completely uniform. Therefore, it seems the best conclusion that

can be made is only to qualitatively say that PAA composite particles will fall in the

range of about 40-70 emu/gram. The bright side is that this is sufficient enough to say









that since the particles are in this range they are still better than many of the other

magnetic particle systems out there.

Fluorescence Labeling

Once the particles have been well characterized, it is important to demonstrate their

functionality. Fluorescence microscopy combined with bioconjugation techniques is very

useful for labeling desired targets. In the case of these composite particles, the labeling

target was the poly(acrylic acid). Specifically, the carboxylic acid functional group was

utilized to attach a fluorescent probe. The ability to conjugate this probe to the particles

and observe its fluorescence demonstrates that the polymer is indeed functional and can

therefore be conjugated to other probes. This is an important step toward producing these

particles for a specific application.

Procedure

The protocol for this labeling procedure was taken from a study that labeled the

carboxyl groups hanging off of a poly(methyl methacrylate) substrate [69]. This method

was actually a modification of another protocol [70], but bioconjugation chemistries for

fluorescence analysis in general have been well studied [71]. The fluorescein probe will

not conjugate to the carboxylic acid group of the poly(acrylic acid) on its own. Instead, a

linker is necessary. This linking molecule, EDAC, is a water-soluble carbodiimide that

reacts with the amine group of the fluorescein and the -COOH group of the PAA. It is

by this reaction that the fluorescein is conjugated to the PAA. The actual procedure to do

this is as follows. The EDAC and fluorescein were dissolved in a 100 mM phosphate

buffer solution (pH 7) to make a concentration of 0.5 mM each. A few milligrams of

PAA composite particles were put into a small vial and covered with the solution. The

vial was then gently agitated in darkness for a 15-hour incubation period. When this was










complete, the sample was washed with phosphate buffer four times to remove excess

fluorescein molecules that did not conjugate. While suspended in phosphate buffer, the

particles were micropippetted onto a glass microscope slide and a cover slip was placed

on top. A mercury lamp and burner were used in conjunction with an optical microscope

to perform the fluorescence microscopy. Digital images were taken with a microscope-

mounted camera.

Fluorescence microscopy

It was necessary to again run a control with this protocol to ensure that there was

no fluorescence associated with particles that did not contain the PAA. This is indeed the

result shown in Figure 3-13 below.














Figure 3-13: Transmission and fluorescence micrographs of control particles. The light
transmission image on the left shows clumps of particles dispersed on the
slide. The fluorescence micrograph on the right shows the fluorescence (or
lack thereof) of that same area.

Fluorescence microscopy is a very effective means to demonstrate the functionality

of the poly(acrylic acid) in these composite particles. The specific labeling chemistry

used in this research will only target and conjugate to carboxylic acid functional groups.

This functionality is only afforded to the particles by the presence of PAA. Therefore, it

would not be expected for the control sample to fluoresce in any way because no PAA



























































Figure 3-14: Fluorescence images of PAA composite particles. The first two images
come from the sample with MW=15,000 and concentration 100 ug/mL. The
third image comes from sample MW=6,000 and concentration 100 ug/mL.









exists in the particles. The images in Figure 3-13 confirm that this is true. However, the

images in Figure 3-14 do show areas of fluorescence because these are images of

composite particles with the PAA additive. It seems that only parts of the sample are

fluorescing, and this is probably for a few reasons. The first and most likely explanation

is that the particles are not completely uniform in their PAA content. There is evidence

of this in the TEM images where we can see that some particles are grainier than others

within the same image. Also, the PAA may not be completely surface accessible to the

fluorescent probe. Some of the chains of PAA will be wedged into the particle too far

and thus their functional groups are rendered useless. However, the fact that we do

observe fluorescence in the particles at all demonstrates that composite particles are

indeed present and their organic functionalities are intact and reactive.















CHAPTER 4
PARTICLE SYNTHESIS WITH POLY-L-GLUTAMIC ACID ADDITIVE

Introduction

Poly-L-glutamic acid is a poly amino acid. Like poly(acrylic acid), it is an acidic

polymer with a carboxylic acid functional group on every repeat unit. Its structure is

shown in the Eigure below.




~~ NH-C -C-

Figure 4-1: Repeat unit structure of poly-L-glutamic acid.

This polymer was chosen for use as an additive for two main reasons. The first is that it

has the -COOH functional group. This makes it similar to PAA in that it has the same

functionality but with a different backbone structure. The second reason for choosing

poly-L-glutamic acid (which will also be referred to as GLU) is because it has this

backbone that is characteristic of all poly amino acids and proteins. Since a possible

future goal with these particles is to incorporate specifically sequenced polypeptides, it is

an important step to Birst try to incorporate a homopolymer of a poly amino acid.

Therefore, poly-L-glutamic acid seemed like a fitting choice for this obj ective.

The range of molecular weights explored with this additive was smaller than what

was used for the PAA trials. The reason for this is that it is significantly more expensive

to produce and consequently purchase poly amino acids. Therefore, only two molecular

weights were used in synthesis trials with the GLU additive, however varying

concentrations were also explored.









This chapter will present the results of the GLU additive trials using the same

characterization techniques utilized in the previous chapter. Transmission electron

microscopy (TEM) will show the composite particle morphology, x-ray diffraction will

confirm the presence of magnetite, and superconducting quantum interference device

analysis (SQUID) will show the magnetic properties of the composite particles.

Comparisons to control particles or PAA composite particles will be made where

necessary. In addition, fluorescence labeling was performed on these particles because

they have carboxylic functional groups. The presence of fluorescence demonstrates that

the organic functionality of the composite particles remains intact.

Materials

Poly-L-glutamic acid sodium salt was obtained from Sigma in molecular weights of

7,500 and 13,600. The rest of the materials used for these trials are the same as those

used for the PAA trials described in chapter 3. The following is a brief summary of those

materials. Iron (II) sulfate heptahydrate (Aldrich Chemical) was dissolved in deionized

water to make a solution with a final concentration of 0.255 M, which corresponds to an

excess ion concentration of [Fe2+ ex = 5 x 10-3. A concentrated potassium hydroxide

(KOH) solution was purchased from Acros Organics and diluted to a 0.5 M solution.

Potassium nitrate, KNO3 (Fisher Scientific), was dissolved in deionized water to make a

solution with a final concentration of2.0 M. The KOH and KNO3 Solutions were stored

separately and kept as stock solutions. The FeSO4 Solution was remade just before each

synthesis trial because it begins to precipitate out of solution after sitting for a few days.

A pyrogallol (C6H3(OH)3) and sodium hydroxide (NaOH) solution mixture was used as a

deoxygenation solution. A 5.0 M NaOH solution was prepared, and then solid pyrogallol

was added to create a 1.0 M concentration. This solution was placed into a three-necked









flask that was subsequently sealed at all three necks with septa. A fluorescein-based

probe was used for the fluorescence labeling study, specifically 5-(aminoacetamido)

fluorescein. Another chemical, 1 -ethyl-3-(3-dimethylaminopropyl) carbodiimide, or

EDAC, was also necessary for the fluorescence study. Both were obtained from

Molecular Probes, Inc.

Methods

The method for particle synthesis--including the setup and separation

procedures--were the same for the trials with poly-L-glutamic acid as they were for the

PAA trials. In addition, the fluorescence labeling technique is also the same. A quick

synopsis of these methods follows, but the most complete description can be found in the

Methods section of chapter 3.

The pyrogallol solution was poured into a three-necked flask. The necks were

plugged with septa and then the septa were punctured through with 18-gauge septum

needles. The needles from the two outside necks were then punctured through other

septa that were sealing vials--one containing the FeSO4 Solution and one containing the

KOH/KNO3 Solution. The needle in the middle neck of the flask served to deliver the

nitrogen from the gas cylinder to the pyrogallol solution. A picture of this setup can be

seen in Figure 3-2 of the previous chapter.

For the actual synthesis, the FeSO4 and KOH/KNO3 Solutions were purged with

nitrogen for two hours. 8 mL of the FeSO4 WAS then inj ected into 2 mL KOH/KNO3

solution, precipitating the Fe(OH)2 gel. The GLU additive was dissolved in the

KOH/KNO3 Solution prior to the two-hour purging with nitrogen. This vial was then

aged in a 900C oil bath for four hours to create the Fe304 HanOparticles. The particles









were separated out from contaminants and solution using a handheld magnet and then

centrifugation. The particles were completely dried and stored as a powder.

Table 4-1: Experimental design of trials with poly-L-glutamic acid.
Molecular Quantity Added Concentration of
Weight (mg) Polymer (ug/mL)
7,500 0.8 80
7,500 1.0 100
13,600 0.6 60
13,600 1.0 100
13,600 1.5 150

The table above shows the experimental trials performed using the GLU additive.

Concentrations only ranged from 60 to 150 ug/mL because it seemed the reaction would

only go to completion within these limits. It was during the optimization of this synthesis

when it was found that the reaction would only seem to produce composite particles

using these molecular weights when a concentration in this range was used. Therefore,

any possible trends discussed in the results section will be based on these trials.

Although only two molecular weights were explored, it is still sufficient enough to see

that composite particles are indeed formed using the poly-L-glutamic acid additive.

Characterization Results and Discussion

Transmission Electron Microscopy (TEM)

The first step toward characterizing the composite particles is performing TEM to

show their composite morphology. A TEM grid was prepared for each sample by

dispersing some particles in ethanol and then depositing them on a formvar coated copper

grid using a micropipette. The grids were placed in an oven for a few minutes to help the

particles better adhere to the formvar coating. The microscopy was performed at an

accelerating voltage of 200 kilovolts.
















.0 00n nm






I~rI







Figure 4-2: TEM micrographs of particles with poly-L-glutamic acid, MW = 7,500. The
top two images are taken from the sample with a GLU concentration of 80
ug/mL. The bottom two images had a GLU concentration of 100 ug/mL.

Examination of the images in the figure above reveals that the morphology

indicative of composite particles is obtained. We can see the grainy texture of the

particles created by the GLU additive disrupting the normal aggregation and

recrystallization mechanism of the primary particles. Thus, the TEM analysis gives

evidence that the polymer additive has indeed been incorporated into the magnetite

particle. The images also show that the particles are in the size range of 150-300

nanometers. This size is small enough so that the particles can flow through blood

capillaries and large enough so that they are not superparamagnetic.

It is difficult to say whether or not there is a trend with these trials based on the

images. The main reason is that there are only two samples at this molecular weight, and

more samples would be needed to make a more conclusive statement about trends.

However, it appears as though the particles with a concentration of 100 ug/mL are










slightly more granular than those with a concentration of 80 ug/mL. This would make

sense because a higher concentration would imply that more GLU additive is present and

thus has a better chance of more incorporating into the particles. So even if it is too

difficult to make a conclusive statement about the effect of concentration, it seems that

these images at least support what would be expected to happen. And most importantly,

these images demonstrate that the composite morphology is indeed obtained, which was

the primary information desired of the TEM analysis in the first place.


Figure 4-3: TEM micrographs of particles with poly-L-glutamic acid, MW = 13,600. The
top two, middle two, and bottom two images have concentrations of 60
ug/mL, 100 ug/mL, and 150 ug/mL, respectively.









The images in Figure 4-3 seem to show subsequently grainier particles as the

concentration of polymer is increased. This is intuitively what would be expected to

happen, as more GLU additive would be available for incorporation into the particles if

its concentration were higher. However, it is hard to say whether or not this is the case

for sure without quantitative data showing the amount of polymer actually incorporated

into the particles. It is important to mention this because if we look at the trial with

MW=13,600 and concentration 100 ug/mL, for example, it seems that the resulting

particles are not as uniform in morphology as the other two samples in that figure. This

could suggest that some particles have incorporated more of the GLU additive than

others, thus making it difficult to say on a whole that this trial contains either more

granular or less granular particles. Therefore, it is not conclusive that a trend is seen

here, however the TEM images give strong support to the affirmative claim. The same

should be said about a trend in molecular weight. Actually, it is hard to say even

qualitatively that the polymer content is increased or decreased due to molecular weight

based on the images in the figures above. And since only two molecular weights were

explored, it makes it even harder to say what the effect might be. It is important,

however, not to lose track of the most important evidence obtained from the TEM

analysis: the nanocomposite particle morphology has indeed been obtained.

X-Ray Diffraction (XRD)

X-ray diffraction is a very useful technique used to determine the crystal structure

of a material. It is used on these composite particles to confirm that the mineral phase is

magnetite (Fe304) and not a different iron oxide or derivative. This technique is

appropriate for this application because the polymer additive is not crystalline so it is not

expected to contribute anything to diffraction. Sample preparation was a straightforward










process of adhering the sample powder to a glass slide via double-sided tape. XRD was

performed on only one of the GLU samples because it is representative of all the GLU

samples given that the reaction went to completion; only those samples where the

reaction did indeed go to completion are presented in this chapter.


Particles with Poly-L-Glutamic Acid





7501* 10










20 25 30 35 40 45 50 55 60 65 70
2 Theta (degrees)
-Standard -GLU MW=7,500

Figure 4-4: X-ray diffraction spectrum of composite particles with GLU additive. The
data for the magnetite standard is also plotted for comparative purposes. The
peaks are labeled with their corresponding planes of diffraction.

As shown in the figure above, the experimental spectrum was plotted along with a

standard for Fe304. The peaks for the sample with the poly-L-glutamic acid additive

match up with the peaks for the standard in both relative intensity and angle. This

confirms that the crystalline phase of the composite particles is magnetite. There seems

to be an extra small peak at about 28 degrees. This does not match up with the standard,

however it does not match up with any of the high intensity peaks for other phases of iron

oxide. There are two likely explanations. Perhaps the standard used for comparison did

not express that plane strongly. However, the most probable cause for this peak is from

the small amount of steel that may have been deposited onto the sample during











preparation. A steel spatula was used to press the powder firmly onto the tape. The

peaks on the spectrmm are also somewhat broader than expected for pure magnetite. This

is likely due to the lattice strains associated with the GLU additive. These strains will

cause the x-rays to slightly diffract differently and thus broaden the peaks.

Superconducting Quantum Interference Device Analysis (SQUID)

This analysis measures the magnetic properties of the composite particles by

creating a hysteresis curve. Since these samples show mirrored behavior at positive and

negative magnetic fields, only half of a full hysteresis loop was obtained for all of the

samples. The curve is generated by slowly applying a magnetic Hield to magnetize the

sample, and then slowly removing and reversing the magnetic Hield until the

magnetization is zero. The maximum applied Hield was 7 Tesla. Samples were prepared

by packing the particle powder into a gelcap, and the curves were given a 5% error based

on weight.



M vs H, Particles with Poly-L-Glutamic Acid T = 300 K
100
-0- MW=7,500, 100 ug/mL
C-a-- MW=7,500, 80 ug/mL
80 *-MW=13,600, 150 ug/mL
-a- MW=13 ,600, 100 ug/mL
E --v- MW=13,600, 60 ug/mL




E 40-


20-





0 2 4 6 8
B (T)


Figure 4-5: Hysteresis curves for GLU modified composite particles.










The values for coercivity and remnant magnetization are not discernable on the plot

above because it is not zoomed in on the origin. However, the zoomed-in plot is not

presented here because there are so many data points in that area that it becomes

confusing to look at. Instead, the values for all of the critical magnetic quantities for each

sample are summarized in the following table.

Table 4-2: Magnetic quantities determined from SQUID analysis of GLU particles
Saturation Remnant
Coercivity,
Sample Magnetization, Magnetization,
He (Gauss)
Msat (emu/gram) Mr, em 0 ELSE)
MW=7,500 80
48.3 2.3 46
ug/mL
MW=7,500
60.0 2.1 41
100 ug/mL
MW=13,600
33.1 1.4 50
60 ug/mL
MW=13,600
34.1 2.7 50
100 ug/mL
MW=13,600
36.4 1.0 33
150 ug/mL

It can be seen from this table that the values for remnant magnetization and

coercivity are approximately in the same range of each other for each of the trials. This

makes sense from the plot because all Hyve of the curves have the same general shape.

The magnetization increases sharply and then begins to level out toward a maximum

value as the Hield strength is increased. Remnant magnetization results from the magnetic

crystals maintaining some of their net magnetic moment after the Hield has been removed.

This property may lead to some particle aggregation problems since they will create an

effect on each other.

The magnetic property that varies most among the trials is the saturation

magnetization. We can see from Figure 4-5 that the three trials with MW=13,600 are









within the error range of each other, effectively showing that there is no significant

difference between them. This could imply many things. It could mean that the

concentration of polymer at this molecular weight has very little or no effect on the

magnetic properties of the particles. It could also mean that in each of the trials the

particles incorporated the same amount of polymer additive. I think that the latter

possibility is on the right track because--as it was noted in the TEM discussion--it is

difficult to say whether or not one trial produced grainier particles as a whole than

another. It is possible that a limit on polymer concentration was reached early (i.e. 60

ug/mL), where any extra polymer in solution would just not be incorporated during

synthesis. But it would seem likely that an excess of additive would better inhibit the

synthesis and perhaps prevent significant conversion during the aging process. It is

impossible to say what the reason is for sure, so further experimentation would be

necessary to elucidate their differences--or lack thereof.

It is clear from the plot that there are significant differences between the

MW=13,600 trials and the MW=7,500 trials. First of all, if we go back to chapter 3 and

look at the control trials, we can see that the controls still have higher values for Msat than

any of the GLU trials. However, the MW=7,500 trials both have a higher Msat than the

MW=13,600 samples, and the one with concentration 100 ug/mL is higher than the one

with concentration 80 ug/mL. This does not make intuitive sense. It would be expected

that the higher concentration sample would have more incorporated additive and thus

have a reduced overall saturation magnetization. A similar result was seen with the PAA

trials in chapter 3, however the more intuitive result was also observed. What this means

is that a definitive trend can not be speculated based on this data for the MW=7,500 trials.









However, it does make sense that the MW=13,600 trials all have lower values because

the additive is a larger molecule. A larger molecule would be more inhibitory to

synthesis, likely creating greater space between primary particles during the aging,

thereby creating particles possessing a lower saturation magnetization. This result is

indeed observed and is really the most conclusive evidence for particle differences based

on this analysis.

Fluorescence Labeling

This method of analysis is being used in order to demonstrate the reactivity and

functionality of the carboxylic acid groups of the poly-L-glutamic acid additive. The

procedure for labeling the GLU modified composite particles is the same procedure

described in chapter 3 for labeling the PAA composite particles. Therefore, only a brief

outline of that protocol will be presented here. The fluorescent probe (5-

(aminoacetamido) fluorescein) is conjugated to the poly-L-glutamic acid chains via a

linker molecule, EDAC. Both of these materials were dissolved in phosphate buffer to

create a 0.5 mM concentrated solution. Particles were then submersed in this solution

and gently agitated overnight to allow coupling. After four subsequent washings with

phosphate buffer, the particles were mounted onto a glass slide and covered with a cover

slip. A mercury lamp and burner were used with an optical microscope to perform

fluorescent imaging. Pictures were captured with a scope-mounted digital camera.

The labeling protocol used for this analysis is specifically targeted to conjugate to

carboxylic acid groups only. Since -COOH groups are the only type of functional group

that exist on the particles (afforded by the poly-L-glutamic acid), then any fluorescence

indicates that these groups are present and reactive. Figure 4-6 supports that this is the

case because we can see that there are areas of fluorescence present in the images.












































Figure 4-6: Fluorescence images of GLU composite particles. Both images come from
the sample with MW=13,600 and concentration 150 ug/mL.

The fluorescence in these images is clearly not blinding by any means, nor does it

seem to be present on all areas of the particle clumps. The reason for lack of complete

ubiquity is that the polymer is likely not sufficiently surface accessible to the fluorescent

probe on all particles; therefore the probe would not have been able to get to all or any of

the additive and conjugate to it. It also makes sense that the fluorescence is not very

strong because--as can be seen in the TEM images--the particles have a very high

mineral loading, i.e., the quantity of polymer additive present in each particle is small. If










the particles had a greater quantity of polymer, like a poly-L-glutamic acid coating, for

example, then a much brighter and widespread fluorescence would be expected. It is

important to note that all observed fluorescence can only be from GLU incorporated in

particles because any excess GLU that may have been left over from synthesis would

have been washed away during both the separation process and the sample preparation.

So, to summarize, the observation of fluorescence does indeed demonstrate the presence

and reactivity of the poly-L-glutamic acid within the magnetite particles.















CHAPTER 5
PARTICLE SYNTHESIS WITH POLY-L-LYSINE ADDITIVE

Introduction

Poly-L-lysine is an amino acid homopolymer with a primary amine as its functional

group. Unlike the polymers used in the previous two chapters, poly-L-lysine is basic in

solution. The structure of its repeat unit is shown in the figure below.


H,NCH,CH2CH,CHe ~HO
~~ NH C C-

Figure 5-1: Repeat unit structure of poly-L-lysine.

This polymer was chosen for use as an additive primarily because of its functionality.

The maj ority of bioconjugation protocols out there seem to make use of an amino group

in order to attach drugs, antibodies, fluorescent molecules, or other probes. Therefore,

the reasoning was simply that the particles would become more useful for conjugation

purposes if they contained a free amine. In addition, the fact that poly-L-lysine is a poly

amino acid lends to the future goal of incorporating sequenced polypeptides.

Only one molecular weight of poly-L-lysine (which will be referred to also as LYS)

was investigated for reasons that will become apparent in the discussion later in this

chapter. However, for the trials that were performed, the samples were characterized

with the same techniques used in the previous two chapters. The only exception to this is

the fluorescence labeling study because the protocol used for the previous two polymer

additives targeted the carboxylic acid functionality, but the LYS additive does not

possess this trait. So, the particles were imaged with transmission electron microscopy










(TEM), measured for magnetic quantities using a superconducting quantum interference

device (SQUID), and the Fe304 phase of the particles was confirmed using x-ray

diffraction (XRD). Similarities to control and/or composite particle samples will be

discussed where necessary.

Materials

Poly-L-lysine sodium salt was obtained from Sigma in a molecular weight of

27,000. The other materials used for the synthesis were the same as those described in

chapters 3 and 4, but for the sake of completeness a brief summary of those materials is

again listed here. Iron (II) sulfate heptahydrate (Aldrich Chemical) was dissolved in

deionized water to make a solution with a final concentration of 0.255 M, which

corresponds to an excess ion concentration of [Fe2+ ex = 5 x 10-3. A concentrated

potassium hydroxide (KOH) solution was purchased from Acros Organics and diluted to

a 0.5 M solution. Potassium nitrate, KNO3 (Fisher Scientific), was dissolved in deionized

water to make a solution with a final concentration of2.0 M. The KOH and KNO3

solutions were stored separately and kept as stock solutions. The FeSO4 Solution was

remade just before each synthesis trial because it begins to precipitate out of solution

after sitting for a few days. A pyrogallol (C6H3(OH)3) and sodium hydroxide (NaOH)

solution mixture was used as a deoxygenation solution necessary to remove trace

amounts of oxygen and carbon dioxide from the nitrogen gas used for purging. A 5.0 M

NaOH solution was prepared, and then pyrogallol powder was added to create a 1.0 M

concentration. This solution was placed into a three-necked flask that was subsequently

sealed at all three necks with septa. The setup and usage of these materials is described

in the following section.









Methods

Just as the materials for these trials were primarily the same as those for the

previous two chapters, so too were the methods for particle synthesis. Refer to chapter 3

for the most complete description of the setup, synthesis, and separation procedures

because only a short description of these methods will be given here.

The pyrogallol solution was poured into a three-necked flask that was then plugged

with septa. The septa were punctured with 18-gauge septum needles and then the needles

from the two outside necks were punctured through other septa that were sealing vials--

one containing FeSO4 Solution and one containing an equal mixture of KOH and KNO3

solution (1 mL of each). The needle in the middle neck of the flask served to deliver the

nitrogen from the gas cylinder to the pyrogallol solution. A picture of this setup is in

Figure 3-2 of chapter 3.

For the actual synthesis, the FeSO4 and KOH/KNO3 Solutions were purged with

nitrogen for two hours. 8 mL of the FeSO4 WAS then inj ected into the 2 mL KOH/KNO3

solution, precipitating the Fe(OH)2 gel. The poly-L-lysine additive was dissolved in the

KOH/KNO3 Solution prior to the two-hour purging period. This vial was then aged in a

900C oil bath for four hours to create the Fe304 HanOparticles. The particles were

separated from contaminants using a handheld magnet and then centrifugation, using

ethanol as a wash. The particles were completely dried in air and stored as a powder.

The following table shows the experimental trials that were performed using the

poly-L-lysine. Only two trials were analyzed with this additive because--as we will see

in the results--particles did not seem to form with the desired morphology. However, the

results of these trials still give useful insight to the synthesis process and are therefore









worthwhile to analyze. A more in-depth discussion will come with the presentation of

the results in the following section.

Table 5-1: Experimental trials with poly-L-lysine.
Molecular Quantity Added Concentration of
Weight (mg) Polymer (ug/mL)
27,000 1.0 100
27,000 2.5 250

Characterization Results and Discussion

Transmission Electron Microscopy (TEM)

TEM was used to obtain images of the particles and show their morphology. To

prepare a sample, some particle powder was dispersed in ethanol by sonication, then

pippetted onto a formvar coated copper grid. The grids were placed in an oven for a few

minutes to help the particles better adhere to the grids. The microscope was operated at

an accelerating voltage of 200 kilovolts.






















Figure 5-2: TEM micrographs of particles with poly-L-lysine, MW = 27,000. The top
two images are taken from the sample with a LYS concentration of 100
ug/mL. The bottom two images had a LYS concentration of 250 ug/mL.









The images above seem to indicate that the LYS additive has not been incorporated

into the particles because the composite particle morphology is not observed. They are

well-defined particles without a grainy or fuzzy-looking texture (for the most part), a

morphology that is reminiscent of the control particles. In fact, the two images on the

right were the most granular-like particles that could be found on the TEM grid for each

of the two trials. The pictures on the left are the most representative of the entire sample.

So the question arises: why did this occur? The answer probably lies in the fact

that poly-L-lysine is basic in solution. Recall that in the case of the PAA and GLU

additives the polymer chains had a negative charge due to the acidic carboxyl group.

This negative charge likely helped attract the additive molecules to the positively charged

iron ions during the particle synthesis. Therefore, the opposite probably occurred with

the LYS additive, which becomes positively charged in solution. The particle

constituents would not have attracted the poly-L-lysine and thus made it very difficult for

the additive to be incorporated and create composite particles. Instead, the result is the

same as (or at least very similar to) the control samples. Further analysis, however, is

key in discerning whether or not there actually are differences from the control particles.

From this analysis it seems as though they are one in the same. It should also be noted

that there is clearly no discernable trend in the effect of concentration because it seems

that the additive has no effect on the synthesis at all.

X-Ray Diffraction (XRD)

This technique was used for the same purpose as in the other two experimental

chapters: to confirm that the crystalline phase of the particles is magnetite (Fe304). Only

one of the samples was used for this analysis because it would be superfluous to do both











of them. The sample was prepared by spreading some of the particle powder onto a piece


of double-sided tape that had been adhered to a glass slide.



Particles with Poly-L-Lysine Additive


90 *

80 *
70 *
S60 *


~40 1 n
30 *





20 25 30 35 40 45 50 55 60 65 70
2 Theta (degrees)
-Standard -LYSMW=27,000

Figure 5-3: X-ray diffraction spectrum of composite particles with LYS additive. The
data for the magnetite standard is also plotted for comparative purposes. The
peaks are labeled with their corresponding planes of diffraction.

The x-ray diffraction spectrum in the figure above shows that the peaks for the


particle sample do indeed match up with the peaks for the magnetite standard, thus

confirming that the particles are composed of Fe304. If We gO back and compare this


spectrum to that of the PAA and GLU composite particles from the previous chapters, we

can see that these peaks do not seem to be quite as broad. This would correlate with the


TEM observation that the particles are not really granular. The lack of a polymer


additive in the particles would not create an added lattice strain, and thus the x-rays


would not be deflected differently and the resulting peaks would be sharper, as observed.


This spectrum again seems to show an extra small peak at an angle of about 28


degrees. This can probably be attributed again to the use of a steel spatula in spreading










the particles on the tape during sample preparation. If this is not the root of the cause,

then at least it seems to be a constant that all of the XRD spectra are showing.

Superconducting Quantum Interference Device Analysis (SQUID)

The TEM analysis yielded an interesting result in that the particles synthesized with

the LYS additive looked very much like control particles. The SQUID analysis will give

further insight to any differences that these particles may have with the controls. This

analysis is used to measure magnetic quantities by plotting a hysteresis curve.

Magnetization is measured as a function of an applied magnetic field to create the plot.

The maximum applied field was 7 Tesla. Samples were prepared by packing the particle

powder into a gelcap. A 5% error was allotted to the data based on weight uncertainty.



M vs H, Particles with Poly-L-Lysine T = 300 K
100

80 1






E 40-

r o MW=27,000; 100 ug/mL
20 MW=27,000; 250 ug/mL-





0 2 4 6 8
B (T)


Figure 5-4: Hysteresis curves for particles with LYS additive.

The hysteresis curves for the particles with the poly-L-lysine additive show the

same characteristic shape that has been observed for all of the other samples. They show









soft magnetic behavior with a low coercivity and remnant magnetization. The table

below shows the actual values for the magnetic quantities.

Table 5-2: Magnetic quantities determined from SQUID analysis of LYS particles
Saturation Remnant
Coercivity, He
Sample Magnetization, Magnetization,
(Gauss)
Msat (emu/gram) Mr, em 0 ELSE)

MW=27,000
76.8 1.1 16
100 ug/mL
MW=27,000
71.9 1.5 20
250 ug/mL

This table also reflects the high saturation magnetization achieved by these

samples. These are the most interesting quantities because they correlate so well with the

values obtained for the control samples. To demonstrate this, the figure below shows the

control samples plotted along with the LYS samples.


T = 300 K


0 2 4 6
B (T)


Figure 5-5: Hysteresis curves for both LYS and control particles









All four of the samples shown in Figure 5-5 are within the error of the other three,

effectively showing that there is no significant difference between any of them. This

result further supports what was suspected by the TEM analysis. The particles with the

poly-L-lysine additive did not form composite particles, but rather formed just like a

control sample. Again, this is likely due to the nature of the poly-L-lysine in solution. Its

charge electrostatically repels it from the magnetite particles during synthesis, thus

preventing it from inhibiting the reaction or incorporating into the particles.

It should now be clear why only two samples with the LYS additive were analyzed.

Since the additive does not incorporate and form composite particles, it would have been

ultimately useless to try to explore a multitude of polymer concentrations and molecular

weights; clearly, no trends would be present since every trial would likely form control-

like particles. It should also be clear now why fluorescence microscopy was not

performed. It would have required the purchase of a different probe that would target the

amino functionality, but since the polymer is not present in the particles, there would be

nothing for the probe to conjugate to and consequently there would be no observable

particle fluorescence.















CHAPTER 6
CONCLUSIONS AND FUTURE WORK

This study of the effect of various polymer additives on the solution synthesis of

magnetite nanoparticles has demonstrated many important findings. First and foremost,

it has been shown that composite particles with organic functionality can indeed be

formed when an appropriate polymer additive is used at a reasonable concentration and

molecular weight. It was found that the two additives with the acidic carboxyl groups

worked well to achieve the composite morphology. During the aging process of the

synthesis, these polymer chains are attracted to the iron ions that are precipitating the

crystalline magnetite particles from the amorphous intermediate. When the primary

particles aggregate and undergo a surface recrystallization phenomenon, the polymer

additive becomes permanently incorporated into the final particle.

The composite particles were characterized to confirm many things. First,

transmission electron microscopy has shown that the composite particles have a grainy

morphology, indicating the presence of a polymer additive. It was further shown that this

additive is not only present, but also functional. This was demonstrated by the

conjugation of a fluorescent probe to the functional carboxyl groups of the incorporated

polymer. In addition, it can be concluded that the composite particles are indeed the

same material as the control, Fe304, thereby showing that the additive did not have an

effect on what iron oxide phase was formed. Furthermore, it seems that some general

trends were established based on the usage of varying molecular weights and polymer

concentrations. Care must be taken in making the following statements because they may









not hold universally true. In fact, even this study found some disagreements with these

general trends. What seemed to be observed was that as the additive concentration was

increased, so did the degree of granular morphology, and consequently there was a loss in

saturation magnetization. This all makes intuitive sense too because the greater the

concentration, the more polymer there is available for incorporation, thus creating grainer

particles with more organic content, thereby lowering the magnetization per unit weight.

A similar line of reasoning would hold true for increases in molecular weight, however

the results make it much harder to speculate this trend. This study has also shown

evidence that there are upper and lower limits to both the concentration and molecular

weight of the additive. Too high a concentration or molecular weight can completely

inhibit the synthesis and prevent particles from forming at all. Too low a concentration

or molecular weight and the maj ority of particles will simply form as a control with no

incorporation of additive. Future work on an increased number of different molecular

weights and concentrations would help solidify the observation of the general trends and

hopefully delineate some upper and lower limits for the different polymer additives.

It has also been demonstrated that the composite particles will not form with just

any polymer additive. The trials with poly-L-lysine have shown this. Although the

molecular weight and concentration was on par with the other additives, none of the

polymer was incorporated into the particles. The characterization methods confirm that

this is true. The TEM images look like those of control particles, and the SQUID

measurements show no significant difference from the control sample data. This all

points to the fact that no polymer was incorporated, likely due to the positive charge

obtained by the poly-L-lysine chains in solution. This works to repel the additive from









the nuclei during synthesis, therefore a composite particle is not created. A future

experiment could be performed where the pH is adjusted during synthesis so that the

charge can be flipped and hopefully create an attractive force that will allow the additive

to be incorporated.

Universally, a few things can be concluded. The first is that all trials produce

magnetite as the crystalline phase. This has been confirmed by x-ray diffraction. This

research has also shown that the shape of the hysteresis curves remains the same for each

sample. The magnetic properties are characterized by a low coercivity and a high

saturation magnetization. Although the composite particles demonstrate a lower overall

Msat as compared to the controls, they are still stronger than many other magnetic particle

systems. Finally, this study has shown that the particles created are indeed small enough

for biomedical applications. Although the size distribution is not as monodisperse as

desired, even the largest particles get up to only about 500 nanometers;, this is still plenty

small considering that the inner diameter of capillaries are typically about 5-10 microns.

However, it is still important to work at creating a more monodisperse colloid of

composite particles. In addition, the size distribution needs to be better characterized.

This would be best achieved by the use of a light scattering technique.

So where do we go from here? It would be very useful to quantify the amount of

polymer actually present in the composite particles. This is especially important if the

particle system should ever need to be FDA approved. Thermogravimetric analysis

(TGA) should be a convenient method to obtain this data, however a complete and

thorough separation process must be established. If the remaining unconverted gel is not










completely removed, it would be difficult to distinguish in a TGA plot which weight loss

can be attributed to which material.

The presence of magnetic remanence may also cause a problem that needs more

work to correct. The particles maintain a small magnetization even when there is no

applied field. This will cause the particles to be attracted to each other and therefore

aggregate. Use of various deflocculates need to be explored to help prevent or eliminate

this problem. Too much aggregation could have disastrous effects if the particles are

used in the bloodstream because they could potentially get to the point where they block

a blood vessel. This is clearly a danger that needs to be remedied by further work.

The ultimate goal of this work is to eventually incorporate specifically sequenced

polypeptides that can target specific cell receptors or other molecules. If a composite

morphology can be successfully obtained using such an additive, then the particles would

then of course need to be fully characterized. Of the greatest interest would be to

demonstrate their specific targeting capability. This may be difficult because amino acid

sequences often have both spatial and chemical functionality. Therefore, the sequence

would have to be sufficiently accessible on the surface of the particles in order for its

targeting capabilities to work. If this did not work out, then it is very conceivable that the

same sequence could just be conjugated to the functional particles that have been created

in this study. In this way, the specifically sequenced peptide would not feel the effects of

being incorporated in the crystalline matrix.

All in all, this research has shown some very interesting results. A new particle

system has been found in which a highly magnetic particle possesses organic

functionality. This functionality affords the particles many possible applications,









including biomedical and environmental targeting and retrieval capabilities. There is a

myriad of directions that could be taken with this system, but this research serves to lay

the foundation for these magnetic nanocomposite particles.

















LIST OF REFERENCES


1. Frankel, R.B. and R.P. Blakemore, h on Biontinerals. 1991, New York: Plenum
Press.

2. Kirschvink, J.L., Mapletite Biontineralization and Geontapletic Sensitivity in
Higher Aninzals-an Update and Reconanendations for Future Study.
Bioelectromagnetics, 1989. 10(3): p. 239-259.

3. Hautot, D., Q. Pankhurst, N. Khan, and J. Dobson, Preliminary Evaluation of
Nanoscale~NNN~~~NNN~~~NNN Biogenic Mapletite in Alzheinter's Disease Brain Tissue. Proceedings of
the Royal Society of London Series B-Biological Sciences, 2003. 270: p. S62-S64.

4. Frankel, R.B. and D.A. Bazylinski, Mapletotaxis and Mapletic ParticlesP~~PPP~~~PP~~~PP in
Bacteria. Hyperfine Interactions, 1994. 90(1-4): p. 135-142.

5. Lowenstam, H.A. and S. Weiner, On Biontineralization. 1989, New York: Oxford
University Press.

6. Mann, S. and J.P. Hannington, Formation oflron Oxides in Unilante~ll~~lllar1~~~111~ Vesicles.
J. Colloid & Interface Science, 1988. 122:2: p. 326-335.

7. Gorby, Y.A., T.J. Beveridge, and R.P. Blakemore, Characterization of the
Bacterial2\aplgetosonse Membrane. Journal of Bacteriology, 1988. 170(2): p. 834-
841.

8. Bazylinski, D.A., Structure and Function of the Bacterial Maplgetosonse. Asm
News, 1995. 61(7): p. 337-343.

9. Bazylinski, D.A., Controlled Biontineralization of2apletic M~inerals by
Mapletotactic Bacteria. Chemical Geology, 1996. 132(1-4): p. 191-198.

10. Bazylinski, D.A., B.R. Heywood, S. Mann, and R.B. Frankel, Fe304 andFe3S4 in
a Bacterium. Nature, 1993. 366(6452): p. 218-218.

11. Devouard, B., M. Posfai, X. Hua, D.A. Bazylinski, R.B. Frankel, and P.R. Buseck,
Mapletite From Mapletotactic Bacteria: Size Distributions and Dr inning.
American Mineralogist, 1998. 83(11-12): p. 1387-1398.

12. Frankel, R.B., D.A. Bazylinski, and D. Schuler, Biontineralization of2apletic Iron
Minerals in Bacteria. Supramolecular Science, 1998. 5(3-4): p. 383-390.










13. Frankel, R.B., J.P. Zhang, and D.A. Bazylinski, Single Mapletic Domains in
Mapletotactic Bacteria. Journal of Geophysical Research-Solid Earth, 1998.
103(Bl2): p. 30601-30604.

14. Majhi, P., B. Devouard, M. Posfai, X. Hua, D.A. Bazylinski, R.B. Frankel, and
P.R. Buseck, Structural Features of2apletite From Mapletotactic Bacteria.
Meteoritics & Planetary Science, 1997. 32(4): p. A83-A84.

15. Meldrum, F.C., S. Mann, B.R. Heywood, R.B. Frankel, and D.A. Bazylinski,
Electron M~icroscopy Study of2apletosontes in a Cultured Coccoid Maplgetotactic
Bacterium. Proceedings of the Royal Society of London Series B-Biological
Sciences, 1993. 251(1332): p. 231-236.

16. Proksch, R.B., T.E. Schaffer, B.M. Moskowitz, E.D. Dahlberg, D.A. Bazylinski,
and R.B. Frankel, Mapletic Force M~icroscopy of the Subnzicron Mapletic
Assembly in a 2apletotactic Bacterium. Applied Physics Letters, 1995. 66(19): p.
2582-2584.

17. Germann, W.J. and C.L. Stanfield, Principles ofHuntan Physiology. 2002: Pearson
Education, Inc.

18. Spaldin, N., Mapletic Materials: Fundamentalsd~~~dd~~~ddd~~ and Device Applications. 2003,
Cambridge: Cambridge University Press.

19. Hummel, R.E., Electronic Properties of2aterials. Third ed. 2001, New York:
Springer-Verlag.

20. Kronmuller, H. and M. Fahnle, M~icrontapletisn; and the M~icrostructure of
Ferrontapletic Solids. 2003, Cambridge: Cambridge University Press.

21. Mohn, P., Mapletisn; in the Solid State: An Introduction. 2003, New York:
Springer-Verlag.

22. Richerson, D.W., M~odern Ceramic Engineering. Second ed. 1992, New York:
Marcel Dekker, Inc.

23. Matij evic, E., Usefd~ Applications of2~onodispersed Colloids Have Come of Age .
Abstracts of Papers of the American Chemical Society, 1997. 214: p. 9-24.

24. Afremov, L.L. and A.V. Panov, Mapletic States and Hysteresis Properties of Snall
Mapletite Particles.~~PPP~~~PP~~~PP Fizika Metallov I Metallovedenie, 1998. 86(3): p. 65-73.

25. Muxworthy, A.R. and W. Williams, M~icrontapletic Calculation ofHysteresis as a
Function of Temperature in Pseudo-Single Domain Mapletite. Geophysical
Research Letters, 1999. 26(8): p. 1065-1068.

26. Ozdemir, O., Coercive Force of Single Crystals of2apletite at Low Temperatures.
Geophysical Journal International, 2000. 141(2): p. 351-356.










27. Roger, J., J.N. Pons, R. Massart, A. Halbreich, and J.C. Bacri, Some Biomedical
Applications ofFerrofluids. European Physical Journal-Applied Physics, 1999.
5(3): p. 321-325.

28. Montagne, F., O. Mondain-Monval, C. Pichot, H. Mozzanega, and A. Elaissari,
Preparation and Characterization ofNarrow Sized (o w) Mapletic Emulsion.
Journal of Magnetism and Magnetic Materials, 2002. 250(1-3): p. 302-3 12.

29. Fu, L., V.P. Dravid, and D.L. Johnson, Self-assenabled (SA) Bilayer M~olecular
Coating on 2apletic Nanoparticles. Applied Surface Science, 2001. 181(1-2): p.
173-178.

30. Ratner, B.D., A.S. Hoffman, F.J. Schoen, and J.E. Lemons, Biontaterials Science:
An hItroduction to Materials in M~edicine. 1996, San Diego: Academic Press.

31. Jordan, A., R. Scholz, P. Wust, H. Fahling, and R. Felix, Mapletic Fluid
Hyperthernzia (M~FH): Cancer 7i Ieunwin~ll n ithr AC Mapletic Fieldlnduced
Excitation ofBioconspatible Superparanzapletic Nanoparticles. Journal of
Magnetism and Magnetic Materials, 1999. 201: p. 413-419.

32. Hsieh, D.S., R. Langer, and J. Folkman, Mapletic M~odulation ofRelease of
Macronsolecules From Polymers. Proceedings of the National Academy of Science,
1989. 78: p. 1863-1867.

33. Mornet, S., F. Grasset, J. Portier, and E. Duguet, Ma'~ghentite@Silica Nanoparticles
for Biological Applications. European Cells and Materials, 2002. 3(2): p. 110-113.

34. Gruttner, C. and J. Teller, New Types of Silica-Fortified Maple tic Nanoparticles as
Tools for M~olecular Biology Applications. Journal of Magnetism and Magnetic
Materials, 1999. 194(1-3): p. 8-15.

35. Imshennik, V.K., I.P. Suzdalev, O.N. Stavinskaya, N.I. Shklovskaya, V.
Schunemann, A.X. Trautwein, and H. Winkler, Preparation and Characterization
of Porous Carbon Loaded al ithr Iron Particles:~~PPP~~~PP~~~PP A Possible Mapletic Carrier of
M~edicalDrugs. Microporous Materials, 1997. 10(4-6): p. 225-230.

36. Rettenmaier, M., J. Stratton, P. Disaia, M. Berman, A. Senyei, and K. Widder,
Treatment of a Syngeneic Rat-Tunzor ~I ithr Mapletically Responsive Albumin
M~icrospheres Labeled al ithr Adrianzycin or Protein-A. Gynecologic Oncology,
1985. 20(2): p. 268-268.

37. Rettenmaier, M.A., J. Stratton, M. Berman, A. Senyei, and K. Widder, D. White,
and P. Disaia, Treatment of a Syngeneic Rat-Tunzor ~I ithr Mapletically Responsive
Albumin M~icrospheres Labeled al ithr Doxorubicin or Protein-A. Gynecologic
Oncology, 1987. 27(1): p. 34-43.

38. Senyei, A., K. Widder, and G. Czerlinski, Mapletic Guidancedd~~~~~ddddd~~~~ ofDrug-Carrying
M~icrospheres. Journal of Applied Physics, 1978. 49(6): p. 3578-3583.










39. Widder, K., G. Flouret, and A. Senyei, Magnetic 2icrospheres? ynthet~i\ ofa
NovelParenteralIDrug Carrier. Journal of Pharmaceutical Sciences, 1979. 68(1):
p. 79-82.

40. Ar shady, R., M~icrospheres for Biomedical Applications-Preparation of Reactive
and Labeled2\~icrospheres. Biomaterials, 1 993 14(1): p. 5- 15 .

41. Driscoll, C.F., R.M. Morris, A. Senyei, K. Widder, and G.S. Heller, Magnetic
Targeting of2~icroshperes in Blood-Flow. Microvascular Research, 1984. 27(3): p.
353-369.

42. Fruitwala, M.A. and S. N.M., Site-Specific Drug TaIS''ill goin ithr Fluorouracil
M~icrospheres. Drug Delivery, 1996. 3: p. 5-8.

43. Gupta, P.K. and C.T. Hung, Albumin M~icrospheres. y. Evaheation ofParantetersP~~~PP~~PP~~PP
Controlling the Efficacy of Magnetic M~icrospheres in the Taixgeted Delively of
Adrianzycin in Rats. International Journal of Pharmaceutics, 1990. 59: p. 57-67.

44. Hassan, E.E., R.C. Parish, and J.M. Gallo, Optimized Formula~tion of2agnetic
Chitosan M~icrospheres Containing the Anticancer Agent, Oxantrazole.
Pharmaceutical Research, 1992. 9(3): p. 390-397.

45. Mosbach, K. and U. Schroder, Preparation and Application of2agnetic Polymers
For TargtingTTTT~~~~~~~TTTTTT ofDrugs. Febs Letters, 1979. 102(1): p. 112-116.

46. Pulfer, S.K. and J.M. Gallo, Enhanced Brain Tumor Selectivity ofCationic
Magnetic Polysaccharide M~icrospheres. Journal of Drug Targeting, 1998. 6(3): p.
215-227.

47. Widder, K. and A. Senyei, Magnetic M~icrospheres: A Vehicle for Selective Drug
Targeting. Pharmaceutical Therapeutics, 1983. 20: p. 377-395.

48. Weishaupt, D., D.Kochli, V, Marincek, B, How doesM2RI work? An hItroduction to
the Physics and Function of2agnetic Resonance Imaging. 2003, New York:
Springer-Verlag.

49. Jung, C.W. and P. Jacobs, Physical and Chemical Properties ofSuperparantagnetic
Iron-Oxide M~R Contrast Agents--Feruntoxides, Feruntoxtr anrt~t~rt~r~rt~r Feruntoxsil.
Magnetic Resonance Imaging, 1995. 13(5): p. 661-674.

50. Bachgansmo, T., Ferrintagnetic Susceptibility Contrast Agents. Acta Radiologica,
1993. 34: p. 1-13.

51. Tiefenauer, L.X., G. Kuhne, and R.Y. Andres, Antibody Magnetite Nanoparticles:
In-Vitro Characterization ofa Potential Tunzor-Specific Contrast Agent for
Magnetic Resonance Imaging. Bioconjugate Chemistry, 1993. 4(5): p. 347-3 52.










52. Suwa, T., S. Ozawa, M. Ueda, N. Ando, and M. Kitajima, Magnetic Resonance
Imaging ofEsophageal Squamous Cell Carcinoma Using Magnetite ParticlesP~~PPP~~~PP~~~PP
Coated al ithr Anti-Epidermal Gi In thr Factor Receptor Antibody. International
Journal of Cancer, 1998. 75(4): p. 626-634.

53. Sugimoto, T. and E. Matij evic, Formation of Uniform Spherical Magnetite
Particles by Crystallization from Ferrous Hydroxide Gels. Journal of Colloid and
Interface Science, 1980. 74(1): p. 227-243.

54. Matij evic, E., M\~onodispersed M~etal (Hydrous) Oxides~ Fascinating Field of
Colloid Science. Accounts of Chemical Research, 1981. 14(1): p. 22-29.

55. Matij evic, E., Interactions 11 ithr Monodispersed Colloids. American Ceramic
Society Bulletin, 1984. 63(8): p. 994-994.

56. Matij evic, E., Production of2~onodispersed Colloidal Particles. Annual Review of
Materials Science, 1985. 15: p. 483-516.

57. Matij evic, E., M~onodispersed Colloids--Art and Science. Langmuir, 1986. 2(1): p.
12-20.

58. Matij evic, E., M~onodispersed Colloids-Achievements and Problems. Abstracts of
Papers of the American Chemical Society, 1990. 200: p. 51-59.

59. Atkinson, R.J., A.M. Posner, and J.P. Quirk, Crystal Nucleation in Fe(3) Solutions
andHydrxide Gels. Journal of Inorganic & Nuclear Chemistry, 1968. 30(9): p.
2371-&.

60. Atkinson, R.J., A.M. Posner, and J.P. Quirk, Crystal Nucleation and Giron thrl in
Hydrolyzing Iron(III) Chloride Solutions. Clays and Clay Minerals, 1977. 25(1): p.
49-&.

61. Her, Y. S., S.H. Lee, and E. Matij evic, Continuous Precipitation of2~onodispersed
Colloidal Particles:~~PPP~~~PP~~~PP SiO2, Al(OH) (3), and BaTiO3. Journal of Material s Research,
1996. 11(1): p. 156-161.

62. Park, J., V. Privman, and E. Matij evic, M~odel ofFormation of2~onodispersed
Colloids. Journal of Physical Chemistry B, 2001. 105(47): p. 1 1630-1 1635.

63. Privman, V., D.V. Goia, J. Park, and E. Matij evic, M~echanism ofFormation of
M~onodispersed Colloids by Aggregation ofNanosize Precursors. Journal of
Colloid and Interface Science, 1999. 213(1): p. 36-45.

64. Sada, E., H. Kumazawa, and H.M. Cho, Formation of2agnetite Fine ParticlesP~~PPP~~~PP~~~PP by
Chemical A bsorption into Aqueous Suspensions of Ferrous Hydroxide Chemical
Engineering Communications, 1989. 75: p. 89-99.










65. Sun, S.H. and H. Zeng, Size-Controlled Sy inhes~ i\ of2agnetite Nanoparticles.
Journal of the American Chemical Society, 2002. 124(28): p. 8204-8205.

66. Taylor, R.M., Influence of Chloride on the Formation oflron-Oxides from Fe(II)
Chloride: Effect of[Cl]/Fej on the Formation of2agnetite. Clays and Clay
Minerals, 1984. 32(3): p. 167-174.

67. Taylor, R.M., Influence of Chloride on the Formation oflron-Oxides from Fe(II)
Chloride: Effect of [Cl on the Formation ofLepidocrocite and Its Crystallinity.
Clays and Clay Minerals, 1984. 32(3): p. 175-180.

68. Lee, S.H., Y. S. Her, and E. Matij evic, Preparation and Gia th~ Ml mechanism of
Uniform Colloidal Copper Oxide by the Controlled Double-Jet Precipitation.
Journal of Colloid and Interface Science, 1997. 186(1): p. 193-202.

69. Johnson, T.J., E.A. Waddell, G.W. Kramer, and L.E. Locascio, Chemical2apping
ofHot-Embossed and UV-Laser Ablated M~icrochannels in Poly(M~ethyl
M~ethacrylate) Using Carboxylate Specific Fluorescent Probes. Applied Surface
Science, 2001. 181(1-2): p. 149-159.

70. Hermanson, G., Bioconjugate Techniques. 1996, San Diego: Academic Press.

71. Kapanidis, A.N. and S. Weiss, Fluorescent Probes and Bioconjugation Chemistries
for Single M~olecule Fluorescence Analysis ofBiomolecules. Journal of Chemical
Physics, 2002. 117(24): p. 10953-10964.
















BIOGRAPHICAL SKETCH

Barry William Miller was born on July 12, 1981, in Akron, Ohio, to William and

Carol Miller. He is the only son and youngest of three children. At age two, his family

moved to Florida, and after a short stint in a rental home where he climbed his nostalgic

first tree and learned to ride his first banana-seat bicycle, he moved to New Port Richey,

Florida, where he grew up and attended school. During this period of his life he had to

share a bathroom with his two elder sisters, Leigha and Lindsay, and he consequently

learned at an early age the importance of putting the toilet seat down after each use--a

quality that will assuredly make him a great man, at least in the eyes of most women.

Barry was very active as a young man and participated in many sports. After

multiple head and body collisions with errant baseballs in Little League, he decided it

was time for a change in sport. He made a few pursuits in basketball; however he later

discovered his prowess in the swimming pool at age 14. Although the leg-shaving

escapades were met with widespread peer ridicule, the form-fitting design of a Speedo

never created the need for embarrassment. Everything else aside, he was very successful

with swimming, garnering many district and conference championships as well as many

commendable performances at state.

After graduating from River Ridge High School near the top of his class in 1999,

he matriculated at the University of Florida later that fall. A few years earlier, after being

inspired by a discovery of how to bake Thirty-Minute Brownies in only twenty minutes,

he decided that he would maj or in the field of engineering. He chose to pursue materials






82


engineering with specialties in polymers and biomaterials. Never satisfied with the pace

of his learning, he entered an accelerated program that allowed him to earn his master' s

degree in coincidence with his bachelor' s. These degrees will be awarded, with honors,

in May 2004. After graduation he plans to pursue an engineering career in industry.