Synthesis and bio-functionalization of nanoparticles for biosensing and biorecognition

Transcription

1 DOTTORATO DI RICERCA IN NANOSCIENZE XXIII CICLO Sede Amministrativa Università degli Studi di MODENA e REGGIO EMILIA TESI PER IL CONSEGUIMENTO DEL TITOLO DI DOTTORE DI RICERCA Synthesis and bio-functionalization of nanoparticles for biosensing and biorecognition Candidato: Paola Serena D Agostino Relatore: Dr. Paolo Facci

2

3 INDEX PREFACE INTRODUCTION TO NANOMATERIALS Defining the nanoscale Aims Natural magnetic nanoparticles Methods of fabricating nanostructures and self- assembly Magnetic oxides Biomedical applications of magnetic nanoparticles (MNPs) Characterization of nanoparticles Bioconjugation The chemistry of reactive groups Semiconductor Quantum Dots (QDs) QDs synthesis Water- soluble colloidal nanocrystals Conjugation to QDs References METHODS OF SYNTHESIS Synthesis and stabilization of magnetic iron oxide nanoparticles Surface modification Aminopropyltriethoxysilane (APTES) coating Surface functionalization via ligand exchange Silane exchange Dopamine exchange Magnetic fluid properties Bioconjugation by glutaraldehyde Objectives References... 42

4 3 AQUEOUS SOLUTION SYNTHESIS OF MAGNETIC NANOPARTICLES Materials and methods Materials Synthesis of magnetite- APTES nanoparticles in aqueous solution Characterization Results and discussion Characterization of MNPs obtained by aqueous solution synthesis Magnetic properties FTIR characterization XPS characterization TEM characterization Conclusion References ORGANIC PHASE SYNTHESIS OF MAGNETIC NANOPARTICLES Materials Preparation of MNPs by thermo- decomposition method Synthesis of 6 nm Fe3O4 nanoparticle seeds Reaction of ligand exchange Silane ligand exchange Improving dispersion of MNPs- APTES Synthesis of magnetite- dopamine nanoparticles Characterization TEM, SEM and AFM characterization FTIR Elemental analysis XPS characterization Conclusion References BIOFUNCTIONALIZATION OF NANOPARTICLES (MNPs AND QDs)... 66

5 5.1 Bioconjugation of insulin on MNPs Materials and methods Materials Synthesis of bigger magnetic nanoparticles (seed- mediated growth) Synthesis of 8 nm Fe3O4 Nanoparticles via 6 nm Fe3O4 Seeds Silane ligand exchange Coupling of MNPs- APTES with glutaraldehyde Bioconjugation of MNPs- APTES with insulin Characterization AFM, TEM and SEM characterization Magnetic characterization of bigger APTES- MNPs Quartz crystal microbalance (QCM) characterization Experimental procedure Binding kinetics of MNPs and insulin Bioconjugation of modified peptidic linker on QDs Materials Quantum Dots Synthesis of succinimidyl ester iodoacetamide Synthesis of (His) 6 - NH OSu- iodoacetate coupling to Rink- (His) 6 - NH Peptide cleavage Conjugation of compound (5) to thiolated- DNA Self- assembly of (His) 6 - peptide modified DNA with QDs Fluorescence resonance energy transfer (FRET) analysis Discussion Gel electrophoresis characterization AFM characterization FRET analysis QDs- MB construct to discriminate between different sequences of DNA Conclusion References... 94

6 6 MAGNETIC FORCE MICROSCOPY (MFM) CHARACTERIZATION Atomic force microscopy (AFM) Fundamental elements of the atomic force microscope How does the AFM work? Cantilevers and tips Imaging modes Magnetic force microscopy (MFM) Materials and methods AFM and MFM characterization Preliminary results in MFM Tests of magnetized tip LIFT mode in liquid on MNPs- APTES LIFT mode with reversed field of tip Larger MNPs embedded in PMMA (poly(methyl methacrylate)) in order to ensure the independence from topography Comparison of magnetic and not magnetic area to access the true magnetic character of the imaged features Conclusion References CONCLUSION

7 PREFACE The aim of this PhD thesis is to integrate nanoparticles, magnetic nanoparticles (MNPs) and quantum- dots (QDs), with biological molecule for achieving constructs to be used in nanobiosensors and in receptor mapping. The main part of my PhD dissertation regards the synthesis and functionalization of NPs with peptidic ligands, such as insulin and thiol- reactive iodoacetyl hexahistidine peptidyl linker modified with DNA. A very intensive investigation has been conducted to look for new strategies to prepare magnetic nanoparticles with tailor- made properties through appropriately attached functional moieties. In most of their potential applications, the quality (it is important to accurately control the size, shape, and (bio)chemical coating and to retain the thermal and chemical stability) and structure of the nanoparticles surface will play a crucial role in determining their function. Therefore, a key challenge for future bioapplications of MNPs is the development of surface chemistry that provides a versatile, stable, and well defined interface while facilitating the incorporation of chemical functional groups. I synthesized MNPs in organic- phase producing highly uniform and monodisperse particles. The nanoparticles resulting from these synthetic procedures are stable in nonpolar solvents (such as hexane) and capped with nonpolar endgroups (oleic acid). Oleic acid is chemisorbed on the particle surface as a carboxylate with its nonpolar CH 3 endgroup exposed to the solution. To render these MNPs dispersible in water, the hydrophobic ligands were exchanged for hydrophilic ones. Ligand exchange is a well- known method for improving the surface properties of nanoparticles. It involves adding an excess of ligand to the nanoparticle solution, which results in the displacement of the original ligand on the nanoparticles surface. The binding of oleates is noncovalent, which means that they can be more easily desorbed from the surface. As ligand molecules, 3- aminopropyltriethoxysilane (APTES) and dopamine were used. The bioconjugation was performed between MNPs- APTES and insulin. The coupling of MNPs and insulin was made by an omobifunctional linker, glutaraldehyde, which present two carbonyl groups at both ends. The reaction is based on imino- group formation (C=N). The bioconjugate finally will be used in nanobiosensors (gravimetric sensors) and in magnetic force microscopy (MFM), in liquid mode, towards surface cell receptor mapping. 1

8 Other bioconjugation was made between CdSe/ZnS core/shell QDs and a peptidic linker conjugated to thiolated DNA for subsequent self- assembly with QDs to obtain a construct for use as multifunctional probe. The linker was conjugated to thiolated DNA by thioetheric bond and the versatility of this approach demonstrated by self- assembling a fluorescence resonance energy transfer (FRET)- based QD- DNA molecular beacon that was able to sense the presence of its cognate DNA complement. A part of the obtained results has been the object of a publication and another publication, including the results on Magnetic Force Microscopy, is at the moment in preparation. L. Berti, P.S. D Agostino, K. Boeneman, I. Medintz; Improved Peptidyl Linkers for Self- Assembly of Semiconductor Quantum Dot Bioconjugates, Nano Res., 2 (2009)

9 Chapter 1. Introduction to nanomaterials Nanotechnology involves the study, control, and manipulation of materials at the nanoscale, typically having dimensions less than 100 nm. This is a truly multidisciplinary area of research and development, bringing together the disciplines of chemistry, biology, engineering, and medicine. The recent interest in nanostructures results from their numerous potential applications such as in materials development, biomedical sciences, electronics, optics, magnetism, energy storage, and electrochemistry. This thesis concentrates on synthesis of magnetic iron oxide nanoparticles (MNPs) and on biofunctionalization of MNPs and quantum dots (QDs). In the following section, MNPs and QDs will be discussed. 1.1 Defining the nanoscale First of all, it is necessary to consider the general concepts related to the nanosized objects. A nanoobject is a physical object differing appreciably in properties from the corresponding bulk material and having at least one dimension in the nanometer range(no more than 100 nm). When dealing with nanoparticles, magnetic properties (and other physical ones) are size dependent to a large extent. Therefore, particles whose sizes are comparable with (or lesser than) the sizes of magnetic domains in the corresponding bulk materials are the most interesting from a magnetism scientist viewpoint. Nanotechnology is the technology dealing with both single nanoobjects and materials, and devices based on them, and with processes that take place in the nanometer range. Nanomaterials are those materials whose key physical characteristics are dictated by the nanoobjects they contain. Nanomaterials are classified into compact materials and nanodispersions. The first type includes so- called nanostructured materials [1], i.e., materials isotropic in the macroscopic composition and consisting of contacting nanometer- sized units as repeating structural elements [2]. Unlike nanostructured materials, nanodispersions include a homogeneous dispersion medium (vacuum, gas, liquid, or solid) and nanosized 3

10 inclusions dispersed in this medium and isolated from each other. The distance between the nanoobjects in these dispersions can vary over broad limits from tens of nanometers to fractions of a nanometer. In the latter case, we are dealing with nanopowders whose grains are separated by thin (often monoatomic) layers of light atoms, which prevent them from agglomeration. Materials containing magnetic nanoparticles, dispersed in nonmagnetic matrices at the distances longer than their diameters, are most interesting for magnetic investigations. A nanoparticle is a quasi- zero- dimensional (0D) nanoobject in which all characteristic linear dimensions are of the same order of magnitude (not more than 100 nm). Nanoparticles can basically differ in their properties from larger particles, for example, from long- and well- known ultradispersed powders with a grain size above 0.5 μm. As a rule, nanoparticles are shaped like spheroids. Nanoparticles with a clearly ordered arrangement of atoms (or ions) are called nanocrystallites. Nanoparticles with a clear- cut discrete electronic energy levels are often referred to as quantum dots or artificial atoms ; most often, they have compositions of typical semiconductor materials, but not always. Many magnetic nanoparticles have the same set of electronic levels. Figure 1.1 The classification of metal containing nanoparticles by the shape. Nanoparticles are of great scientific interest because they represent a bridge between bulk materials and molecules and structures at an atomic level. The term cluster, which has been widely used in the chemical literature in previous years, is currently used to designate small nanoparticles with sizes less than 1 nm. Magnetic polynuclear coordination compounds (magnetic molecular clusters) belong to the special type of magnetic materials 4

11 often with unique magnetic characteristics. Unlike nanoparticles, which always have a certain size variability, molecular magnetic clusters are fully identical small magnetic nanoparticles. Their magnetism is usually described in terms of exchange- modified paramagnetism. Nanorods and nanowires, as shown in figure 1.1, are quasi- one- dimensional (ID) nanoobjects. In these systems, one dimension exceeds by an order of magnitude the other two dimensions, which are in the nanorange. The group of two- dimensional objects (2D) includes planar structures nanodisks, thin- film magnetic structures, magnetic nanoparticle layers, etc., in which two dimensions are an order of magnitude greater than the third one, which is in the nanometer range. The nanoparticles are considered by many authors as giant pseudomolecules having a core and a shell and often also external functional groups. The superparamagnetic properties are usually inherent to the particles with a core size of 2 30 nm. For magnetic nanoparticles, this value coincides approximately with the size of a magnetic domain in most bulk magnetic materials. 1.2 Aims Among many known nanomaterials, a special position belong to those, in which isolated magnetic nanoparticles (magnetic molecular clusters) are dispersed in a dielectric nonmagnetic medium. These nanoparticles can be regarded as giant magnetic pseudoatoms with a huge overall magnetic moment and collective spin. In this regard nanoparticles fundamentally differ from the classic magnetic materials characterized by their domain structure. As a result of recent investigations, a new physics of magnetic phenomena nanomagnetism was developed. Nanomagnetism advances include superparamagnetism, ultrahigh magnetic anisotropy and coercive force, and giant magnetic resistance. A fundamental achievement in the last time is the development of the solution preparation of the objects with advanced magnetic parameters. Currently, unique physical properties of nanoparticles are subject of intensive research [4, 5]. A special place is occupied by magnetic properties in which the difference between a massive (bulk) material and a nanomaterial is especially pronounced. In particular, it was shown that magnetization and the magnetic anisotropy of nanoparticles could be much greater than those of a bulk specimen, while differences in the Curie or Ne él temperatures 5

12 between nanoparticle and the corresponding microscopic phases reach hundreds of degrees. The magnetic properties of nanoparticles are determined by many factors, the key of these including chemical composition, type and degree of defectiveness of the crystal lattice, particle size and shape, morphology (for structurally inhomogeneous particles), interaction of particles with the surrounding matrix and neighboring particles. By changing nanoparticle size, shape, composition, and structure, one can control the magnetic characteristics of the material based on them. However, these factors cannot always be controlled during the synthesis of nanoparticles nearly equal in size and chemical composition; therefore, the properties of nanomaterials of the same type can be markedly different. In addition, magnetic nanomaterials were found to possess a number of unusual properties giant magnetoresistance, abnormally high magnetocaloric effect, and so on. Nanomagnetism usually considers so- called single- domain particles; typical values for the single- domain size range from 15 to 150 nm. Recently the researchers focused their attention on the particles, whose sizes are smaller than the domain size range; a single particle of size comparable to the minimum domain size would not break up into domains; there is a reason to call these particles domain free magnetic nanoparticles (DFMN). Each such particle behaves like a giant paramagnetic atom and shows superparamagnetic behavior when the temperature is above the so- called blocking temperature. Experiments show that the last one can vary in a wide range, from few kelvin to temperatures higher than room temperature. It is important to mention that the intensity of interparticle interactions can dramatically affect the magnetic behavior of their macroscopic ensemble. Now it is possible to prepare individual nanometer metal or oxide particles not only as ferromagnetic fluids (whose preparation was developed back in the 1960s) [5, 6] but also as single particles covered by ligands or as particles included into rigid matrices (polymers, zeolites, etc.) Natural magnetic nanoparticles Interstellar space, lunar samples, and meteorites have inclusive magnetic nanoparticles. The geomagnetic navigational aids in all migratory birds, fishes and other animals contain magnetic nanoparticles. The most common iron storage protein ferritin ([FeOOH]n containing magnetic nanoparticle) is present in almost every cell of plants and animals 6

13 including humans. The human brain contains over 108 magnetic nanoparticles of magnetite maghemite per gram of tissue [7]. Readers who are interested in more detailed information about the physical properties, magnetic behavior, chemistry, or biomedical applications of magnetic nanoparticles are referred to specific reviews [8]. 1.3 Methods of fabricating nanostructures A series of general methods for nanoparticles synthesis have been developed to date [9, 10]. Most of them can also be used for the preparation of magnetic particles. An essential feature of their synthesis is the preparation of particles of a given size and shape; at least, the dispersity should be small, 5% 10%, and controllable, since the blocking temperature (and other magnetic characteristics) depends on the particle size. The shape control and the possibility of synthesis of anisotropic magnetic structures are especially important. In order to eliminate (or substantially decrease) the interparticles interactions, magnetic nanoparticles often need to be mutually isolated by immobilization on a substrate surface or in the bulk of a stabilizing matrix or by coating with long chain ligands. It is important that the distance between the particles in the matrix be controllable. Finally, the synthetic procedure should be relatively simple, inexpensive and reproducible. The development of magnetic materials is often faced with the necessity of preparing nanoparticles of a complex composition, namely, ferrites, FePt, NdFeB or SmCo5 alloys, etc. In these cases, the range of synthetic approaches substantially narrows down. For example, the thermal evaporation of compounds with a complex elemental composition is often accompanied by a violation of the stoichiometry in the vapor phase, resulting in the formation of other substances, while the atomic beam synthesis does not yield a homogeneous distribution of elements in the substrate. The mechanochemical methods of powder dispersion also violate (in some cases, substantially) phase composition: in particular, ferrites do not retain homogeneity and oxygen stoichiometry. Furthermore, there is a difficulty of synthesis of the heteroelement precursors required composition. For example, no precursors for SmCo5 with a Sm atom bonded to five Co atoms are known; the maximum chemically attainable element ratio in Sm[Co(CO)4]3 is 1:3. It is even more difficult to propose a stoichiometric precursor for the synthesis of NdFeB nanoparticles. The 7

14 overview of general aspects of nanoalloys preparation and characterization and resulting difficulties is reported in [11]. The physical characteristics of nanoparticles are known to be substantially dependent on their size. Unfortunately, most of the currently known methods of synthesis yield nanoparticles with rather broad size distributions (dispersion>10%). The thorough control of reaction parameters (time, temperature, stirring velocity, and concentrations of reactants and stabilizing ligands) does not always allow one to narrow down this distribution to the required range. Therefore, together with the development of methods for synthesis of nanoparticles with a narrow size distribution, the techniques of separation of nanoparticles into rather monodisperse fractions can be improved. This is done using controlled precipitation of particles from surfactant- stabilized solutions followed by centrifugation. The process is repeated until nanoparticles fractions with specified sizes and dispersion are obtained. The methods of nanoparticles preparation cannot be detached from stabilization methods. For 1 10 nm particles with a high surface energy, it is difficult to select a really inert medium [12], because the surface of each nanoparticles bears the products of its chemical modification, which affect appreciably the nanomaterial properties. This is especially important for magnetic nanoparticles in which the modified surface layer may possess magnetic characteristics markedly differing from those of the particle core. Nevertheless, the general methods for nanoparticles synthesis are not related directly to the stabilization and special methods exist where nanoparticles formation is accompanied by stabilization (in matrices, by encapsulation, etc.). 1.4 Magnetic oxides Iron oxides Iron oxides have received increasing attention due to their extensive applications, such as magnetic recording media, catalysts, pigments, gas sensors, optical devices, and electromagnetic devices [13]. They exist in a rich variety of structures (polymorphs) and hydration states; therefore until recently, knowledge of the structural details, thermodynamics and reactivity of iron oxides has been lacking. Furthermore, physical (magnetic) and chemical properties commonly change with particle size and degree of hydration. By definition, superparamagnetic iron oxide particles are generally classified with 8

15 regard to their size into superparamagnetic iron oxide particles (SPIO), displaying hydrodynamic diameters larger than 30 nm, and ultrasmall superparamagnetic iron oxide particles (USPIO), with hydrodynamic diameters smaller than 30 nm. USPIO particles are now efficient contrast agents in biomedical imaging used to enhance relaxation differences between healthy and pathological tissues, due to their high saturation magnetization, high magnetic susceptibility, and low toxicity. The biodistribution and resulting contrast of these particles are highly dependent on their synthetic route, shape, and size [14]. There has been much interest in the development of synthetic methods to produce high- quality iron oxide systems. The synthesis of controlled size magnetic nanoparticles is described in multiple publications. High- quality iron oxide nanomaterials have been generated using high- temperature solution phase methods similar to those used for semiconductor quantum dots. Other synthesis methods such as polyol- mediated, sol gel [15] and sonochemical [16] were also proposed. The effectiveness of the nonaqueous routes for the production of well- calibrated iron oxide nanoparticles was shown in [17]. Magnetite nanocrystals (and other) were easily purified using standard methods also developed for quantum dots. For achieving a variety of magnetic nanomaterials properties different morphologies including spheres, rods, tubes, wires, belts, cubes, starlike, flowerlike, and other hierarchical architectures were fabricated by various approaches. Finally, some bacteria couple the reduction of Fe(III) with the metabolism of organic materials, which can include anthropogenic contaminants, or simply use iron oxides as electron sinks during respiration [18]. Figure 1.2 Schematic of a partial unit cell and magnetic ordering of spinel ferrite structure. A and B are two cationic sites in the spined structure through which coordination occurs with oxygen. Fe3O4 as spinel ferrite is FeFe2O4 (AB2O4). 9

16 Fe3O4 (magnetite) Among all iron oxides, magnetite Fe3O4 possess the most interesting properties because of the presence of iron cations in two valence states, Fe 2+ and Fe 3+, in the inverse spinel structure (figure 1.2). The cubic spinel Fe3O4 is ferrimagnetic at temperatures below 858 K. The synthetic route to these particles most often involves treatment of a solution of a mixture of iron salts (Fe 2+ and Fe 3+ ) with a base under an inert atmosphere. For example, the addition of an aqueous solution of ammonia to a solution of FeCl2 and FeCl3 (1 : 2) yields nanoparticles, which are transferred into a hexane solution by treatment with oleic acid [19]. Repeated selective precipitation gives Fe3O4 nanoparticles with a rather narrow size distribution. The synthesis can be performed starting only from FeCl2, but in this case, a specified amount of an oxidant (NaNO2) should be added to the aqueous solution apart from alkali. This method allows one to vary both the particle size ( nm) and, to a certain extent, the particle shape [20]. In some cases, thermal decomposition of compounds containing Fe 3+ ions under oxygen- deficient conditions is accompanied by partial reduction of Fe 3+ to Fe 2+. Thus thermolysis of Fe(acac)3 in diphenyl ether in the presence of small amounts of hexadecane- l,2- diol (probable reducer of a part of Fe 3+ ions to Fe 2+ ) gives very fine Fe3O4 nanoparticles (about 1 nm), which can be enlarged by adding excess Fe(acac)3 into the reaction mixture [21]. Fe3O4 nanoparticles can be also prepared in uniform sizes of about 9 nm by autoclave heating of a mixture, consisting of FeCl3, ethylene glycol, sodium acetate, and polyethylene glycol [22]. For partial reduction of Fe 3+ ions, hydrazine has also been recommended [23]. The reaction of Fe(acac)3 with hydrazine is carried out in the presence of a surfactant. This procedure resulted in superparamagnetic magnetite nanoparticles with controlled sizes, 8 and 11 nm. Also so- called dry methods are used alongside with the solution ones. Thus, Fe3O4 nanoparticles with an average size of 3.5 nm have been prepared by thermal decomposition of Fe2 (C2 O4 )3 5H2 O at T > 400 C. Furthermore, the controlled reduction of ultradispersed α- Fe2O3 in a hydrogen stream at 723 K (15 min) is a more reliable method of synthesis of Fe3O4 nanoparticles. Particles with ~ 13 nm size were prepared in this way [24]. The stabilization in the water media is interesting for bioapplications, but at the same time it represent a problem. For solving it cyclodextrin was used to transfer obtained organic ligand stabilized iron oxide nanoparticles to aqueous phase via forming an inclusion complex between surface- bound surfactants and cyclodextrin [25]. 10

17 In contrast, higher nanoparticles (20 nm < d < 100 nm) are of great interest, mainly for hyperthermia, because of their ferrimagnetic behavior at room temperature. However, there are some difficulties encountered when trying to get a monodisperse magnetite particle of size larger than 20 nm while controlling the stoichiometry. 1.5 Biomedical applications of magnetic nanoparticles Nanotechnology is an enabling technology that deals with nanometer- sized objects. It is expected that nanotechnology could impact on several applications: materials, devices, and systems. At present, nanomaterials application is the most advanced one, both in scientific knowledge and in commercial applications. A decade ago, nanoparticles were studied because of their size- dependent physical and chemical properties. Now they have entered a commercial exploration period. Magnetic nanoparticles offer some attractive possibilities in medicine. Living organisms are built of cells that are typically 10 μm in diameter. However, the cell parts are much smaller and in the submicron size domain. First advantage in medicine is that nanoparticles have controllable sizes ranging from a few nanometers up to tens of nanometers, which places them in a size range that is smaller than that of a cell ( μm), or comparable to that of a virus ( nm), a protein (5 50 nm), or a gene (2 nm wide and nm length). This means that they can get close to a biological entity of interest. This simple size comparison suggests the idea of using nanoparticles as very small probes that would allow us to spy at the cellular machinery without introducing too much interference. Indeed, they can be coated with biological molecules to make them interact with or bind to a biological entity, thereby providing a controllable means of tagging or addressing it. Second, if nanoparticles are magnetic, they can be manipulated by an external magnetic field gradient. This action at a distance, combined with the intrinsic penetrability of magnetic fields into human tissue, opens up many applications involving the transport and immobilization of magnetic nanoparticles, or of magnetically tagged biological entities. In this way, they can be made to deliver a package, such as an anticancer drug, to a targeted region of the body, such as a tumor. Third, the magnetic nanoparticles can be made to respond resonantly to a time- varying magnetic field, with advantageous results related to the transfer of energy from the exciting field to the nanoparticles. For example, the particle can be made to heat up, which leads to 11

18 their use as hyperthermia agents, delivering toxic amounts of thermal energy to targeted bodies such as tumors; or as chemotherapy and radiotherapy enhancement agents, where a moderate degree of tissue warming results in more effective malignant cell destruction. These, and many other potential applications, are made available in biomedicine as a result of the special physical properties of magnetic nanoparticles. Understanding of biological processes on the nanoscale level is a strong driving force behind development of nanotechnology. Nanoparticles have a size (mass) between single molecules and cells, i.e., a size of nm, or 500 to 10oo g/mol particle mass. The size is between that of large protein complexes (5 10 nm), e.g., ATP- synthase, and cells. The corresponding native biostructures are cellular compartments, i.e., mitochondria, chloroplasts, and the cytoskeleton elements, i.e., actin fibers and microtubuli with the associated molecular motor systems, supplying active motion and transport. There are many instruments that are able to measure nanoparticles sizes. Some system uses dynamic light scattering and can determine particle diameter due to differences in scattering from solid and liquid phases. Magnetic nanoparticles can be a promising tool for several applications in vitro and in vivo. In medicine, many applications were investigated for diagnostics and therapy and some practical approaches were choosen. Magnetic immunobeads, magnetic streptavidine DNA isolation, cell immunomagnetic separation (IMS), magnetic resonance imaging (MRI), magnetic targeted delivery of therapeutics, or magnetically induced hyperthermia are approaches of particular clinical relevance. Investigations on applicable particles induced a variability of micro- and nanostructures with different materials, sizes, and specific surface chemistry [26]. Nanoparticles in medicine are useful for therapy, imaging, and diagnostics of cancer and other diseases leading an entrapped or bound therapeutic or diagnostic target material to the area of interest, e.g., a tumor. The destination targeted delivery may be found by physical forces (magnetic) or with surface- bound antibodies (cell/tissue- specific). Motile polymers and membranes a long- term concept for technical application of molecular motion (polymers and chimerical membranes) are capable of active motion. Nanoparticles are structure components of these motile systems, which can supply the system with the energy required for motion, e.g., by magnetic forces. The nanoparticles for medical applications as well as motile polymers use the following nanoparticle structure elements as components 12

19 1. Magnetic liposomes liposomes with an internal ferromagnetic iron oxide shell, entrapped magnetic particles or lipid- bound paramagnetic ions. These magnetic target carrier particles can be used for cancer therapy (neutron capture of entrapped boron compounds), magnetic drug targeting (drug entrapped in the liposome lumen), bioanalytics (analytical target signal, imaging), and biophysical experiments (membranes, rheology, cellular traffic, and transport). Magnetic liposomes of nm size can be used for targeting in vivo, i.e., magnetic drug targeting (MDT), and magnetic radiation targeting for X- rays (photodynamic X- ray therapy (PXT)), neutrons (neutron capture therapy (NCT)), and isotopes (PET). 2. Ferrofluids contain iron oxide nanoparticles (spheres) covered with biocompatible polymers for magnetic drug targeting (cancer therapy), spectroscopy, magnetic imaging (MRI), and technical applications. Biocompatible ferrofluids are water- based and contain only endogenous or bioinert materials. Small ferrofluid particles (usually single- domain) are suitable for hypothermic cancer therapy (overheating by RF application). For biomedical target applications, the magnetic effect of simple ferrofluids is too small. Thus only polyferrofluids of nm size, featuring a large macroscopic magnetic moment and magnetic structure generation, can be used for targeting in vivo, i.e., magnetic drug targeting MDT, and magnetic radiation targeting for X- rays (PXT) and neutrons (NCT). Some current applications of nanomaterials in biology and medicine (figure 1.3) are fluorescent biological labels [27 29], drug and gene delivery [30, 31], biodetection of pathogens [32], detection of proteins [33], probing of DNA structure [34], tissue engineering [35, 36], tumor destruction via heating (hyperthermia) [37], separation and purification of biological molecules and cells [38], MRI contrast enhancement [39], and phagokinetic studies [40]. As mentioned above, nanomaterials are suitable for biotagging or labeling because they are of the same size of proteins. The other interesting feature to use nanoparticles as biological tags is their biosusceptibility. In order to interact with a biological target, a molecular linker should be attached to the nanoparticle, acting as a bioinorganic interface. Examples of biological coatings may include antibodies, biopolymers like collagen [41], or molecule monolayers (amino acids, sugars) that make the nanoparticles biocompatible [42]. In addition, as optical detection techniques are wide spread in biological research, it is better if nanoparticles show fluorescence or have other optical features. 13

20 Figure 1.3 Schematic representation of the therapeutic strategy using magnetic iron oxide nanoparticles, MRI: magnetic resonance imaging and MI: magnetoimpedance. Nanoparticles usually form the core of nanobiomaterials. It can be used as a convenient surface for molecular assembly and may be composed of inorganic or polymer materials. It can also be in the form of nanovesicle surrounded by a membrane or a layer. The shape is not automatically spherical but sometimes cylindrical or platelike. Even more complicated shapes are possible. The size and size distribution might be important in some cases, for example, if penetration through a pore structure of a cellular membrane is required. The size and size distribution are becoming extremely critical when quantum- sized effects are used to control material properties. A tight control of the average particle size and a narrow distribution of sizes allow creating very efficient fluorescent probes that emit narrow light in a very wide range of wavelengths. This helps creating biomarkers with many well- distinguished colors. The core itself might have several layers and be multifunctional. For example, by combining magnetic and luminescent layers one can both detect and manipulate the particles. The core particle is often protected by several monolayers of inert material, for example, silica. Organic molecules that are adsorbed or chemisorbed on the surface of the particle are also used for this purpose. The same layer might act as a biocompatible material. However, more often an additional layer of linker molecules is required to proceed with further functionalization. This linear linker molecule has reactive groups at both ends. One group is aimed at attaching the linker to the nanoparticle surface and the other is used to bind 14

21 various moieties like biocompatible (for example, dextran), antibodies, fluorophores etc., depending on the function required for the application. Functionalized magnetic nanoparticles have found many applications including cell separation and probing. Most of the magnetic particles studied so far are nearly spherical, which can limit the possibilities to make these nanoparticles multifunctional. Alternative cylindrically shaped nanoparticles can be created by employing metal electrodeposition into nanoporous alumina template [43]. As surface chemistry for functionalization of metal surfaces is well- developed, different ligands can be selectively attached to different parts of nanoparticle surface. It is possible to produce magnetic nanowires by spatially segregated fluorescent parts. In addition, because of the large aspect ratios, the residual magnetization of these nanowires can be high. Hence, weaker magnetic field can be used to drive them. It has been shown that a self- assembly of magnetic nanowires in suspension can be controlled by weak external magnetic fields. This would potentially allow controlling cell assembly in different shapes and forms. Moreover, an external magnetic field can be combined with a lithographically defined magnetic pattern ( magnetic trapping ). 1.6 Characterization of nanoparticles The size of the particle core of nanoparticles can be determined by transmission electron microscopy (TEM) images [44, 45]. This technique reports the total particle size of the core (crystalline and amorphous parts) and gives access to a number- weighted mean value. It provides details on the size distribution and the shape. However, this technique needs an analysis by image treatment, and it has to be performed on a statistically significant number of particles. Moreover, the sample preparation can induce aggregation of the colloids, and the TEM measurements may consequently not reflect the size and the distribution in solution; aggregates of smaller particles can be discerned [46]. High- resolution transmission electron microscopy (HRTEM) gives access to the surface atomic arrangement. It can be used to study local microstructures (such as lattice vacancies and defects) of crystalline nanoparticles [44-46]. Other physicochemical techniques used to investigate the surface properties of coated nanoparticles are; atomic force microscopy (AFM), X- ray photoelectron spectroscopy (XPS), fourier transform infrared spectroscopy (FTIR), secondary ion mass spectra (SIMS), 15

22 conductimetry, potentiometry, thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and solid- state NMR. These techniques are used to describe the nature and strength of the bonding between the nanoparticle surface and the capping and to understand the influence of the capping on the properties of the nanoparticles. TEM, XPS, FTIR, AFM are employed in this work. 1.7 Bioconjugation Biofunctionalization can be defined as a surface modification technique which attaches biological entities to synthetic materials. To biofunctionalize magnetic nanoparticles, a number of biological entities can be used, such as proteins, ligands, antibodies, enzymes, etc. Biofunctionalized magnetic nanoparticles consisting of a specific ligand coating on the outer surface can be used to isolate cells, cell organelles, nucleic acids, proteins, etc [47, 48]. For biosensing applications, magnetic nanoparticles possess several advantages over microparticles in detecting biological molecules. They have a higher magnetization density, and they are able to form much more stable suspensions. Also, they have less tendency to settle in microfluidics with faster velocities in solution. For the purpose of sensing, the recognition and specific capture of biological entities on the magnetic nanoparticle surface are critical for numerous bioanalytical applications in bio- and immunosensor diagnostic devices. The immobilization of biorecognition molecules (e.g. nucleic acids, proteins and other ligands) on magnetic nanoparticles offers modified surface properties and this biofunctionalization plays an important role in the development of such devices The Chemistry of reactive groups Every chemical modification or conjugation process involves the reaction of one functional group with another, resulting in the formation of a covalent bond. The creation of bioconjugate reagents with spontaneously reactive or selectively reactive functional groups forms the basis for simple and reproducible crosslinking or tagging of target molecules. Of the hundreds of reagent systems described in the literature or offered commercially, most utilize common organic chemical principles that can be reduced down to a couple dozen or so primary reactions. An understanding of these basic reactions can provide insight into the properties and use of bioconjugate reagents even before they are applied to problems in the 16

23 laboratory. Some of the reagents are not themselves crosslinking or modification compounds, but may be used to form active intermediates with another functional group. These active intermediates subsequently can be coupled to a second molecule that possesses the correct chemical constituents, which allows bond formation to occur. Amine reactions Reactive groups able to couple with amine- containing molecules are by far the most common functional groups present on crosslinking or modification reagents. An amine- coupling process can be used to conjugate with nearly all protein or peptide molecules as well as a host of other macromolecules. The primary coupling reactions for modification of amines proceed by one of two routes: acylation or alkylation. Most of these reactions are rapid and occur in high yield to give stable amide or secondary amine bonds. Schiff base formation Aldehydes and ketones can react with primary and secondary amines to form Schiff bases, a dehydration reaction yielding an imine. However, Schiff base formation is a relatively labile, reversible interaction that is readily cleaved in aqueous solution by hydrolysis. The formation of Schiff bases is enhanced at alkaline ph values, but they are still not stable enough to use for crosslinking applications unless they are reduced by reductive amination. Reductive amination Reductive amination (or alkylation) may be used to conjugate an aldehyde or ketone containing molecule with an amine- containing molecule. Schiff base formation between aldehydes and amines occurs readily in aqueous solutions, especially at elevated ph. This type of linkage, however, is not stable unless reduced to secondary or tertiary amine bonds. A number of reducing agents can be used to convert specifically the Schiff base interaction into an alkylamine linkage. Once reduced, the bonds are highly stable and will not readily hydrolyze in aqueous environments. The use of reductive amination to conjugate an aldehyde containing molecule to an amine containing molecule results in a zero- length crosslinking procedure where no additional spacer atoms are introduced between the molecules. 17

24 Figure 1.4 Carbonyl groups can react with amine nucleophiles to form reversible Schiff base intermediates. In the presence of a suitable reductant, such as sodium cyanoborohydride, the Schiff base is stabilized to a secondary amine bond. 1.8 Semiconductor Quantum Dots (QDs) Quantum dots (QDs) are nanoparticles typically made of a semiconductor arranged in a spherical crystalline core and capped with a shell consisting of a second metal alloy composition (figure 1.5). Figure 1.5 The structure of a typical QD nanocrystal includes a semiconductor alloy core surrounded by a shell consisting of a different ally structure. Early QD composition involved the use of CdSe core with a ZnS shell, but many different alloy compositions have been used and are possible. 18

25 The size of this raw core/shell construct is usually less than 10nm in diameter or about the same sizes as many globular protein molecules. Upon exposure to light at the appropriate wavelength range, QDs are able to absorb a photon of energy, which results in the excitation of an electron within the core. The exciton is confined within the nanocrystal, because its core diameter is less than the exciton Bohr radius, which thus leads to quantum confinement. The shell structure aids in this confinement and prevents the electron from tunneling out of the core and escaping into the outer medium or undergoing non- radiative deactivation. The size and shape of a QD governs the discrete energy levels that the excited- state electron can attain within it, thus dots can be tuned to have desired electronic properties by careful adjustment of core diameter and composition. Upon return of the electron to its ground state, a radiative QD emits a photon of light, the wavelength of which is dependent on the alloy material type and its diameter [49] (figure 1.6). Figure 1.6 The size of a QD directly affects its emission wavelenght. Careful control of nanocrystal diameter during the manufacturing process can result in discrete QD populations having emission properties ranging from the blue to the red region of the visible spectrum. As a result of their unique optical and electronic properties, particularly their ability to fluoresce at discrete wavelengths directly proportional to their sizes and material compositions, QDs have found use in many fields, including electronics, biology, medicine, 19

26 and even cosmetics. The first attempts to modify their surface characteristics to make them water- soluble and bio- compatible eventually led to their use as fluorescent labels for biomolecules in many applications [50, 51, 52]. QDs have a number of advantages over other fluorescent molecules such as organic dyes, including: (1) resistance to photobleaching, which allows them to be imaged over long periods without loss of fluorescence; (2) narrow, nearly symmetrical emission peaks with no red- shift tail typical of organic fluors, thus creating a bright fluorescence signal at characteristic wavelengths; (3) a broad absorbance band, which increases almost exponentially toward shorter wavelengths with extremely high extinction coefficients ( M - 1 cm - 1 ); (4) the ability to excite at a single wavelength an entire family of QDs having different emission characteristics, thus providing multiplexed assay capability; (5) the capacity to design QDs with emission characteristics ranging from the low visible wavelengths to well within the IR region; and (6) the potential for relatively high quantum yield (QY) of fluorescence ( for CdSe). The emission properties of QDs can be adjusted based upon core diameter and nanoparticles composition. Nanoparticles diameters typically are carefully controlled during manufacture to be between 2 and 10 nm. In addition, the band gap energy or energy of fluorescence emission is inversely proportional to the diameter of the QD particle. Thus, the smaller the particle, the more blue- shifted is its emission and the larger the QD, the more red- shifted is its emission bands. QDs also have an intrinsic color to their solutions that corresponds to the size of the particles and their fluorescence emission characteristics. However, to create a single particle population with a tight fluorescence emission pattern, the diameter of the particles must be controlled to well within a nanometer. The emission peak width is directly proportional to the size distribution of a particle population. This makes manufacturing reproducible QDs a constant challenge for most suppliers that rely on size to control fluorescence properties. However, as opposed to the difficulty of tuning emission properties by particle diameter, QD alloy composition instead may be adjusted independent of size to control the wavelength of emission for a given particle population. In a QD having a concentration gradient composition, the concentration of an alloy of a first semiconductor gradually increases from the core to the surface of the particle, while the concentration of a second semiconductor gradually decreases from the core to the surface [53]. A third semiconductor type also may be added to fine- tune further the emission properties. By careful adjustment 20

27 of these semiconductor concentration gradients, QD populations can be made having discrete emission properties without changing the particle size. Therefore, tuning QD spectral characteristics can be done using a single particle size and by making selective changes to the alloy composition. This avoids the difficulties in manufacturing particles of uniform size, because all particle populations can have the same size, but only vary in their relative semiconductor gradient concentrations to attain particles having discrete fluorescence character. The material types making up the core of a QD also affect the range of emission wavelengths that can be attained QDs synthesis QDs have been made using a number of techniques. Nearly all of the Groups semiconductors have been prepared in colloidal form, in a variety of different approaches [54, 55]. QDs synthesis can be tailored to specific requirements; the achievement of desired particle sizes over the largest possible wavelength range, narrow size distributions within 2% [56], good crystallinity, desired surface properties and in some cases high luminescence quantum yields, as well as adjustable electronic properties, are all results that are considered to be characteristics of a good preparation. Choice of the QDs and capping (or stabilizer) are gaining importance, as the type of QD to some extent alters the photophysical properties, whilst the capping confers properties to the QD which allow its incorporation into a desired application. The nature and concentration of the initial materials, the capping, which keeps the particles in solution, solvent, ph value, reaction time, temperature and atmosphere, are some of the parameters that influence a colloidal synthesis of QDs nanocrystals [57]. The most successful preparations are based on the reliable separation of nucleation and growth with following size- selective precipitation. Most procedures developed for obtaining high- quality QDs are based on variations of the high- temperature pyrolytic reaction [58] which involves the fast injection of liquid precursors (combination of an appropriate metallic or organometallic precursor (zinc, cadmium or mercury species) with a corresponding chalcogen (sulfur, selenium or tellurium species) into a coordinating solvent e.g. trioctylphosphine oxide (TOPO), trioctylphosphine (TOP) or hexadecylamine, at high temperatures under inert conditions [59-61]. The QDs obtained in this way are hydrophobic. 21

28 1.8.2 Water- soluble colloidal nanocrystals To use QDs in biological applications, the particles must be rendered biocompatible by coating with a hydrophilic layer that masks the surface, thus preventing aggregation and nonspecific binding. This is not a trivial problem, as the successful commercialization of QDs for biomolecule labeling took at least 5 years from the time the first two papers appeared in Science describing water- soluble particles for bioconjugation [51, 52]. The fact is, these early particles were not very soluble in aqueous environments and tended to clump together or bind nonspecifically with biomolecules. The initial modifications done to covalently link molecules to QDs need to displace the TOPO or detergent coating on the raw nanocrystal surface with a new organic derivative imparting water solubility. The first attempts at making biocompatible QDs all involved the use of simple monothioacids, such as thioacetic acid, which can link to the shell through thiol dative bonding and provide a terminal carboxylate for further conjugation (figure 1.7). Figure 1.7 Schematic image of a QD nanocrystal core surrounded by solubilizing ligands. However, monothiol linkers easily can oxidize back off QDs and leave behind surface gaps, which become hydrophobic sites for particle clumping and nonspecific binding to biomolecules. A better approach is to use a dithiol compound, which forms two dative bonds per linker on the QD surface. This makes the linkage to the QD resistant to oxidation and prevents nonspecific surface gaps from forming. One such dithiol compound that is particularly useful is dihydrolipoic acid (DHLA), which contains a carboxylate group on the 22

29 other end. QDs modified with DHLA subsequently can be modified with polyethylene glycol (PEG) groups to provide increased hydrophilicity of the surface or directly linked to proteins via electrostatic interactions or through an EDC- mediated reaction [62-66]. The negative charge character of DHLA- modified QDs has been used to link noncovalently positively charged proteins, such as avidin [67] or a recombinant protein containing a positively charged fusion peptide. In this regard, the highly positive leucine zipper peptide has been used as a fusion tag [68] as well as a penta- histidine peptide tag [69] and esa- histidine tag [70, 71]. Combinations of positively charged fusion proteins and avidin also have been used to control the resultant biotin binding density on a DHLA- QD surface for use in live cell imaging [72]. One major advantage of DHLA modification is that the diameter of the QD remains as small as possible, while still creating a water- soluble particle. QDs of 10nm diameter have been created using this process and were successfully used to image intracellular proteins [73]. Another type of simple surface modification involves the noncovalent coating of the QDs with detergents or lipids. The hydrophobic tails of these molecules bind to the QD particle, while the hydrophilic portions interact with the aqueous phase and render the dots dispersible. Still other surface modification schemes use polymeric coatings containing multiple binding points to the QD, thus eliminating the possibility for leaching. Coatings containing PEG spacers also can be used to create a highly hydrophilic layer on top of the semiconductor surface. All of these modification strategies provide QDs that are water- soluble (or dispersible and stable in suspension) and that contain functional groups for covalent attachment of proteins or other affinity molecules. Masking the surface of semiconductor QD particles also can be done by adding another inorganic layer to the outer shell alloy structure. This layer can take the form of a silica coating formed by the reaction of a silane derivative with the shell [74]. For instance, a pure silica surface can be created by controlled polymerization of the raw nanocrystals with tetraethyl orthosilicate (TEOS), which forms a siliceous sphere with silanol groups on the outer surface. Another organosilane derivative that is appropriate for use with metallic particles is mercaptopropyl- tris- hydroxy- silane. The thiol groups on the silane compounds datively bind to the surface while the hydroxy- silane groups polymerize to form a new silica coating. The resultant silanol- containing surface then can be functionalized using other organosilane compounds containing functional groups or reactive groups for further conjugation with biomolecules. The only disadvantage of this approach is the increasingly 23

30 greater particle diameter that results from building successive layers on the initial QD core, which may inhibit their use for probing within cells or tissues. Most QD surfaces for biological applications contain negatively charged carboxylates for conjugation with amine- containing molecules via a carbodiimide reaction with EDC and (sulfo)nhs. The negative charges on the QD surface prevent particle aggregation through like charge repulsion. An alternative method of creating water dispersible dots is to form a hydrophilic coating that carries along with it a layer of hydration consisting of hydrogen- bonded water molecules. This often is done using hydroxylic polymers or PEG modifications. This too prevents aggregation due to the high energy needed to remove the bound water layer. QDs have been used successfully in many biological applications, which exploit their best properties of brightness, photostability, and multiplex capability. There are many publications that use QDs for in cell or whole organism- based imaging, including tracking of targets within cells [75], gene localization within chromosomes [76], embryo developmental monitoring [77], tumor imaging in vivo [78], and multiplexed imaging and assays [69, 79], including FRET signaling [80]. For a review on the use of QDs for cancer imaging and treatment, see [81] Conjugation to QDs Many antibodies, proteins, and other targeting or affinity ligands have been conjugated to QDs for biological applications. Antibodies to tumor markers have been used to image cancer cells in vivo [82], fluoroimmunoassays have been developed using antibody- conjugated QDs [83], peptide QD conjugates have been made to target proteins in vivo [84], antibody QD conjugates containing tumor toxic agents have been designed to image and kill tumor cells [85], and sugar QD conjugates have been made to detect carbohydrate binding proteins [86]. The conjugation of proteins and other molecules to QDs involves standard coupling reactions with the added caveat related to the potential difficulties of working with particles. Many of the coupling strategies for dealing with nanoparticles and microparticles are valid for use with QDs, but it is best when using commercially available particles to pay close attention to the manufacturer s suggested protocols. When designing a protein QD conjugate, it is also important to consider the optimal number of proteins to be coupled per particle. In some applications, a low ratio of protein- 24

31 to- QD may result in the highest signal for fluorescence detection. This often is the case for antibody QD conjugates where two to three antibodies coupled per particle are sufficient to target antigens. However, other applications may require a high density of protein on the QD surface. As is the case with most nanoparticles, buffer and salt compatibility should be taken into account when working with QDs. Small charged particles maintain colloidal stability by like charge repulsion, which prevents aggregation by van der Waals or hydrophobic attraction. Any salt or buffer constituents that neutralize or eliminate the surface charge on the QDs will cause clumping or precipitation. Some buffer additives also will cause loss of fluorescence intensity and should be avoided in concentrations above a certain level. Figure 1.8 Schematic illustration of different approaches of conjugation of QDs to biomolecules: (a) Use of a bifunctional ligand to link QDs to biomolecules; (b) TOPO- capped QDs bound to a modified acrylic acid polymer by hydrophobic forces; (c) QDs solubilization and bioconjugation using a mercaptosilane compound; (d) Electrostatic attraction of positively charged biomolecules and negatively charged QDs (e) Incorporation of QDs into microbeads and nanobeads. References [1] P. Moriarty, Rep. Prog. Phys., 2001, 64, 297. [2] A.I. Gusev, A.A. Rampel, Nanokristallicheskie Materialy (Nanocrystalline Materials), Moscow: Fizmatlit, [3] J.- T. Lue, J. Phys. Chem. Sol., 2001, 62, [4] J. Jortner, C.N.R. Rao, Pure Appl. Chem., 2002, 74, [5] V.E. Fertman, Magnetic Fluids Guide- Book: Properties and Application, Hemisphere, New York,

32 [6] B.M. Berkovsky, V.F. Medvedev, M.S. Krakov, Magnetic Fluides: Engineering Applications, Oxford University Press, Oxford, [7] J.L. Kirschvink, A. Kirschvink- Kobayashi, B.J. Woodford, Proc. Natl. Acad. Sci., 1992, 89, [8] D.G. Rancourt, Rev. Mineral. Geochem., 2001, 44, 217. [9] S.P. Gubin, Yu.A. Koksharov, G.B. Khomutov, G.Yu. Yurkov, Russian Chem. Rev., 2005, 74, 489. [10] S.P. Gubin, Yu.A. Koksharov, Neorg. Mater., 2002, 38, [11] R. Ferrando, J. Jellinek, R.L. Johnston, Chem. Rev., 2008, 108, 845. [12] S.P. Gubin, Ros. Khim. Zh., 2000, 44(6), 23. [13] R.M. Cornell, U. Schwertmann, The Iron Oxides: Structure, Properties, Reactions, Occurrences and Uses, 2nd ed.; Wiley- VCH: Weinheim, [14] P. Tartaj, M.P. Morales, S. Veintemillas- Verdaguer, T. Gonzalez- Carren, C.J. Serna, J. Magn. Magn. Mater., 2005, , 28. [15] C. Feldmann, H.- O. Jungk, Angew. Chem., Int. Ed. 2001, 40, 359. [16] R. Vijaya Kumar, Yu. Koltypin, Y.S. Cohen, Y. Cohen, D. Aurbach, O. Palchik, I. Felner, A. Gedanken, J. Mater. Chem., 2000, 10, [17] Z. Li, H. Chen, H. Bao, M. Gao, Chem. Mater., 2004, 16, [18] D.R. Lovley, Microbiol. Rev., 1991, 55, 259. [19] T. Fried, G. Shemer, G. Markovich, Adv. Mater., 2001, 13, [20] I. Nedkov, T. Merodiiska, S. Kolev, K. Krezhov, D. Niarchos, E. Moraitakis, Y. Kusano, J. Takada, Monatsh. Chem., 2002, 133, 823. [21] S. Sun, H. Zeng, J. Am. Chem. Soc., 2002, 124, [22] X. Wang, J. Zhuang, O. Peng, Y. Li, Nature, 2005, 437, 121. [23] Y. Hou, J. Yu, S. Gao, J. Mater. Chem., 2003, 13, [24] Yu.F. Krupyanskii, I.P. Suzdalev, Zh. Eksp. Teor. Fiz., 1974, 67, 736. [25] W.W. Yu, X. Peng, Angew. Chem. Int. Edn., 2002, 41, [26] C. Alexiou, R. Jurgons, C. Seliger, H. Iro, J. Nanosci. Nanotechnol., 2006, 6, [27] M. Bruchez, M. Moronne, P. Gin, S. Weiss, A.P. Alivisatos, Science, 1998, 281, [28] W.C.W. Chan, S.M. Nie, Science, 1998, 281, [29] S. Wang, N. Mamedova, N.A. Kotov, W. Chen, J. Studer, Nano Letters, 2002, 2, 817. [30] C. Mah, I. Zolotukhin, T.J. Fraites, J. Dobson, C. Batich, B.J. Byrne, Mol. Therapy, 2000, 1,

33 [31] D. Panatarotto, C.D. Prtidos, J. Hoebeke, F. Brown, E. Kramer, J.P. Briand, S. Muller, M. Prato, A. Bianco, Chem. Biol., 2003, 10, 961. [32] R.L. Edelstein, C.R. Tamanaha, P.E. Sheehan, M.M. Miller, D.R. Baselt, L.J. Whitman, R.J. Colton, Biosens. Bioelectron., 2000, 14, 805. [33] J.M. Nam, C.C. Thaxton, C.A. Mirkin., Science, 2003, 301, [34] R. Mahtab, J.P. Rogers, C.J. Murphy, J. Am. Chem. Soc., 1995, 117, [35] J. Ma, H. Wong, L.B. Kong, K.W. Peng, Nanotechnology, 2003, 14, 619. [36] A. de la Isla, W. Brostow, B. Bujard, M. Estevez, J.R. Rodriguez, S. Vargas, V.M. Castano, Mater. Res. Innovat., 2003, 7, 110. [37] J. Yoshida, T. Kobayashi, J. Magn. Magn. Mater., 1999, 194, 176. [38] R.S. Molday, D. MacKenzie, J. Immunol. Methods, 1982, 52, 353. [39] R. Weissleder, G. Elizondo, J. Wittenburg, C.A. Rabito, H.H. Bengele, L. Josephson, Radiology, 1990, 175, 489. [40]W.J. Parak, R. Boudreau, M.L. Gros, D. Gerion, D. Zanchet, C.M. Micheel, S.C. Williams, A.P. Alivisatos, C.A. Larabell, Adv. Mater., 2002, 14, 882. [41] V.A. Sinani, D.S. Koktysh, B.G. Yun, R.L. Matts, T.C. Pappas, M. Motamedi, S.N. Thomas, N.A. Kotov, Nano Lett., 2003, 3, [42] Y. Zhang, N. Kohler, M. Zhang, Biomaterials, 2002, 23, [43] L.G. Gutwein, T.J. Webster, J. Nanoparticle Res., 2002, 4, 231. [44] N. Gaponik, D. V. Talapin, A. L. Rogach, K. Hoppe, E. V. Shevchenko, A. Kornowski, A. Eychmüller, H. Weller, J. Phys. Chem. B, 2002, 106, [45] X. Wang, J. Zhuang, Q. Peng, Y. Li, Nature, 2005, 437, 121. [46] C. Pascal, J.L. Pascal, F. Favier, M.L. Elidrissi Moubtassim, C. Payen, Chem. Mater., 1999, 11, 141. [47] A.K. Gupta, M. Gupta, Biomaterials, 2005, 26, [48] Pankhurst, Q.A., et al., J. Phys. D: Appl. Phys., 2003, 36(R), 167. [49] A.P. Alivisatos, Science, 1996, 271, 933. [50] A.L. Rogach, L. Katsikas, A. Kornowski, D. Su, A. Eychmuller, H. Weller, Ber. Bunsenges. Phys. Chem., 1996, 100, [51] M. Bruchez, M. Moronne, P. Gin, S. Weiss, A.P. Alivisatos, Science, 1998, 281, [52] W.C.W. Chan, S. Nie, Science, 1998, 281, [53] S. Nie, R.E. Bailey, US Patent Application 2007/ A1. [54] A. Henglein, Chem. Rev., 1989, 89,

34 [55] Y. Wang, N. J. Herron, Phys. Chem., 1991, 95, 525. [56] S. Santra, K.M. Wang, R. Tapec, W.H. Tan., J. Biomed. Opt., 2001, 6, 160. [57] A. Eychmüller, J. Phys. Chem. B., 2000, 104, [58] C.B. Murray, D.J. Norris, M.G. Bawendi, J. Am. Chem. Soc., 1993, 115, [59] L. Qu, Z.A. Peng, X. Peng, Nanolett., 2001, 1, 333. [60] D.V. Talapin, A.L. Rogach, A. Kornowski, M. Haase, H. Weller, NanoLett., 2001, 1, 207. [61] X. Peng, J. Wickham, A. P. Alivisatos, J. Am. Chem. Soc., 1998, 120, [62] H. Mattoussi, J.M. Mauro, E.R. Goldman, G.P. Anderson, V.C. Sundar, F.V. Mikulec, M.G. Bawendi, J. Am. Chem. Soc., 2000, 122, [63] H.T. Uyeda, I.L. Medintz, H. Mattoussi, Mater. Res. Soc. Proc. 789, Symposium N, [64] I.L. Medintz, J.H. Konnert, A.R. Clapp, I. Stanish, M.E. Twigg, H. Mattoussi, J.M. Mauro, J.R. Deschamps, Proc. Natl. Acad. Sci. USA, 2004, 101(26), [65] N. Anikeeva, T. Lebedeva, A.R. Clapp, E.R. Goldman, M.L. Dustin, H. Mattoussi, Y. Sykulev, Proc. Natl. Acad. Sci. USA, 2006, 103(45), [66] A.R. Clapp, E.R. Goldman, H. Mattoussi, Nat. Protoc., 2006, 1(3), [67] E.R. Goldman et al., J. Am. Chem. Soc., 2002a, 124, [68] E.R. Goldman et al., Anal. Chem., 2002b, 74, 841. [69] I.L. Medintz, A.R. Clapp, H. Mattoussi, E.R. Goldman, B. Fisher, J.M. Mauro, Nat. Mater., 2003, 2, 630. [70] I.L. Medintz, L. Berti, T. Pons, A.F. Grimes, D.S. English, A. Alessandrini, P. Facci, NanoLett., 2007, 7, [71] L. Berti, P.S. D Agostino, I.L. Medintz, Nano Res., 2009, 2, 121. [72] J.K. Jaiswal, E.R. Goldman, H. Mattoussi, S.M. Simon, Nat. Meth., 2004, 1(1), 73. [73] J.K. Jaiswal, H. Mattoussi, J.M. Mauro, S.M. Simon, Nat. Biotechnol., 2003, 21(1), 47. [74] M. Darbandi, T. Nann, Chem. Mater., 2005, 17, [75] M. Dahan, S. Levi, C. Luccardini, P. Rostaing, B. Riveau, A. Triller, Science, 2003, 302, 442. [76] Y. Xiao, P.E. Barker, Nucleic Acids Res., 2004, 32, e28/1. [77] B. Dubertret, P. Skourides, D.J. Norris, V. Noireaux, A.H. Brivanlou, A. Libchaber, Science, 2002, 298, [78] X. Gao, Y. Cui, R.M. Levenson, L.W. Chung, S. Nie, Nat. Biotechnol., 2004, 22, 969. [79] X. Wu, H. Liu, J. Liu, K.N. Haley, J.A. Treadway, J.P. Larson, N. Ge, F. Peale, M.P. Bruchez, Nat. Biotechnol., 2003, 21,

35 [80] M. Han, X. Gao, J.Z. Su, S. Nie, Nat. Biotechnol., 2003, 19, 631. [81] S.K. Vashist, R. Tewari, R.P. Bajpai, L.M. Bharadwaj, R. Raiteri, J. Nanotechnol. Online, 2006, 2, 1. [82] H. Tada, H. Higuchi, T.M. Wanatabe, N. Ohuchi, Cancer Res., 2007, 67, [83] E.R. Goldman, H. Mattoussi, G.P. Anderson, I.L. Medintz, J.M. Mauro, Meth. Mol. Biol., 2005, 303, 19. [84] J.M. Ness, R.S. Akhtar, C.B. Latham, K.A. Roth, J. Histochem. Cytochem., 2003, 51, 981. [85] X. Gao, Y. Cui, R.M. Levenson, L.W. Chung, S. Nie, Nat. Biotechnol., 2004, 22, 969. [86] P. Babu, S. Sinha, A. Surolia, Bioconjugate Chem., 2007, 18,

36 Chapter 2. Methods of synthesis To accomplish the design criteria for nanoparticles to be used in biomedical applications, we have synthesized magnetic nanoparticles such as magnetite (Fe3O4) by different techniques such as the co- precipitation method, from which we have observed the formation of small clusters of magnetic nanoparticles, and the thermo- decomposition method, from which we have obtained single nanoparticles with a narrow size distribution. These magnetic nanoparticles were modified with APTES and dopamine covalently bonded onto the nanoparticles surface using ligand exchange. 2.1 Synthesis and stabilization of magnetic iron oxide nanoparticles Physical methods such as gas phase deposition and electron beam lithography can be used to synthesize magnetic nanoparticles but such methods are elaborate and suffer from the inability to control the size of particles in the nanometer size range [1]. Numerous chemical methods such as microemulsions [2], sol- gel syntheses [3], hydrothermal reactions [4], hydrolysis and thermolysis of precursors [5], flow injection syntheses [6] and electrospray syntheses [7] can also be used in synthesizing magnetic nanoparticles. The most common method for the production of magnetic iron oxide nanoparticles is the wet chemical coprecipitation [8-11] of Fe 2+ and Fe 3+ (molar ratio of 1:2) aqueous salt solutions by addition of a base as shown by Equation 1 with a reaction mechanism shown in Scheme 2.1 [12]. The precipitated magnetite is black in colour. Appreciable control over size, composition and sometimes even the shape of the nanoparticles is possible in this method. Fe Fe OH - Fe 3 O 4 + 4H 2 O (1) 30

37 Complete precipitation of Fe3O4 should be expected at a ph between 8 and 14, with a stoichiometric ratio of 2:1 (Fe 3+ /Fe 2+ ) [13], otherwise, Fe3O4 can be oxidized as shown by equation 2, or transformed into maghemite (γfe2o3) in the presence of oxygen due to Fe3O O H2O 3Fe(OH)3 (2) its instability and sensitivity to oxygen as shown by equation 3. Fe3O4 + 2H + γfe2o3 + Fe 2+ + H2O (3) Figure 2.1 Reaction mechanism of formation of magnetite nanoparticles through precipitation. This can critically affect the physical and chemical properties of the magnetic particles. In order to prevent magnetite from agglomeration and possible oxidation in air, great care is taken to carry out the reaction under inert conditions so as to control the reaction kinetics, which is strongly related with the oxidation speed of iron species. This also helps in reduction of the particle size when compared with methods that are not oxygen free [11]. Fe3O4 nanoparticles produced by co- precipitation reaction are also usually coated with organic anions like carboxylate e.g. citric acid, polymer surface complexing agents e.g. dextran or inorganic materials like silica, gold or gadolinium [14-16] during the precipitation process. This provides stability in solution, size control and also, ideal anchorage at the nanoparticles surface for binding various biological ligands for biomedical applications. They are also known to provide better protection against toxicity. Iron oxide nanoparticles 31

38 can be capped by in situ and post- synthesis capping with much care to avoid formation of a non- magnetic surface [17]. The relative success in the control of size, shape, growth process and composition of nanoparticles depends on the type of salts used (e.g. chlorides, sulphates, nitrates, perchlorates, etc.), Fe 2+ to Fe 3+ concentration ratio, ph, temperature and ionic strength of the media [18-20]. Studies have shown that magnetic nanoparticles prepared with Fe 2+ /Fe 3+ ions having an optimum concentration between 39 and 78 mm are effective enough to be used as contrast agents [20-22]. It has also been documented that the higher the ph and ionic strength, the smaller the particle size, size distribution and width, because these parameters determine the chemical composition of the crystal surface and consequently the electrostatic surface charge of the particles [23]. Figure 2.2 Scheme of procedure of coprecipitation method. A widely used technique for magnetite nanoparticle synthesis is the high- temperature (~ 265 C) reaction of Fe(acac) 3, iron (III) acetylacetonate, in phenyl ether with alcohol, oleic acid and oleylamine [24]. This approach is able to produce monodisperse magnetite nanoparticles. The dimension of nanoparticles can be controlled. Large monodisperse magnetite nanoparticles (up 2 to 20 nm in diameter) can be synthesized by simply using the smaller nanoparticles as the seeds for further particle growth. The as- prepared nanoparticles have a narrow size distribution, hence no subsequent size- selection process is required after the synthesis. The as- prepared magnetite nanoparticles can be dispersed into nonpolar 32

39 solvent readily. And this seed- mediated growth method has the potential for mass production by scale up processes. The mechanism leading to Fe 3 O 4 in the reactions presented is not yet clear. However, evidence suggests that reduction of the Fe(III) salt to an Fe(II) intermediate occurs, followed by the decomposition of the intermediate at high temperature. Sun et al. [24] reported the synthesis of monodisperse Fe 3 O 4 nanoparticles with sizes varying from 3 to 20 nm in diameter, based on a simple organic phase process. A typical process involves the high- temperature decomposition (> 250 C) of an organic iron precursor, such as iron(iii) acetylacetonate, in the presence of surfactants, such as oleic acid. Figure 2.3 Schematic illustration of organic phase synthesis. The nanoparticles resulting from these procedures are stable in nonpolar solvents (such as hexane) and capped with nonpolar endgroups on their surface. The capping molecules (also called ligands) are typically long- chain alkanes with polar groups that bind to the nanoparticles surface. Oleic acid is widely used in ferrite nanoparticle synthesis because it can form a dense protective monolayer, thereby producing highly uniform and monodisperse particles. Oleic acid is chemisorbed on the particle surface as a carboxylate with its nonpolar CH 3 endgroup sticking out into solution. The nonpolar endgroups render these particles stable in nonpolar solvents because they allow the creation of a surface that is mutually unreactive and repulsive, which is commonly considered to be steric stabilization. From metodologies presented in this section, we have selected the the co- precipitation method [22] to obtain small clusters of magnetic nanoparticles and the thermo- decomposition method to yield single nanoparticles with a narrow size distribution [24]. These nanoparticles can be dispersed in aqueous media after a ligand exchange of the oleic 33

40 acid by a silane- functional molecule [25] and dopamine [26] molecule and modified with biological molecules. Figure 2.4 Set up representation of the thermo- decomposition method. Magnetic nanoparticles with a narrow size distribution were obtained from this process. 2.2 Surface modification Previous studies of surface modification of magnetic nanoparticles to be used in biomedical applications involved the functionalization of these nanoparticles with materials such as poly(ethylene glycol) (PEG) [27-29], aminosilane [30-32], dimercaptosuccinic acid (DMSA) [33, 34], thermo- responsive polymers with upper and lower critical solution temperatures such as polycaprolactone [35] and homopolymers of N- n- propylacrylamide (NNPAM) [36-38], and polysaccharide molecules such as dextran [39-42]. Dextran and PEG are the most commonly selected coatings for biomedical applications such as magnetic resonance imaging (MRI) [43-45], magnetic drug targeting [46-48], and magnetic fluid hyperthermia (MFH) [30, 45, 49, 50], because they can enhance the plasma half- life of the magnetic nanoparticles in the blood stream, avoid nanoparticle aggregation, and improve cellular uptake of the particles. Dextran coated superparamagnetic iron oxide nanoparticles (SPIO) have been approved for the Food and Drug Administration (FDA) as contrast agents for MRI applications [51]. Moreover, surface functionalization of magnetic nanoparticles with thermo- responsive polymers has been studied for drug delivery [14, 35, 34

41 52], due to changes in their matrix structure at specific critical solution temperatures resulting in release of a drug Aminopropyltriethoxysilane (APTES) coating Silanes are bifunctional molecules with the general chemical formula Y- (CH 2 ) n - Si- R 3, where Y represents the headgroup functionality, (CH 2 ) n an alkane chain, and Si- R 3 the anchor group by which the silane will be grafted to the oxide surface. First of all, why silica as a coating? The main reason is that the surface chemistry of the silica is very well known. This means that one can easily functionalize the silica surface and adapt this nanoparticles to many applications. Surface modification with organosilanes is an attractive approach in this context as it is compatible with many of the materials used in a biological context, i.e., silica gel, glass slides, or silicon wafers. Silanization in solution [25] is the most common approach since it does not require the use of volatile compounds and to be performed under vacuum. Typical silane concentrations lie between 0.1 and 2% (v/v) and the presence of a small amount of water is necessary for the reaction to occur. Figure 2.5 APTES activation in water and linkage with MNPs. During the surface activation reaction, two reactions take place simultaneously with respect to the silane: the hydrolysis of the n silane alkoxy groups to the highly reactive silanol species, and the condensation of the resultant silanols with the free OH groups of the surface to render stable Si- O- Si bonds. Oligomerization in solution also occurs as a 35

42 competing reaction with covalent binding to the surface and constitutes an important consideration specially when n = 2 or 3. The procession of these reactions and consequently the characteristics of the final surface layer depend on reaction variables such as solvent type, temperature, or time, as well as on the catalyst and organosilane concentrations used. The ways in which these parameters influence the surface modification process are complicated and various. Frequently a factor can lead to opposing effects by positively and negatively influencing different stages of the overall process. As an example, hydrophilic and protic solvents (like alcohols) usually accelerate hydrolysis and condensation kinetics, thus promoting the surface modification process. However, they can also compete with the silane for surface silanol groups by H- bonding. Solvent molecules can also invert the silane hydrolysis reaction or even form stable complexes with the hydrolyzed species that can lower their reactivity. Additionally, the reaction temperature can affect the outcome of the process. For example, raising the reaction temperature accelerates the silane condensation kinetics, both onto the surface and in solution, and can lead to the formation of silane aggregates. Small silane aggregates that result from condensation in solution may still be able to react with the surface (through unreacted silanol groups), but if solution- phase oligomerization goes too far, large aggregates will be inhibited sterically from reacting with surface - OH groups. This situation is even more complicated in the case of surface modification of nanoparticles. Particle aggregation phenomena and interparticle cross- linking through siloxane bridges may appear as a consequence of uncontrolled reaction conditions. Moreover, issues such as the extent of the modification (the thickness of the surface layer and the density of functional groups), and the homogeneity of the coating may be crucial for the final performance of the surface- activated material. As an example, the sensitivity of a diagnostic test will be limited by the number (density) and distribution of surface groups that are able to interact with the complementary/target molecular species. 2.3 Surface functionalization via ligand exchange For magnetic particles covered by small molecules, a direct method to replace the small molecules with polymer surfactants is by a ligand exchange process. As shown in figure 2.5, the ligand exchange process involves a competing reaction where a second ligand that binds strongly onto the nanoparticle surface is used to replace the weakly bound ligand [25]. In order to improve the ligand exchange efficiency, the second ligand should have a high 36

43 concentration. After the exchange process, free ligands are removed by magnetic separation, precipitation or other methods. Figure 2.6 Exchange of the chemisorbed oleic acid ligands on the nanoparticle surface by silanes and dopamine Silane exchange To make MNPs dispersible in water, the hydrophobic ligands must be exchanged by hydrophilic ones. It involves adding an excess of silane to the nanoparticles solution, which results in the displacement of the oleic acid (original ligand) on the nanoparticles surface. Figure 2.7 Exchange of the chemisorbed oleic acid ligands on the nanoparticle surface by silanes. 37

44 The binding of oleates is noncovalent, while silanes are known to form a densely packed film, which is covalently linked to the surface Dopamine exchange The ligand exchange process was carried out using dopamine [26]. Dopamine presents un amino- group necessary for biofunctionalization. Both oleic acid and dopamine are covalently bound to the surface of MNPs via a chelating bidentate interaction to the iron species [14, 53]. At room- temperature ligand exchange reaction using dopamine can be used to convert the NPs from the hydrophobic to the hydrophilic state. The dopamine coated nanoparticles are stable in water thanks to electrostatic stabilization of the amino groups. 2.4 Magnetic fluid properties Ferrofluids are stable colloidal suspensions of magnetic nanoparticles dispersed in a liquid. In figure 2.7 one can figure out what a ferrofluid is. The magnetic properties of a material depend on the interaction between its magnetic dipole moments and an external magnetic field [35]. Materials that have weak dipole moments with random orientation are called paramagnetic. In this case, magnetic moments are only aligned under the application of a magnetic field [36]. A variation of this magnetic behavior is known as superparamagnetism, in which a large magnetization is observed even at low magnetic fields. Ferrofluids are superparamagnetic at room temperature. Figure 2.8 Each nanoparticle constitutes a single magnetic domain.here you can see a single nanoparticle. What is important is that each nanoparticle constitutes a single magnetic domains. This means that each nanoparticle reacts to an external magnetic field separately, by orienting its magnetic moment along the external field. In the figure, a nanoparticle stabilized by the hydrophobic tails of a surfactant. 38

45 One important advantage for the MNPs is their superparamagnetism that enables their stability and dispersion upon removal of the magnetic field as no residual magnetic force exist among the particles. Below approximately 15 nm, such particles are so small that the cooperative phenomenon of ferromagnetism is no longer observed and no permanent magnetization remains after the particles have been subject to an external magnetic field. However, the particles still exhibit very strong paramagnetic properties (hence the name of the phenomenon) with a very large susceptibility. In MR imaging, superparamagnetic particles made of iron oxide can be used as contrast agents. They strongly influence T1 relaxation and T2 relaxation, the latter depending strongly on the size and coating of the particles [54]. Particles whose unpaired electron spins align themselves spontaneously so that the material can exhibit magnetization without being in a magnetic field are called ferromagnetic particles. Ferromagnetism is a so- called cooperative phenomenon, as single atoms cannot exhibit ferromagnetism, but once a certain number of atoms are bound together in solid form, ferromagnetic properties a rise. When the ferromagnetic particles are removed from the field, they exhibit permanent magnetization. Upon field reversal, the ferromagnetic material will initially oppose the field change, but eventually most domains will have switched their magnetization vectors and the same inverse magnetization is attained. Ferromagnetic materials, which are ground down to particle dimensions smaller than a particular domain, are no longer ferromagnetic but exhibit superparamagnetism. In case of paramagnetic particles, a magnetic field is altered by the magnetic materials present in it. If a particle contains magnetic moments that can be aligned in an external magnetic field, this will amplify the field. Such substances exhibit the property of paramagnetism. In contrast to ferromagnetic materials (ferromagnetism), no permanent magnetization remains in paramagnetic materials when they are removed from the magnetic field. Paramagnetism can be understood by postulating permanent atomic magnetic moments, which can be reoriented in an external field. These moments can be either due to orbiting electrons or due to atomic nuclei. The torque applied by an external magnetic field on these moments will tend to orientate them parallel to the field, which then reinforce it. 2.5 Bioconjugation with glutaraldehyde Glutaraldehyde is the most popular bis- aldehyde homobifunctional crosslinker in use today. 39

46 Glutaraldehyde has been used extensively as a homobifunctional crosslinking reagent, especially for antibody enzyme conjugations [55, 56]. However, a glance at glutaraldehyde s structure is not indicative of the complexity of its possible reaction mechanisms. Reactions with proteins and other amine- containing molecules would be expected to proceed through the formation of Schiff bases. Glutaraldehyde modification readily proceeds at alkaline ph. The higher the ph, the more efficient is Schiff base formation. Using a reductant like sodium cyanoborohydride that does not affect the aldehyde groups, while efficiently transforming Schiff bases into a secondary amines, provides the best possible yields. Glutaraldehyde in aqueous solutions can form polymers- containing points of unsaturation [57, 58, 59] (figure 2.9). Figure 2.9 Glutaraldehyde in aqueous solution may polymerize at either acid or basic ph. Such δ,δ- unsaturated glutaraldehyde polymers are highly reactive toward nucleophiles, especially primary amines. Reaction with a protein results in alkylation of available amines, forming stable secondary amine linkages. These glutaraldehyde- modified proteins may still react with other amine- containing molecules either through the Schiff base pathway or through addition at other points of unsaturation (figure 2.9). The proposed reaction mechanism of conjugation using these polymer conjugates explains the stability of proteins crosslinked by glutaraldehyde that have not been reduced. Schiff base formation alone would not yield stable crosslinked products without reduction. Crosslinking using glutaraldehyde polymers is difficult to reproduce. Since the glutaraldehyde polymer size and structure is unknown, the exact nature of the conjugates 40

47 formed by this method is unknown too. In many cases, the degree of glutaraldehyde- induced crosslinks is so severe that conjugate precipitation occurs. This is especially well documented in antibody enzyme conjugation schemes employing this reagent. The age of a glutaraldehyde solution is another variable, because the older the solution the more polymer will be formed. Fresh glutaraldehyde often will not yield the same results as aged solutions. To help overcome its tendency to form large- molecular- weight polymers upon crosslinking two proteins, a two- step protocol often is employed. Figure 2.10 Glutaraldehyde may react by several routes to form covalent crosslinks with amine-containing molecules. 2.6 Objectives The scope of current project was to optimize the surface- engineering procedures of superparamagnetic nanoparticles (MNPs). Monocrystalline iron oxide nanoparticles were synthesized and coated with macromolecules such as APTES or dopamine, in order to increase the colloidal stability, improve bio- interfacial properties. To achieve this aim, the project was performed in five stages below: i. Optimization of APTES- and Dopamine- coating procedures of iron oxide nanoparticles and their biofunctionalization with Insulin; ii. iii. Optimization of a peptidyl- DNA conjugate be attached to QDs; Characterization of uncoated-, APTES- and Dopamine- coated MNPs by transmission electron microscopy (TEM), scanning electron microscope (SEM), X- Ray Photoelectron Spectroscopy (XPS) and Fourier Transform Infrared (FTIR) and; 41

48 iv. Characterization of bio- interfacial properties of insulin- MNPs by Quarz Crystal Microbalance (QCM) for future applications in biosensing; v. Investigation of magnetic properties of MNPs by Magnetic Force Microscopy (MFM) to obtain in liquid a method to be used in receptor mapping of cells. References [1] A. Rishton, Y. Lu, R.A. Altman, A.C. Marley, C. Bian Hahnes, R. Viswanathan, G. Xiao, W.J. Gallagher, S.S.P. Parkin, Microelectron Eng., 1997, 35, 249. [2] A. B. Chin, I. I. Yaacob, J. Mater. Process. Technol., 2007, 191, 235. [3] C. Albornoz, S. E. Jacobo, J. Magn. Magn. Mater., 2006, 305, 12. [4] J. Wan, X. Chen, Z. Wang, X. Yang, Y. Qian, J. Cryst. Growth, 2005, 276, 571. [5] M. Kimata, D. Nakagawa, M. Hasegawa, Powder Technol., 2003, 132, 112. [6] G. S. Alvarez, M. Muhammed, A. A. Zagorodni, Chem. Eng. Sci., 2006, 61, [7] S. Basak, D.R. Chen, P. Biswas, Chem. Eng. Sci., 2007, 62, [8] I. Martinez- Mera, M. E. Espinosa, R. Perez- Hernandez, J. Arenas- Alatorre, Mater. Lett., 2007, 61, [9] S.A. Morisson, C. L. Cahill, E. Carpenter, S. Calvin, V. G. Harris, J. Nanosci. Nanotechnol., 2005, 5, [10] J. Qiu, R.Yang, M. Li, N. Jiang, Mater. Res. Bull., 2005, 40, [11] A.K. Gupta, A.S.G. Curtis, Biomaterials, 2004, 25, [12] A.K. Gupta, M. Gupta, Biomaterials, 2005, 26, [13] J.P. Jolivet, C. Chaneac, E. Tronc, Chem. Commun., 2004, 5, 481. [14] J. Zhang, R. Srivastava, R. Misra, Langmuir, 2007, 23, [15] M. Chen, S. Yamamuro, D. Farrell, S. A. Majetich, J. Appl. Phys., 2003, 93, [16] A.M. Morawski, P. M. Winter, K. C. Crowder, S. D. Caruthers, R. W. Fuhrlop, M. J. Scott, J. D. Robertson, D. R. Abendschein, G. M. Lanza, S. A. Wickline, Magn. Res. Med., 2004, 51, 480. [17] F. Tourinho, R. Franck, R. Massart, R. Perzynski, Coll. Polym. Sci., 1989, 79, 128. [18] J.F. Banfield, S.A. Welch, H. Zhang, T.T. Ebert, R.L. Penn, Science, 2000, 289, 751. [19] J.F. Banfield, R.L. Penn, Science, 1998, 281, 969. [20] C.E. Sjogren, K. Briley- Saebo, M. Hanson, C. Johansson, Magn. Res. Med., 1994, 31, 68. [21] E. Tronc, P. Belleville, J.P. Jolivet, J. Livage, Langmuir, 1992, 8,

49 [22] R. Massart, V. Cabuil, J. Chim. Phys., 1987, 84, 7. [23] S. Sun, C. B. Murray, D. Weller, L. Folks, A. Moser, Science, 2000, 287, [24] S. Sun, H. Zeng, J. Am. Chem. Soc., 2002, 124, [25] N.R. Jana, C. Earhart, J.Y. Ying, Chem. Mater., 2007, 19, [26] D.K. Nagesha, B.D. Plouffe, M. Phan, L.H. Lewis, S. Sridhar, S.K. Murthy, J. Appl. Phys., 2009, 105, 07B317. [27] C. Barrera, A.P. Herrera, C. Rinaldi, Journal of Colloid and Interface Science, 2009, 329, 107. [28] M. Butterworth, L. Illum, S. Davis, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2001, 179, 93. [29] Y. Zhang, N. Kohler, M. Zhang, Biomaterials, 2002, 23, [30] A. Jordan, R. Scholz, P. Wust, H. Schirra, T. Schiestel, H. Schmidt, R. Felix, Journal of Magnetism and Magnetic Materials, 1999, 194, 185. [31] M. Chastellain, A. Petri, A. Gupta, R. Rao, H. Hofmann, Advanced Engineering Materials, 2004, 6, 235. [32] Y. Jaemoon, L. Jaewon, K. Jinyoung, C. Chan- Hwa, L. Kwangyeol, S. Jin- Suck, Y. Ho- Geun, H. Yong- Min, H. Seungjoo, Nanotechnology, 2008, 19, 1. [33] P. Morais, J. Santos, L. Silveira, C. Gansau, N. Buske, Journal of Magnetism and Magnetic Materials, 2004, 272, [34] J. Roger, J. N. Pons, R. Massart, A. Halbreich, J. C. Bacri, The European Physical Journal of Applied Physics, 1999, 5, 321. [35] A. Schmidt, Journal of Magnetism and Magnetic Materials, 2005, 289, 5. [36] J. Cai, J. Guo, M. Ji, W. Yang, C. Wang, S. Fu, Colloid and Polymer Science, 2007, 285, [37] N. Shamim, L. Hong, K. Hidajat, M. S. Uddin, Colloids and Surfaces B: Biointerfaces, 2007, 55, 51. [38] S. Uchiyama, Y. Matsumura, A. de Silva, K. Iwai, Analytical Chemistry, 2004, 76, [39] T. Kawaguchi, T. Hanaichi, M. Hasegawa, S. Maruno, Journal of Materials Science: Materials in Medicine, 2001, 12, 121. [40] T. Kawaguchi, M. Hasegawa, Journal of Materials Science: Materials in Medicine, 2000, 11, 31. [41] S. Mornet, J. Portier, E. Duguet, Journal of Magnetism and Magnetic Materials, 2005, 293,

50 [42] S. Palmacci, L. Josephson, in United States Patent. vol. 5, 262, 176 U.S.A: Advanced Magnetics, Inc. (Cambridge, MA), [43] B. Feng, R. Hong, L. Wang, L. Guo, H. Li, J. Ding, Y. Zheng, D. Wei, Colloids and Surfaces A: Physicochemical and Engineering Aspects, 2008, 328, 52. [44] H. Liu, J. Li, Iranian Polymer Journal, 2008, 17, 721. [45] S. Mornet, S. Vasseur, F. Grasset, P. Veverka, G. Goglio, A. Demourgues, J. Portier, E. Pollert, E. Duguet, Progress in Solid State Chemistry, 2006, 34, 237. [46] M. Liong, M. Kovochich, T. Xia, S. Ruehm, A. Nel, F. Tamanoi, J. Zink, ACS Nano, 2008, 2, 889. [47] J. Young, P. Daksha, G. Ho, S. Woo, Y. Chang, Nanotechnology, 2008, 19, 1. [48] M. Koneracka, M. Muckova, V. Zavisova, N. Tomasovicova, P. Kopcansky, M. Timko, A. Jurikova, K. Csach, V. Kavecansky, G. Lancz, Journal of Physics: Condensed Matter, 2008, 20, [49] D. H. Kim, D. E. Nikles, D. T. Johnson, C. S. Brazel, Journal of Magnetism and Magnetic Materials, 2008, 320, [50] S. Wada, K. Tazawa, I. Furuta, H. Nagae, Oral Diseases, 2003, 9, 218. [51] J. A. Frank, H. Zywicke, B. Miller, A. Arbab, E. K. Jordan, B. K. Lewis, L. H. Bryant, J. W. M. Bulte, Proceeding of the International Society of Magnetic Resonance in Medicine, 2002, vol. 10. [52] C. De las Heras, S. Pennadam, C. Alexander, Chemical Society Reviews, 2005, 34, 276. [53] C. J. Xu, K. M. Xu, H. W. Gu, R. K. Zheng, H. Liu, X. X. Zhang, Z. H. Guo, B. Xu, J. Am. Chem. Soc., 2004, 126, [54] M.P. Morales, S. Veintemillas- Verdaguer, M.I. Montero, C.J. Serna, A. Roig, L. Casas, B. Martinez, F. Sandiumenge, Chem. Mater., 1999, 11, [55] S. Avrameas, Immunochemistry, 1969, 6, 43. [56] S. Avrameas, T. Ternynck, Immunochemistry, 1971, 8, [57] P.M. Hardy, A.C. Nicholls, H.N. Rydon, Chem. Comm., 1969, 65, 525. [58] P.M. Hardy, A.C. Nicholls, H.N. Rydon, J. Chem. Soc., Perk Trans, 1976, 1, 958. [59] P. Monsan, G. Puzo, H. Mazarguil, Biochimie, 1975, 57,

51 Chapter 3. Aqueous solution synthesis of magnetic nanoparticles In this section, colloidal chemical synthesis will be treated. Efforts have been made to synthesize different types of magnetic nanoparticles/nanocrystals using colloidal chemical synthetic approaches. Generally, in a colloidal system, separated nanoparticles with ultra- fine dimensions are stabilized by adding surfactant reagents resulting in a uniform suspension in an aqueous or organic solvent. Magnetic colloidal systems consist of magnetic nanoparticles/nanocrystals. In these colloidal systems the contributions from interparticle interactions are negligible and their unique magnetic behaviors are attributed to their reduced size. The surfactant coating plays an important role in colloidal systems to prevent magnetic nanoparticles clustering owing to steric repulsion. At the same time, assisted by surfactant molecules, reactive species can be added or removed during the synthesis of magnetic nanoparticles. In addition, these surfactant- attached magnetic nanoparticles can be suspended in various aqueous or organic solvents uniformly and powder forms can be obtained again by removing the solvent. For all technological and biomedical applications superparamagnetic nanoparticles with sizes smaller than 20 nm and a narrow size distribution are required. For this reason, the synthesis of magnetite nanoparticles with controlled size/shape and uniform physical/chemical properties is needed. Magnetite nanoparticles were functionalized with 3- aminopropyltriethoxysilane (APTES) and dopamine covalently bonded to the particle surface. The size and morphology of the magnetite nanoparticles was monitored during the various functionalization steps by TEM, SEM and AFM tecniques. 45

52 3.1 Materials and methods Materials Abbreviations: EtOH=ethanol; APTES= 3- aminopropyltriethoxysilane FeCl 2 * 4H 2 O (SIGMA- Aldrich), FeCl 3 * 6H 2 O (Sigma- Aldrich), NaOH (Fluka), HCl (Fluka), NH 4 OH (Fluka), tetramethyl ammonium hydroxide 25% w/v (Fluka), APTES (Sigma- Aldrich), ethanol (Fluka), Argon 5.0 Pa (Rivoira), Nitrogen 5.0 Pa (Rivoira), ddh 2 O (obtained by Millipore Simplicity 185). All materials were used as received. The samples were weighed by decimal balance (with accuracy of 10-4 g) Gibertini Crystal 100, concentred by Concentrator eppendorf 5301, centrifuged by Centrifuge Allegra X- 12R (Beckman Coulter), heated in water bath TRM 740 and on hot plate ARE Heating Magnetic Stirrer (CDL) e VELP Scientific, and ph was measured by phmeter PhMETER Basic Synthesis of magnetite- APTES nanoparticles in aqueous solution Synthesis of magnetite nanoparticles Magnetite nanoparticles were synthesized through co- precipitation method (Massart method modified) of an aqueous iron solution (Fe +3 :2Fe +2 ) using sodium hydroxide [1]. To the mixture consisting of 10 ml of 0.25 M FeCl 2 and 10 ml of 0.5 M FeCl 3 to ph 11 solution of 1.5 M NaOH was slowly added. The synthesis was carried out at room temperature for one hour under continuous stirring with bubbling argon to avoid oxidation of the ferrous ions. The solution was placed on the permanent magnet to obtain magnetic separation of the particles from the supernatant and washed 3 times with ddh 2 O to ph 7 to obtain a solution with colloidal particles suspension. Nanoparticles were peptized with tetramethyl ammonium hydroxide (1 M) and dried at 65 C using a vacuum oven. Synthesis of magnetite- APTES nanoparticles The ferrofluid suspension was coated with 3- aminopropyltriethoxysilane (APTES) [2]. The MNPs were dispersed in ethanol/ddh 2 O by sonication, APTES was added and sonicator was used again to homogenize; finally the sample was stirred overnight at 50 C. The APTES- bound magnetic nanoparticles were separated from the reaction mixture by placing the bottle on a permanent magnet, obtaining precipitation within 1-2 min. The supernatant was 46

53 removed and the precipitates were washed several times with ethanol. The obtained particles appear as in figure 3.1. Figure 3.1 Schematic representation of MNPs coated with APTES. 3.2 Characterization Fe, N, Si elemental analyses of the as- synthesized nanoparticle powders were performed. To prepare samples for elemental analysis, the particles were precipitated from their hexane dispersion by ethanol, centrifuged, rinsed in ethanol, and dried. Samples for transmission electron microscopy (TEM) analysis were prepared by drying a dispersion of the particles on amorphous carbon- coated copper grids. Particles were imaged using a Philips CM 12 TEM (120 kv). Sample for scanning electron microscopy (SEM) analysis were prepared by drying a dispersion of the particles on silicon wafer for several hours. Particles were imaged using a FIB-SEM Dual Beam (Fei Strata DB 235M). The ImageJ program (distributed by NIH) was used to measure the diameters of one hundred nanoparticles at the different functionalization steps, which were analyzed using a lognormal distribution. Magnetic studies were carried out using a MPMS Quantum Design SQUID magnetometer with fields up to 7 T and temperatures at 298 K. Infrared spectra of dried particles pressed into KBr pellets were obtained on a Varian 800 FTIR spectrometer. 47

54 3.3 Results and discussion Characterization of MNPs obtained by aqueous solution synthesis The size and morphology of the magnetite nanoparticles was monitored during the various functionalization steps by TEM and SEM techniques. The elemental analysis confirm the presence of iron. Figure 3.2 Magnetite (Fe3O4) nanoparticles synthesized by co- precipitation method. (a, b) TEM images; (c) SEM image; (d) Elemental Analysis. In the absence of any surface coating, magnetic iron oxide particles have hydrophobic surfaces with a large surface area to volume ratio. Due to hydrophobic interactions between the particles, these particles agglomerate and form large clusters, resulting in increased particle size. Since particles are attracted magnetically, in addition to the usual flocculation due to van der Waals force, surface modification is often indispensable. Strategy to avoid flocculation is to coat them with ligand molecules, such as APTES and dopamine. These agglomerates could be broken easily by sonication. 48

55 Figure 3.3 Hydrophobic and magnetic interaction between MNPs and their electrostatic and steric repulsion Magnetic properties NaOH NH4OH Figure 3.4 Hysteresis cycle of the magnetization for magnetic nanoparticles obtained by co- precipitation method, using as base NaOH or NH 4 OH. Magnetic measurements on all Fe 3 O 4 nanoparticles indicate that the particles are superparamagnetic at room temperature, meaning that the thermal energy can overcome the anisotropy energy barrier of a single particle, and the net magnetization of the particle assemblies in the absence of an external field is zero. At room temperature there is no 49

56 hysteresis. Under a large external field, the magnetization of the particles aligns with the field direction and reaches its saturation value (saturation magnetization). For Fe 3 O 4 nanoparticles, we noticed that the saturation magnetization was dependent on the size of the particles. Fe 3 O 4 nanoparticles in presence of oxygen reduced their saturation magnetization, suggesting the transformation of Fe 3 O 4 to Fe 2 O FTIR characterization Transmittanc Figure 3.5 IR spectra of APTES- MNPs synthesized at different temperature (30 C and 50 C). The IR spectrum of pure magnetite is included for comparison. To demonstrate the successful silane coating onto MNPs, the nanoparticles using FTIR were inspected (Figure 3.5). The strong IR band at 585 cm - 1 is characteristic of the Fe- O vibrations related to the iron core. After coating, a broadening and shift of the Fe- O band were observed, which are caused to the formation of Fe- O- Si bonds in the silane layer. Compared to the as- synthesized uncoated- MNPs, several new bands are observed in the spectra for silane- modified MNPs. The most striking difference is the appearance of bands between 1000 and 1150 cm - 1. These bands are characteristic of silane layers and originate from Si- O- Si vibrations. The alkane chain present in all silanes results in the appearance of typical bands at attributed to asymmetric CH 2 stretching. The position of these 50

57 CH 2 stretching vibrations is known to give more information about the crystalline packing of the alkane chains in the silane layer. The measured values can be compared to IR data for highly ordered and crystalline n- decanethiols on planar Au, which show absorptions at 2918 and 2851 cm - 1, respectively. The observed values ( cm - 1 ) imply that the silane layers deposited on the MNPs do not exhibit high crystallinity [3]. This low degree of crystallinity is most probably attributed to the presence of polymerization in the silane layer and the formation of silane multilayers. The amino- silane shows two N- H bendings at 1633 and 1570 cm - 1, characteristic of the presence of NH + 3 groups XPS characterization Although FTIR unambiguously confirms the successful silane coating, XPS measurements were also performed. An example of two survey spectra, before and after silane coating, is presented in figure 3.6. (ev) Figure 3.6 XPS spectra of APTES- MNPs synthesized at different temperature (30 C and 50 C). The XPS spectrum of pure magnetite is included for comparison. Inset show zoom of low energy area. After silane coating, the appearance of characteristic atoms such as sulfur and nitrogen could be observed. For all silanes, the total amount of Si was found to increase drastically. The increase in Si can be explained by the deposition of a Si- rich silane layer, whereas the decrease in Fe is due to the attenuation of the Fe2p photo electrons through the silane layer. 51

58 To further examine the chemical structure after silane coating, the high- resolution XPS spectra of Fe2p and O1s were deconvoluted. Figure 3.7 Deconvoluted spectra of the Fe2p subregion for unmodified and silane- modified magnetic nanoparticles at different temperature. Figure 3.8 Deconvoluted spectra of the O1s subregion for unmodified and silane- modified magnetic nanoparticles at different temperature. Apart from the presence of Si in the XPS spectrum, the successful silane coating also results in the appearance of two shifted peaks in the Fe2p and O1s spectra. For O1s spectrum silane- modified MNPs show peaks at ~ and ~ ev. These peaks are located at 52

59 approximately + 2 ev with respect to the principal Fe- O peak (~ 530 ev) and can be mainly assigned to Fe- O- Si and Si- O bonds, respectively. Shift of + 3 ev respect to Fe2p peak indicate iron oxidation state variation or/and bond variation (it is not loading effect ) TEM characterization This image shows that the aggregation is reduced, but the shape is not so regular. Most of the individual particles tended to form small agglomerates, which were suitable for further surface modification. 20 nm Figure 3.9 TEM image of APTES- MNPs. 3.4 Conclusion The Massart s method to prepare supermagnetic iron oxide nanoparticles (SPIO) has been modified. A strong alkaline solution (NaOH or NH 4 OH) was added to an aqueous mixture of ferric and ferrous salt. The resulting magnetite particles were black and exhibited a strong magnetic behaviour. During the reaction process, the ph was maintained at about 11. The ferrofluid suspension was washed several times by water and ethanol and dissolved in ethanol. The advantages of this method is facility of synthesis; the disadvantages, instead, are size dispersion, aggregation and crystallinity rather poor. For this reason, the nanoparticles obtained with this procedure are not monodispersed in size and their shape is not so regular as can be seen by SEM and TEM images. The silanazization reduced the aggregation, but the quality of MNPs obtained by this synthesis remains low. For this reason, the MNPs in organic phase were synthesized. 53

60 References [1] R. Massart, V. Cabuil, J. Chim. Phys., 1987, 84, 7. [2] N.R. Jana, C. Earhart, J.Y. Ying, Chem. Mater., 2007, 19, [3] M.D. Porter, T.D. Gright, D.L. Allara, C.D. Chidsey, J. Am. Chem. Soc., 1987, 109,

61 Chapter 4. Organic phase synthesis of magnetic nanoparticles Magnetic nanoparticles (Fe3O4) were synthesized with a narrow size distribution using the thermo- decomposition method. 4.1 Materials The synthesis was carried out using standard airless procedures and commercially available reagents. Absolute ethanol, hexane, dichloromethane (99%), benzyl ether (99%), 1,2- hexadecanediol (97%), oleic acid (90%), oleylamine (> 70%), iron(iii) acetylacetonate, APTES, Dopamine were purchased from Aldrich Chemical Co. All materials were used as received. The samples were weighed by decimal balance (with accuracy of 10-4 g) Gibertini Crystal 100, concentred by Concentrator eppendorf 5301, centrifuged by Centrifuge Allegra X- 12R (Beckman Coulter), heated in water bath TRM 740 and on hot plate ARE Heating Magnetic Stirrer (CDL) e VELP Scientific, and ph was measured by phmeter PhMETER Basic Preparation of MNPs by thermo- decomposition method The reaction of Fe- (acac)3 with surfactants at high temperature leads to monodisperse Fe3O4 nanoparticles, which can be easily isolated from reaction byproducts and the high boiling point ether solvent [1]. When phenyl ether was used as solvent, 4 nm Fe3O4 nanoparticles were separated, while the use of benzyl ether as solvent led to 6 nm Fe3O4. As the boiling point of benzyl ether (298 C) is higher than that of phenyl ether (259 C), the larger sized Fe3O4 particle obtained from benzyl ether solution seems to indicate that high reaction temperature will yield larger particles. However, regardless of the size of the 55

62 particles, the key to the success in making monodisperse nanoparticles is to heat the mixture to 200 C first and remain at that temperature for some time before it is heated to reflux at 265 C in phenyl ether or at ~ 300 C in benzyl ether. Directly heating the mixture to reflux from room temperature would result in Fe3O4 nanoparticles with wide size distribution from 4 to 15 nm, indicating that the nucleation of Fe3O4 and the growth of the nuclei under these reaction conditions is not a fast process. The low cost of Fe(acac)3 and the high yields it produces makes it an ideal precursor for Fe3O4 nanoparticle synthesis. Figure 4.1 Schematic illustration of organic phase synthesis. Several different alcohols and polyalcohols have been tested for their reactions with Fe(acac)3. It was found that 1,2- hydrocarbon diols, including 1,2- hexadecanediol and 1,2- dodecanediol, react well with Fe(acac)3 to yield Fe3O4 nanoparticles. Long- chain monoalcohols, such as stearyl alcohol and oleyl alcohol, can also be used, but particle quality is worse and product yield is poorer than those with diols in the synthesis of Fe3O4 nanoparticle seeds. However, in the seed- mediated growth process, these monoalcohols can be used to form larger Fe3O4 nanoparticles. Oleic acid and oleylamine are necessary for the formation of particles. Sole use of oleic acid during the reaction resulted in a viscous red- brown product that was difficult to purify and characterize. On the other hand, the use of oleylamine alone produced iron oxide nanoparticles in a much lower yield than the reaction in the presence of both oleic acid and oleylamine. When the 6 nm particles were oxidized by bubbling oxygen through the dispersion at room temperature, they precipitated from hexane as a red- brown powder. Adding more oleic acid did not cause re- dispersion of this powder into hexane. However, adding oleylamine alone, leading to an orange- brown hexane dispersion. This is consistent 56

63 with the observation that γ- Fe2O3 nanoparticles can be stabilized by alkylamine surfactants, suggesting that - NH2 coordinates with Fe(III) on the surface of the particles [1] Synthesis of 6 nm Fe 3 O 4 nanoparticles Fe(acac) 3 (2 mmol),1,2- hexadecanediol (10 mmol), oleic acid (6 mmol), oleylamine (6 mmol), and dibenzyl ether (20 ml) were mixed and magnetically stirred under a flow of nitrogen. The mixture was heated to 200 C for 2 h and then, under a blanket of nitrogen, heated to reflux (300 C) for another 1 h. The black- brown mixture was cooled to room temperature by removing the heat source. Under ambient conditions, ethanol (40 ml) was added to the mixture, and a black material was precipitated and separated via centrifugation. The black product was dissolved in hexane in the presence of oleic acid (200 ul) and oleylamine (200 ul). Centrifugation (6000 rpm, 10 min) was applied to remove any undispersed residue. The product, 6 nm Fe 3 O 4 nanoparticles, was then precipitated with ethanol, centrifuged (6000 rpm, 10 min) to remove the solvent, and redispersed into hexane [1]. Figure 4.2 Schematic illustration of organic phase synthesis. 4.3 Reaction of ligand exchange Modification of the synthesized magnetic- oleic acid nanoparticles was carried out via ligand exchange [2] with a functional amine- silane (APTES) molecule and dopamine. 57

64 To make MNPs dispersible in water, the hydrophobic ligands were exchanged by hydrophilic ones. This procedure involves adding an excess of ligand to the nanoparticle solution, which results in the displacement of the original ligand on the nanoparticles surface as can be seen in figure 4.3. The binding of oleates is noncovalent, while silanes are known to form a densely packed film, which is covalently linked to the surface. Figure 4.3 Schematic representation of silane ligand exchange Silane ligand exchange Procedure In a plastic container at 50 C, 1% (v/v) APTES solution in EtOH/H 2 O 1:1 was added to a dispersion of hydrophobic Fe 3 O 4 nanoparticles in hexane, previously sonicated for 5 min. The mixture was shaken for 5 h, during which the particles precipitated. The black- brown precipitate was separated using a permanent magnet. During the magnetic separation step, the particles were washed several times with hexane, toluene, to remove all excess silanes. The product was finally redispersed in ethanol/water Improving dispersion of MNPs- APTES In order to improve dispersion the reaction time was increased from 5 to 20 h, and this led to an increase in amine density at the particle surface and this effect is more pronounced at higher reaction temperatures. Also the reaction temperature was increased, which should promote silanol condensation in solution and onto the nanoparticles surface; however, thermal energy may destabilize possible silanol- protic solvent complexes and therefore promote more favorable silanol- surface interactions. Sonication of the particle suspension before commencing the silanization reaction is necessary because it led to a slight increase 58

65 in the final amine density values of the materials and may disrupt the formation of particle aggregates and/or cause the disaggregation of reversible aggregates already formed in the original suspension. The procedure experimental was carried out as follow: in a plastic container at 30 C, 0.5 % (v/v) APTES solution in EtOH/H 2 O 1:1 was added to a dispersion of hydrophobic Fe 3 O 4 nanoparticles in hexane, previously sonicated for 5 min. The mixture was shaken for 20 h, during which the particles precipitated. The black- brown precipitate was separated using a permanent magnet. During the magnetic separation step, the particles were washed several times with hexane, toluene, to remove all excess silanes. The product was finally redispersed in ethanol/water Synthesis of magnetite- dopamine nanoparticles Figure 4.4 Schematic illustration of ligand exchange from oleic acid to dopamine surfactant on Fe3O4 NPs. The ligand exchange procedure with dopamine proceeded easily and subsequent aggregation of the nanocrystals in water was not observed. Bidentate enediol ligands, such as dopamine, have been shown to be excellent capping agents for iron oxide nanostructures. Both oleic acid and dopamine are covalently bound to the surface via a chelating bidentate interaction to the iron species. Furhemore, a ligand exchange procedure was used to demonstrate that dopamine provides a better ligand than carboxylic acid. At room- temperature ligand exchange reaction using dopamine was used to convert the NPs from the hydrophobic to the hydrophilic state. This was carried out as follow: nanoparticles protected with oleic acid in hexane were mixed with an aqueous dopamine solution (10 mg). After two hours of strong agitation, the dopamine ligand displaced the oleic acid molecules as shown in figure 4.4 by transferring the nanoparticles to the aqueous phase. 59

66 4.4 Characterization The size and size distribution of the magnetic nanoparticles were determined at the different functionalization steps using a Philips CM 12 TEM (120 kv) transmission electron microscope (TEM), FIB- SEM Dual Beam (Fei Strata DB 235M) scanning electron microscope (SEM) and atomic force microscopy (AFM). AFM images were obtained with a Multimode- Nanoscope IIIa (Veeco Instruments, Santa Barbara, CA) in tapping mode using TESP silicon cantilevers (resonance frequency ~ 300 khz). Surface modification of the nanoparticles was confirmed by infrared spectra of dried particles pressed into KBr pellets and obtained on a Varian 800 FTIR spectrometer and by XPS technique. Fe, N, Si elemental analyses of the as- synthesized nanoparticle powders were performed TEM, SEM and AFM characterization The size and size distribution of the magnetic nanoparticles was monitored by TEM, SEM and AFM measurements from synthesis in presence of oleic acid until their surface modification with APTES and dopamine molecules. For this purpose, were prepared samples of magnetic nanoparticles coated with oleic acid and suspended in hexane, magnetic- APTES nanoparticles suspended in deionized water at ph 5.0, and magnetic- dopamine nanoparticles suspended in deionized water at ph 7.0. Subsequently, for TEM ultra thin carbon type A grids were immersed in these nanoparticle solutions placed on filter paper and then dried in a vacuum oven for 30 min; for SEM analysis the MNPs was deposited in silicon wafer and then dried in air for several hours; for AFM characterization, the sample a drop was deposited on a freshly cleaved mica surface. After 2 minutes the mica surface was washed with ddh2o and the sample dried under a N2 flux. The ImageJ program (distributed by NIH) was used to measure the diameters of one hundred nanoparticles, which were fitted to a lognormal distribution. No significant nanoparticle agglomeration was observed for magnetic nanoparticles in hexane and as can be seen in figure 4.4 the particles have good size distribution. Significant nanoparticle agglomeration was observed after the ligand exchange of oleic acid with the APTES molecules onto the nanoparticles surface, as observed in figure 4.5. The hydrolysis of APTES led to the formation of silanol (Si- OH) groups, which are condensed around the magnetic nanoparticle surface through siloxane (Si- O) bonds [3]. However, the 60

67 interactions between silanols groups leads these particles to agglomerate and form large clusters. No significant nanoparticle agglomeration was observed after the ligand exchange of oleic acid with the APTES molecules, after variation of reaction parameters or dopamine molecules onto the nanoparticles surface, as observed in figures 4.5, 4.6, 4.7. Figure 4.5 TEM micrograph and size distribution of the un- coated MNPs. The red line represents the best fit of the data according to a normal size distribution. Figure 4.6 SEM micrograph of large clusters of APTES- MNPs. 61

68 Figure 4.7 AFM micrograph of APTES- MNPs. As can be seen by these image was obtained a net improvement. Data scale: 20 nm. Figure 4.8 SEM micrograph of dopamine- MNPs. As can be seen by these image there is not significant agglomeration and the particles coated are well dispersed. 62

69 4.4.2 FTIR characterization Infrared spectroscopy was recorded to confirm the ligand exchange of oleic acid by APTES [3]. We observe characteristic vibrational bands at 2920 and 2850 cm - 1 for the synthesized magnetic nanoparticles, which were attributed to the antisymmetric and symmetric CH2 stretching present in the structure of the oleic acid. Furthermore, we identified a band at 1720 cm - 1 which was attributed to the antisymmetric C=O stretch of the carboxylic acid. After the ligand exchange, we observed for magnetic nanoparticles coated with APTES bands at about 1630, and 1550 cm - 1 characteristic of the NH2 bending mode of free amino groups and C- N stretching mode present in the APTES structure. Moreover, we observe the presence of bands at about 1100, 1000, and 920 cm - 1, which correspond to the condensation of siloxane (Si- O) molecules onto the surface of the magnetic nanoparticles. The presence of these bands and the absence of the bands attributed to the oleic acid indicate successful ligand exchange. Figure 4.9 IR spectra of APTES- MNPs in two different organic solvents Elemental analysis Elemental analysis verified that the nanoparticles were composed of Fe and Si. 63

70 Figure 4.10 Elemental analysis of APTES- MNPs XPS characterization To demonstrate the successful dopamine ligand exchange onto MNPs, the nanoparticles using XPS were inspected. Chloryne (200 ev), Carbon (284.7 ev), Nitrogen (401.5 ev), and Oxygen (532.5 ev) photoelectron peaks (in the order of binding energy from low to high) were observed as can be seen in figure 4.11 and the two peaks corresponding to binding energy of N1s and Cl2p confirm occurred linkage between dopamine and magnetic nanoparticles. Intensity Figure 4.11 XPS spectrum of dopamine- MNPs. 64

71 4.5 Conclusion Presence of NH 2 groups in APTES and dopamine allows the functionalization of magnetic nanoparticles with biomolecules. The particles obtained by these procedures are stable, not aggregates and uniform in size. References [1] S. Sun, H. Zeng, J. Am. Chem. Soc., 2002, 124, [2] N.R. Jana, C. Earhart, J.Y. Ying, Chem. Mater., 2007, 19, [3] R. De Palma, S. Peeters, M.J. Van Bael, H. Van den Rul, K. Bonroy, W. Laureyn, J. Mullens, G. Borghs, G. Maes, Chem. Mater., 2007, 19,

72 Chapter 5. Biofunctionalization of nanoparticles (MNPs and QDs) 5.1 Bioconjugation of insulin on MNPs Insulin is a polypeptide hormone formed, after elimination of C peptide by hydrolysis, of two chains of 21 and 30 amino acids, connected by two disulfide bridges and present several amino- groups. It is secreted by the ß cells of the islets of Langerhans of the pancreas and exerts an hypoglycemic action. It belongs to the group of peptides called IGF (insulin like growth factors) or somatomedins. Figure 5.1 Structure of insulin. Figure 5.2 Structure of glutaraldehyde. 66

73 For coupling APTES- MNPs at insulin an omobifunctional linker, glutaraldehyde, which presents two carbonyl groups at both ends, was used. The reaction is based on imino- group formation (C=N), a Schiff base, that contains a carbon nitrogen double bond. 5.2 Materials and methods Materials The synthesis was carried out using standard airless procedures and commercially available reagents. Absolute ethanol, hexane, dichloromethane (99%), benzyl ether (99%), 1,2- hexadecanediol (97%), oleic acid (90%), oleylamine (> 70%), iron(iii) acetylacetonate, APTES, sodium cyanoborohydride, glutaraldehyde 25% v/v and insulin were purchased from Aldrich Chemical Co. All materials were used as received. The samples were weighed by decimal balance (with accuracy of 10-4 g) Gibertini Crystal 100, concentred by Concentrator eppendorf 5301, centrifuged by Centrifuge Allegra X- 12R (Beckman Coulter), heated in water bath TRM 740 and on hot plate ARE Heating Magnetic Stirrer (CDL) e VELP Scientific, and ph was measured by phmeter PhMETER Basic Synthesis of bigger magnetic nanoparticles (seed- mediated growth) Bioconjugates formed by bigger APTES - MNPs and insulin were assembled. For subsesequent applications larger MNPs than 5 nm are necessary. To obtain them was used the seed mediated growth by means of which was increased their size by steps of 2 nm each. These are AFM images and again their height distribution. These MNPs are hydrophobic and I used as substrate for AFM HOPG back- side Synthesis of 8 nm Fe 3 O 4 nanoparticles via 6 nm Fe 3 O 4 seeds Fe(acac) 3 (2 mmol), 1,2- hexadecanediol (10 mmol), dibenzyl ether (20 ml), oleic acid (2 mmol), and oleylamine (2 mmol) were mixed and magnetically stirred under a flow of N 2. A sample of 6 nm Fe 3 O 4 nanoparticles dispersed in hexane (4 ml) was added. The mixture was first heated to 100 C for 30 min to remove hexane, then to 200 C for 1 h. Under a blanket of nitrogen, the mixture was further heated to reflux (300 C) for 30 min. The black- colored mixture was cooled to room temperature by removing the heat source. Under ambient conditions, ethanol (40 ml) was added to the mixture, and a black material was precipitated 67

74 and separated via centrifugation. The black product was dissolved in hexane in the presence of oleic acid (200 ul) and oleylamine (200 ul). Centrifugation (6000 rpm, 10 min) was applied to remove any undispersed residue. The product, 8 nm Fe 3 O 4 nanoparticles, was then precipitated with ethanol, centrifuged (6000 rpm, 10 min) to remove the solvent, and redispersed into hexane Silane ligand exchange In a plastic container at 30 C, 0.5 % (v/v) APTES solution in EtOH/H 2 O 1:1 was added to a dispersion of hydrophobic Fe 3 O 4 nanoparticles in hexane, previously sonicated for 5 min. The mixture was shaken for 12 h, during which the particles precipitated. The black- brown precipitate was separated using a permanent magnet. During the magnetic separation step, the particles were washed several times with hexane, toluene, to remove all excess silanes. The product was finally redispersed in ethanol using a diluition 1: Coupling of MNPs- APTES with glutaraldehyde Amine groups of APTES on MNPs is reacted with glutaraldehyde and purified away from excess reagent. Insulin then is added to effect the conjugate formation. We used the following protocol that describes a two- step coupling method using glutaraldehyde on amine particles based on the procedures of Kaplan et al. (1983) and Bang s Laboratories. FIRST STEP 1. Wash 10 mg of amine particles 3 times with 10 mm sodium phosphate, ph 7.4 (coupling buffer). 2. After the final wash, suspend the particles in coupling buffer containing 10 percent glu- taraldehyde and mix well to dissolve. 3. React while mixing for 1 hour at room temperature. 4. Wash the particles with coupling buffer at least several times using centrifugation to remove excess glutaraldehyde and separate by surnatant placing the bottle on the permanent magnet. Resuspend in 1 ml of the same buffer. 68

75 Figure 5.3 Glutaraldehyde conjugation to amine particles via two routes. Reaction of the aldehyde groups with the amines using sodium cyanoborohydride results in secondary (or tertiary) amine linkages with modifications containing a terminal aldehyde for further coupling to ligands. Alternatively, glutaraldehyde polymers can react with amine particles via addition to double bonds, resulting in a polymeric coating that contains both aldehydes and additional double bonds for further coupling with amine- containing molecules Bioconjugation of MNPs- APTES and insulin Amino groups on MNPs was reacted with the bis- aldehyde compound glutaraldehyde to form activated derivatives able to crosslink with other proteins, such as insulin. For this coupling we proceeding with procedures of Kaplan et al. (1983) and Bang s Laboratories. SECOND STEP 5. Add insulin to be coupled to the particle suspension in an amount equal to 1 10 molar excess over the calculated monolayer. Mix thoroughly to dissolve. Low concentrations of protein may result in particle aggregation, because a single protein molecule can react with more than one particle. 6. React with mixing for 2 4 hours. 7. Add to the particle suspension a final concentration of 0.2 M glycine (or another amine- containing quench molecule, such as ethanolamine or Tris) and 10 mm sodium cyanoborohydride. The blocking agent will couple to any remaining glutaraldehyde- reactive 69

76 sites and the reducing agent will convert all the resultant Schiff bases into stable secondary amine linkages. 8. Remove excess protein and reactants by washing with coupling buffer at least 3 times using centrifugation and separate by surnatant placing the bottle on the permanent magnet. Store particles in a suitable buffer containing a preservative. Figure 5.4 Bioconjugation between APTES- MNPs and insulin. The reaction proceeds in two steps: 1) The first carbonyl group of glutaraldehyde binds the amino groups of MNPs giving rise to the formation of a Schiff base. 2) The same reaction occurs with insulin, incubate with appropriate amount. Bond is reduced by sodium cyano- borohydride, because the bond is at equilibrium and can be easily detached. Schiff base interactions between aldehydes and amines typically are not stable enough to form irreversible linkages. These bonds may be reduced with sodium cyanoborohydride or a number of other suitable reductants to form permanent secondary amine bonds. 70

77 5.3 Characterization The size and size distribution of the magnetic nanoparticles were determined at the different functionalization steps using transmission electron microscope (TEM), scanning electron microscope (SEM) and atomic force microscopy (AFM). AFM images were obtained with a Multimode- Nanoscope IIIa (Veeco Instruments, Santa Barbara, CA) in tapping mode using TESP silicon cantilevers (resonance frequency ~ 300 khz). Surface modification of the nanoparticles was confirmed by QCM (Quarz Crystal Microbalance) AFM, TEM and SEM characterization For characterization, were prepared samples of magnetic nanoparticles coated with oleic acid and suspended in hexane, magnetic- APTES nanoparticles suspended in deionized water at ph 5.0, and insulin- magnetic nanoparticles suspended in deionized water at ph 7.0. Subsequently, for TEM ultra thin carbon type A grids were immersed in these nanoparticle solutions placed on filter paper and then dried in a vacuum oven for 30 min; for SEM analysis the MNPs was deposited in silicon wafer and then dried in air for several hours; for AFM characterization, the sample a drop was deposited on a freshly cleaved mica surface. After 2 minutes the mica surface was washed with ddh2o and the sample dried under a N2 flux. For MNPs- oleate HOPG back- side substrate was used because HOPG is hydrophobic such as MNPs- oleate. The ImageJ program (distributed by NIH) was used to measure the diameters of one hundred nanoparticles, which were fitted to a lognormal distribution. No significant nanoparticle agglomeration was observed for magnetic nanoparticles in hexane and, as it can be seen in figure 5.5, the particles have a good size distribution. No significant nanoparticle agglomeration was observed after the ligand exchange of oleic acid with the APTES molecules or insulin molecules onto the nanoparticles surface, as observed in figures 5.7, 5.8, 5.9. AFM characterization of un- coated MNPs obtained by seed mediated growth By seed- mediated growth MNPs with size about 12 nm were obtained and no significant nanoparticle agglomeration was observed for magnetic nanoparticles in hexane and, as it can be seen in figure 5.5, the particles have a good size distribution. 71

78 Figure 5.5 AFM micrograph of (a) HOPG substrate and (b) hydrophobic un- coated MNPs. Figure 5.6 Size distribution of the un- coated MNPs and 3D AFM image. TEM characterization of APTES- MNPs In this image we see that no significant nanoparticles agglomeration is present and that their shape is regular. Average size: ± 6.65 nm Figure 5.7 TEM micrograph of APTES- MNPs. 72

79 AFM and SEM characterization of bioconstruct formed by insulin and MNPs No significant nanoparticle agglomeration was observed after the bioconjugation of insulin molecules onto the nanoparticles surface and their size is about 50 nm. Figure 5.8 AFM micrograph of insulin- MNPs. Average size: 41.0 ± 21.2 nm Figure 5.9 SEM micrograph and size distribution of the Insulin- MNPs. The red line represents the best fit of the data according to a normal size distribution. No significant nanoparticle agglomeration was observed and their size is about 50 nm Magnetic characterization of bigger APTES- MNPs Magnetic measurements on bigger APTES- Fe 3 O 4 nanoparticles indicate that the particles maintaned superparamagnetic behaviour at room temperature 73

80 Figure 5.10 Hysteresis cycle of the magnetization for APTES- magnetic nanoparticles Quartz crystal microbalance (QCM) characterization In order to confirm bioconjugation between MNPs and insulin, QCM was utilized. QCM The quartz crystal microbalance (QCM) is an ultra- sensitive weighing device that utilizes the mechanical resonance of quartz. Quartz is one of the ionic crystalline solids that crystallizes in structures lacking a center of inversion. Therefore, a single crystal will possess a polar axis associated with the orientation of atoms in the crystalline lattice. As a consequence, if the crystal is set under mechanical pressure, an electric signal will be generated. This is known as the piezoelectric nature of quartz. Instead, if an external electric field was applied across the crystal this induce a mechanical strain in a material, where the direction of the induced strain can be controlled via the orientation of the cut in the crystal with respect to the crystal lattice. The QCM operates under thickness shear mode. For this particular case, the crystal, named AT- cut crystal, is cut at about 35 from the crystal axis. This type of cut ensures high temperature stability and pure shear motion when subjected to an electric field. In the most common configuration, a thin circular crystal disc is sandwiched between a pair of circular metal electrodes. 74

81 By applying an AC voltage, resonance is excited when the frequency of the applied voltage corresponds to the natural or resonance, f 0, of the crystal. This resonance condition occurs when the thickness of the disc is an odd integer number of half- wavelengths of the standing wave induced between the electrodes, causing the mechanical oscillation to have its anti- nodes at each electrode interface. The resonance frequency of the crystal is thus directly proportional to the total mass (thickness density) of the crystal. This can be written as: where f o is the resonance frequency, uq is the acoustic wave velocity in quartz, and tq is the quartz thickness. Provided that a mass is added to one or both electrode surfaces, Sauerbrey, demonstrated in 1959 that if the mass is (i) small compared to the weight of the crystal, (ii) rigidly adsorbed and (iii) evenly distributed over the active area of the crystal, a change in the total mass of the crystal, Δm, induced upon adsorption is, to a first approximation, linearly proportional to a change in frequency, Δf. Here, Δf is the change in frequency due to a change of mass equal to Δm, ρq is the quartz density, and μq is the elastic shear modulus for quartz. Although the QCM technique was originally used for gas phase and vacuum applications, but many work showed that QCM can also be used for liquid- phase applications, has paved the way for many applications within biotechnology, and in particular for various biosensor applications [1-5]. But, in liquid the adsorbed films do not obey the assumptions underlying the Sauerbrey relationship. In particular, the combined effect of hydration water, water trapped between adsorbed species, and the non- rigid character of many polymers/biomolecules, induces frictional (viscous) losses and thus a dampening of the crystal's oscillation. This, in turn, violates the linear relation between Δf and Δm, and calls for technical solutions to provide 75

82 information not only about changes in resonance frequency, but also about changes in energy dissipation, D, of the oscillating system. IgG anti- insulin antibodies were immobilized on gold surface [6] and MNPs- insulin were put on the construct for 30 min before washing Experimental procedure Measurement cell For QCM measurement in liquid a home- made cell has been used (figure 5.12). Figure 5.11 Schematic representation of gold- coated quartz. In this cell, the measurement chamber could be contain a determinate volume of buffer and two tube permit the insertion of required solutions. In measurement cells, the resonator has been put between two Silicon O- Ring that, from the one part do not permit movement into the cell, from the other part allows that only one gold electrode of resonator is in contact with the buffer solutions. The acquisition of frequency and dissipation values, has been performed via an impedance analyzer (IMPEDANCE ANALYSER agilent 4294A 40Hz- 110 MHz). The interfacing software has been created in Labview 8. QCM measurements were performed in in liquid with a homemade cell, using as transducer AT- cut 10 MHz resonators equipped with or Au (International Crystal Manufacturing Co, Inc) electrodes. The system was initially equilibrated at the desired temperature and ph then a syringe pump (PHD 2000, Harvard Apparatus, USA). 76

83 Figure 5.12 Home- made cell for QCM measurement in liquid. IgG construct Gold surface is cleaned in a piranha solution (H2SO4/H2O2 70%/30%) and exposed to plasma cleaner (Diener Femto) for 5 min just before use. For randomly oriented construct (see figure 5.13) measurement, Au electrode were exposed firstly at 2- Mercaptoethylamine (2- MEA Sigma Co.) solution at 5 mm concentration in water, and after at Glutaraldehyde solution 1% v/v (GD 25% v/v Sigma Co.) For oriented construct (used in MFM measureament; see figure 6.6) clean gold surface were incubate with protein A solution (Biovision) at 1 mg/ml concentration in water. After this step, IgG anti- insulin immobilization on exposed IgG binding domain layer was achieved by incubating on glutaraldehyde (random oriented construct) and on uniformly oriented protein A (oriented construct) for typically 30 min with mouse IgG anti- insulin Fc (DiaMed Belgium) at room temperature, followed by a rinse in a physiologic buffer [6]. Figure 5.13 Randomly oriented construct on gold surface formed by mercaptoethanolamine, glutaraldehyde and IgG anti-insulin. MNPs-insulin were conjugated with IgG anti-insulin and used for binding kinetic study. 77

84 5.3.5 Binding kinetics of MNPs and insulin Biomolecular binding events are detectable using labelled MNPs due to the increase of the hydrodynamic diameter in consequence of the formation of aggregates during the interaction. Whereas binding of single molecules to the particles generates rather small increases of the effective particle sizes, cross- linking of MNPs via coupling partners yields more pronounced effects. This offers the opportunity to determine kinetic parameters of the underlying processes, i.e. the affinity of binding partners such as antigen and antibody. In kinetic studies, the binding properties of antigen insulin on MNPs and their polyclonal antibody were investigated. In figure 5.14 we see frequency decreases which indicates the effective bioconjugation between insulin- MNPs and construct with antibody anti- insulin. Howewer, in control, consisting in APTES- MNPs, after washing in water the frequency value back to the initial and this indicate that MNPs are not linked on the construct (see figure 5.15). Figure Binding kinetics curve of insulin- MNPs on the construct. The frequency decreases and this indicate the effective bioconjugation. 78

85 Injection of MNPs-APTES Washing Figure 5.15 Binding kinetics curve of control (APTES- MNPs) on the construct. The fact that by washing in water the frequency value back to the initial indicate no bioconjugation between APTES- MNPs and antibody anti- insulin. These MNPs can be used as nanobiosensors with two types of possible applications (in future trend): 1) as gravimetric sensor because they can amplify the mass signal of small analytes which are too light to be detected and 2) as magnetic biosensors on cantilevers transducer which senses the presence of MNPs bonds to the analyte, the number of which is proportional to the concentration of analyte in the sample. 5.4 Bioconjugation of modified peptidic linker on QDs Metal- affinity coordination can drive the rapid self- assembly of various polyhistidine- appended proteins and peptides onto the metal- rich surface of QDs [7, 8]. This conjugation strategy is highly stable (low nanomolar binding constants), rapid and facile to implement, and is steadily being adopted by other research groups [7, 9-11]. Building on our understanding of polyhistidine coordination to the ZnS surface of QDs, we were able to extend the applicability of this strategy by synthesizing a reactive polyhistidine peptidyl linker that could be conjugated to modified DNA for subsequent self- assembly with QDs [12]. The linker consisted of a hexahistidine (His 6 ) sequence followed by a cysteine 79

86 residue that was modified with a pyridyldisulfide on its thiol group. The linker was conjugated to thiolated DNA by disulfide exchange and the versatility of this approach demonstrated by self- assembling a fluorescence resonance energy transfer (FRET)- based QD- DNA molecular beacon that was able to sense the presence of its cognate DNA complement. The liability of this approach is the disulfide bond itself as it is susceptible to both reduction and back- exchange with other free thiols. This precludes its use in reducing environments such as the cellular cytoplasm or conjointly in formats that include gold nanoparticles. Here, we report an improved chemistry for conjugation of peptidyl linkers to similar thiol- modified DNA. The His 6 functionality is retained unperturbed; however, the disulfide motif is replaced with a thiol- reactive iodoacetyl group leading to a more robust nonhydrolyzable peptide- DNA linker that still allows the same facile self- assembly approach for bioconjugation to QDs, as shown schematically in figure The synthesis, conjugation to modified DNA, self- assembly with QDs and characterization of the QD conjugates are described Materials Rink amide AM resin ( mesh), 9- fluor- enylmethoxycarbonyl (Fmoc)- His(trityl (Trt))- OH and benzotriazole- 1- yl- oxy- tris- pyrrolidino- phosphonium hexafluorophosphate (PyBOP) were obtained from Novabiochem (Darmstadt, Germany). Diisopropylethylamine (DIPEA), piperidine, iodoacetic acid, and N- hydroxysuccinimide were obtained from Aldrich (St. Louis, Missouri). Dichloromethane (DCM), dimethylformamide (DMF), triisopropylsilane (TIS), trifluoroacetic acid (TFA), and diisopropylcarbodiimide were obtained from Fluka (Buchs, Switzerland). The reagents for performing the Kaiser test were a generous gift from Dr. Claudio Paolucci of the Organic Chemistry Department A. Mangini at the University of Bologna. HPLC analysis and purification were performed on an Agilent 1100 system equipped with a PDA detector and employing a Zorbax SB- 300, 300 Å, 250 mm 9.4 mm C18 reverse phase column (Agilent Technologies, Palo Alto, CA). Elution of peptides was monitored at 220 nm. HPLC grade acetonitrile (MeCN) was obtained from Aldrich, while ddh2o (18.2 MΩ cm) was obtained from a Millipore Simplicity 185 system (Billerica, Massachusetts). The eluents were filtered through a 0.45 μm nylon membrane prior to use. For the purification of the iodoacetamide- derivatized peptides, the following HPLC gradient was employed at t = 0 min, B = 0; t = 30 min, B = 15%; flow = 2.5 ml/min, where A = ddh 2 O + 0.1% TFA, and B = MeCN + 0.1% TFA. 80

87 5.4.2 Quantum Dots CdSe/ZnS core/shell QDs with emission maxima centered at ~ 520 nm and 530 nm were synthesized using a stepwise reaction of organometallic precursors in a hot coordinating solvent mixture [13]. The QDs were made hydrophilic by exchanging the native organic capping shell with dihydrolipoic acid (DHLA) ligands as described elsewhere [13], and the quantum yield was determined to be ~ 20% Synthesis of succinimidyl ester - iodoacetamide The succinimidyl ester(osu) of iodoacetic acid (3) in figure 5.16, was prepared by mixing iodoacetic acid (1.86 g, 10 mmol) with N- hydroxysuccinimide (1.15 g, 10 mmol) in anhydrous ethyl acetate (50 ml). Diisopropylcarbodiimide (1.55 ml, 10 mmol) was slowly added and the mixture reacted for 4 h at room temperature. The precipitated urea by- product was removed by filtration and the resulting clear solution was first concentrated and then crystallized from anhydrous isopropyl alcohol to obtain (2.47 g, 8.7 mmol, 87%) of a white crystalline solid (m.p C) Synthesis of (His) 6 - NH 2 The hexahistidine backbone was synthesized as described in Ref. [12], see figure Briefly, Rink amide was pre- swelled in DCM for 2 h in a disposable polyethylene cartridge (Supelco, Bellefonte, Pennsylvania) equipped with a polyethylene frit and thoroughly rinsed with DCM (3 ) and with DMF (3 ). For Fmoc deprotection, piperidine (20% in DMF) was added to the resin and reacted for 1 h. The resin was washed with DMF (3 ), DCM (3 ) and methanol (3 ). Removal of the Fmoc protective group was confirmed by a positive Kaiser test. If the Kaiser test did not give a satisfactory result, deprotection and washing were repeated. For residue coupling, the resin was rinsed with DCM (5 ). Fmoc- protected histidine (2.5 eq.), PyBOP (2.5 eq.), and DIPEA (5 eq.) were added directly to the resin in the cartridge with the minimum amount of DCM required to ensure solubilization of the reactants and the mixture reacted for 2 h while shaking. After coupling, the resin was rinsed with DCM (3 ) and methanol (3 ). The outcome of the reaction was monitored by Kaiser test and the coupling and washing steps were repeated if needed. Coupling and deprotection were repeated until the sequence Rink- (His) 6 - NH- Fmoc (compound (1)) was obtained. After 81

88 the attachment of the terminal His residue, the Fmoc protective group was removed as usual and the resin washed as before to yield the free terminal primary amine OSu- iodoacetate coupling to Rink- (His) 6 - NH 2 An excess (> 10 eq.) of OSu- iodoacetate (3) was dissolved in 100 μl of dry DCM and added to 50 mg of the resin- bound peptide, and the mixture shaken for 2 h. The resin was then washed with DMF (3 ), DCM (3 ) and finally with methanol (3 ) to yield Compound (4) Peptide cleavage Figure 5.16 Synthesis of the thiol- reactive iodoacetyl His6 peptidyl linker. (a, b) Histidine residue coupling: Fmoc- (Trt)- His- OH 2.5 eq., PyBOP 2.5 eq., DIPEA 5 eq., in DCM for 2 h and repeated 6 times. (c) Fmoc deprotection of the primary amine: 20% piperidine in DMF, 0.5 h. (d) Modification of the primary amine with OSu- iodoacetate: >10 eq. of OSu- iodoacetate in DCM, 2 h. (e) Cleavage from the resin/trityl deprotection: TFA:TIS:H2O (95:2.5:2.5), 3 h. 82

89 Cleavage from the resin and trityl group deprotection was achieved by treatment with TFA:TIS:H 2 O (95:2.5:2.5) for 3 h. The solution was filtered through the cartridge frit and collected in Eppendorf tubes. The resin was then rinsed twice with TFA. The fractions collected were concentrated under reduced pressure until only a small volume was left. Cold (20 C) diethyl ether (5 volumes) was then added to precipitate the crude peptide which was centrifuged and the supernatant removed. The pellet obtained was rinsed with cold diethyl ether and re- suspended in 1:1 MeCN:ddH2O % TFA. The crude peptide was purified by HPLC to yield the His6- iodoacetamide derivative Compound (5). Mass spectral analysis confirmed the synthesis (ESI- MS: [M + ], [M 2+ ]) Conjugation of compound (5) to thiolated- DNA A synthetic oligonucleotide with the sequence 5 - [ThiSS] [TAMdT]- GAGCTCGTTCGTCTGAAGGTGAATGGCAG- 3 (where ThiSS is a thiol modifier with a six- carbon spacer, and TAMdT is a DNA T- base modified with tetramethyl- 6- carboxyrhodamine/tamra on a six- carbon spacer, with length 30 bp, Mw , and Tm 69.9 C) was obtained from Operon Biotechnologies (Huntsville, Alabama). The oligonucleotide was deprotected by reduction with 0.04 mol/l dithiothreitol (DTT) in 100 mmol/l KH 2 PO 4 buffer ph 8 at 37 C overnight and purified using two consecutive PD- 10 gel permeation columns (GE Healthcare, Piscataway, New Jersey). A two to three fold excess of Compound (5) was immediately added to the freshly deprotected oligonucleotide (see figure 5.17 and 5.18) and reacted for 2 h in KH 2 PO 4 buffer. The reaction was carried out at slightly alkaline ph where thiol- modification by the iodoacetyl is the exclusive reaction, thus negating non- specific reactivity towards amines or other groups [14]. At the end of the reaction the peptide- appended DNA was purified by PD- 10 gel chromatography, estimated quantitatively using the extinction coefficient of both the DNA and TAMRA dye, dried under vacuum and stored at 20 C until use Self- assembly of His 6 - peptide modified DNA with QDs To self- assemble QD His 6 - DNA conjugates or control samples at the desired valence, His- appended DNA at the appropriate molar ratios was added to 0.25 μmol/l of DHLA- capped QDs emitting at 520 nm in 10 mmol/l sodium tetraborate buffer ph 8 (100 μl in total) and allowed to incubate at room temperature for ~ 30 min. Samples for agarose gel analysis were prepared in the same manner but using lower concentrations and smaller volumes. 83

90 5.4.9 Fluorescence resonance energy transfer (FRET) analysis Fluorescence spectra from the QD- peptidyl conjugates were collected on a Tecan Safire Dual Monochromator Multifunction Microtiter Plate Reader (Tecan, Research Triangle Park, NC) using 300 nm excitation. FRET efficiency E was calculated from each sample set using the expression: E= (FD-FDA) / FD (3) where FD and FDA are the deconvoluted fluorescence intensities of the donor alone and the donor in the presence of acceptor(s), respectively [15]. QD- donor dye- acceptor separation distances r were estimated by fitting the FRET efficiencies E using Förster theory and the distribution of the number of acceptors per QD [16-18]. The quenching efficiency from a QD donor conjugated to exactly n acceptors is given by: E(n) = n / n + (r/r0)6 (4) Equation (4) is specific to a centrosymmetric QD donor- fluorophore acceptor conjugate where r is the center- to- center donor- acceptor distance and R0 is the Förster distance corresponding to the donor- acceptor separation that results in 50% energy efficiency [15, 16]. The FRET analysis also accounted for the contribution and any deviation due to Poisson distribution effects on low valence during the self- assembly process as described in Ref. [8, 16, 18]. 5.5 Discussion The schemes in figure 5.16 illustrate the synthetic pathway utilized for generating the iodoacetyl reactive peptide linker and in figure 5.17 and 5,18 its subsequent conjugation to thiolated- DNA. The core His6 sequence was obtained by standard solid phase peptide synthesis (SPPS) and a single terminal primary amine on the Rink solid phase- attached peptide was made available by deprotecting the Fmoc group. This amine was subsequently modified with the OSu- iodoacetate (3) and the peptide cleaved from the solid phase along with deprotection of the imidazole trityl- group by acid treatment to yield the final thiol- reactive peptide (5). Successful modification of the His6 sequence with Compound (3) and 84

91 its stability during the subsequent acid- cleavage/deprotection procedure demonstrates the robustness of this particular group. Further, the small size of the His6 sequence translates into the ability to generate relatively large amounts/ yields by SPPS as opposed to the lower yields that very often result when synthesizing and purifying longer peptide lengths. As with the previous version, the reactive peptidyl- linker could be stored dry for long periods of time until needed and also shipped over long distances. Indeed, samples of reactive linker were stored for more than one year and still remained viable. To demonstrate the general applicability of this improved peptide design, we obtained a TAMRA dye- labeled DNA oligonucleotide with a 5 protected- thiol group, very similar in sequence to that utilized previously [11]. Using standard procedures, the thiol group on the oligonucleotide was deprotected by reduction with DTT, reacted with a slight molar excess of the iodoacetyl- modified peptide (5) and the resulting peptidyl- DNA conjugate (6) purified using simple gravity- flow gel permeation chromatography on a disposable column as described above. Self- assembly of the peptide- DNA with the QDs was performed at room temperature in microcentrifuge tubes and peptide- DNA valence controlled through the molar ratio added relative to QD. Figure 5.17 Schematic representation of synthesis and self- assembly of the thiol- reactive iodoacetyl His6 peptidyl linker modified. 85

92 5.5.1 Gel electrophoresis characterization The self- assembly was first confirmed using agarose gel- electrophoresis; QDs were self- assembled with both the His construct and unmodified TAM DNA control and then separated in a 2% agarose gel. Figure 5.18 Schematic representation of synthesis and self- assembly of the thiol- reactive iodoacetyl His6 peptidyl linker modified Figure 5.19 Agarose gel electrophoresis of 530 nm QDs self- assembled with tetramethyl- 6- carboxyrhodamine (TAMRA) dye- labeled DNA, valence of >30 DNA / QD. 10 pmol of 530 nm DHLA- functionalized QDs were self- assembled with either His6- linker modified TAM DNA (His6TAM DNA) or unmodified TAM DNA and separated in 2% agarose gels supplemented with 1X TBE (0.089 mol/l Tris, mol/l borate, mol/l EDTA ph 8.3) in 1X TBE running buffer. The three resolved species indicated are: (1) His6TAM DNA, (2) 530 nm QDs, and (3) 530 nm QD- His6TAM DNA conjugates. 86

93 As can be seen in the two images, three distinct species are visible and become better resolved with longer separation times due to their differential mobility. Species 1 corresponds to the unmodified TAM DNA oligonucleotide and this has the highest relative migration rate, Species 2 is the 530 nm QDs where the strong negative charge of the DHLA ligand carboxy group promotes electrophoretic migration, and Species 3 is the QD- His6TAM DNA conjugate. The addition of multiple selfassembled DNA moieties to the QDs results, in this case, in retardation of the migration rate making the QD- His6TAM DNA conjugate the slowest moving species. The large molar excess of His6TAM DNA used during self- assembly also results in some unconjugated DNA present along with the QD- conjugates and this can also be seen in the gels. These results confirm that modification of the DNA with the thiol- reactive His6 peptide promotes direct self- assembly on the QD surface. In the absence of the peptide, the mutual negative charges, and thus the repulsion of the DHLA QDs and the DNA, serve to prevent both assembly and non- specific interactions AFM characterization Figure 5.20 AFM images of QDs exposed to peptide linker DNA (following hybridization to a longer dsdna amplicon). The 10 nm scale bar is shown in white. The bright spots are QDs with DNA seen as the lighter trails attached to the QD. 87

94 DNA- QD construct was self- assembled by mixing a 1:1 ratio of His6DNA and 590 nm QD in water. A 489 bp amplicon with a sticky- end complementary to the QD- immobilized oligo was hybridized onto the QD- Oligo construct to facilitate the imaging of what would be otherwise a very small DNA strand. The sample was then diluted to a concentration of 1 µg/ml in 10mM HEPES buffer, ph mm MgCl2 and a drop was deposited on a freshly cleaved mica surface. After 2 minutes the mica surface was washed with ddh2o and the sample dried under a N2 flux. AFM images were obtained with a Multimode- Nanoscope IIIa (Veeco Instruments, Santa Barbara, CA) in tapping mode using TESP silicon cantilevers (resonance frequency ~ 300 khz) FRET analysis Having verified self- assembly we proceeded to analyze the FRET interactions between 530 nm emitting QDs and the TAMRA dye on the proximal selfassembled DNA. Figure 5.21 A Schematic of the peptidyl- DNA conjugate as attached to a QD formed by the covalent linkage of the thiolated- DNA to the reactive iodoacetyl peptide which yields a thioether bond. Self- assembly of the peptidyl- DNA conjugate to the QD surface is mediated by the peptide terminal hexahistidine (His6) sequence through metal- affinity coordination and results in efficient FRET from the QD donor to the dye attached on the proximal DNA. The 530 nm QD hard radius is estimated to be in the range of ~ Å FRET measurements allowed us to test conjugate formation and to gain insight into the conformation of the His6- peptide- DNA self- assembled on the nanocrystal surface by 88

95 providing accurate estimates of the QD- dye donor- acceptor separation distance in these assemblies. Analysis of the QD- TAMRA FRET data gives a corrected average QD center- to- dye center r value of 49 Å and the His6 portion of the peptide is assumed to be directly attached to the QD surface by a minimum of at least 4 histidines. FRET measurements allowed us to test conjugate formation and to gain insight into the conformation of the His6- peptide- DNA self- assembled on the nanocrystal surface Figure 5.22 shows the absorption and emission spectra of QD donor - and TAMRA dye acceptor pair utilized in these experiments. In order to have FRET efficiency spectral overlap between donor s emission and acceptor s absorption is needed. The inset shows the QD- TAMRA excellent spectral overlap. Figure 5.22 FRET analysis. Absorption and emission of 520 nm QDs and tetramethyl- 6- carboxyrhodamine (TAMRA) dye. QD quantum yield (QY) is ~ 20%. TAMRA properties:λabs. max. 555 nm/λem. max. 580 nm, extinction coefficient ~ (mol/L) 1 cm 1, QY ~ 23%. The inset shows the QD- TAMRA spectral overlap. We monitored the FRET efficiency of the QD donor as an increasing ratio of His6TAM DNA was selfassembled onto the central nanocrystals. Figure 5.23 shows representative spectra 89

96 (deconvoluted and corrected for the direct acceptor excitation contribution) collected as the acceptor valence was increased in steps. In absence of construct we have the QD emission spectrum. As the concentration of construct is increased, the QD PL decreases (this behaviour is called quenching and is the proof of the bioconjugation). Since the energy is transferred to the acceptor by resonance the reemission of the acceptor increases. Figure 5.23 FRET analysis. Representative deconvoluted spectra collected from 520 nm QDs self- assembled with increasing ratios of His6TAM DNA conjugates. The QD spectra were fitted with a Gaussian profile. QD quenching and TAMRA reemission are indicated. The rate of the quenching phenomenon can be seen by figure 5.24 that shows the QD quenching efficiency along with TAMRA re- emission derived from the same QD- His6TAM DNA conjugates. QD quenching observed at the same valences in the presence of unmodified TAM DNA controls is also included in figure At just a nominal ratio of 2 DNA acceptors, the QD donor PL was quenched slightly more than 80% and quenching then effectively reached a plateau at ~ 90% for >(more then)4 acceptors/qd. The TAMRA re- emission remained unperturbed after reaching around 20% at a ratio of 2, where it continued to linearly track the QD quenching. From the similarity of the two curves we can 90

97 conclude that TAMRA is an effective quencher of QDs. The modest rate of TAMRA- sensitized re- emission can be attributed to a combination of the intrinsic properties of this dye along with contributions from self- quenching at higher valence. In contrast, exposing the same QDs to unmodified TAM DNA results in substantially less quenching and this result can be ascribed to solution- phase FRET interactions. Figure 5.24 FRET analysis. Plot of 520 nm QD quenching efficiency vs TAMRA ratio after self assembly with His6TAM DNA conjugates along with TAMRA re- emission and QD quenching after exposure to unmodified TAM DNA controls. Analysis of the QD- TAMRA FRET data gives a corrected average QD center- to- dye center r value of 49 Å which is in reasonable agreement with the previous conjugates, where an average separation distance of ~ 40 Å was obtained [12]; the differences can be accounted for by the smaller QDs used in the previous study along with differences in peptide structure and DNA sequence. The 530 nm QD hard radius is estimated to be in the range of ~ Å [19] and the His6 portion of the peptide is assumed to be directly attached to the QD surface by a minimum of at least 4 histidines [7], and thus it will only contribute to the overall lateral extension by <5 Å. The combination of the remaining 6- carbon linker that joins the thiol group to the DNA, along with the modified T- nucleotide and the extended diamino/6- carbon spacer linking the TAMRA dye to this T- base more than account for the remaining 14 91

98 15 Å of lateral peptide- DNA length even when not fully extended (energy minimized) or when accounting for some heterogeneity in conformation. Similar results were obtained when self- assembling the same peptide- DNA conjugates with polyethylene glycol (PEG)- modified QDs (data not shown) [20, 21] QDs- MB construct to discriminate between different sequences of DNA Confident in the ability of the His6- tagged- reactive peptide to specifically mediate DNA self- assembly onto CdSe- ZnS QDs, as verified by gel mobility and FRET data above, we set out to demonstrate a potential use for such conjugates by assembling a QD- MB (QD- molecular beacon conjugate) and testing its ability to specifically sense its target. Molecular beacons are commonly used to perform genetic analysis due to their essentially reagentless function. They can easily signal the presence or absence of target DNA/RNA sequence(s) in a reaction by monitoring changes in FRET interactions between two fluorophores attached at the ends of the MB sequence. For this, the MB was self- assembled onto DHLA- capped QDs. In nondenaturing conditions, the native oligo is designed to form a stem- loop structure, due to complementarity of the sequences at the 3 and 5 ends, which brings the dye close to the QD and initiates FRET interactions between the two fluorophores and produces QD PL loss. Addition of appropriate complementary target DNA, which hybridizes to the loop, opens the stem structure and results in a systematic change in the FRET efficiency that can be directly correlated with the target concentration. Figure 5.25 Schematic depiction of His6- peptide- linker facilitated self- assembly of labeled DNA onto QDs (left) and establishment of a hybrid molecular beacon structure (right). The hairpin DNA stem structure brings the dye- acceptor into close proximity r of the QD establishing efficient FRET. Addition of DNA complementary to the molecular beacon unwinds the stem- loop structure altering the donor- acceptor distance to r and changing FRET efficiency. DHLA- dihydrolipoic acid. 92

99 5.6 Conclusion Peptide- DNA conjugates are relatively well described in the literatures and have been extensively applied for nucleic acid, gene, and molecular beacon delivery to cells [22]. This has led to the development of several different covalent and noncovalent methods to accomplish the linkage or association of these two species. Disulfide bonds have been purposely introduced as the linking group between the peptide and the oligonucleotide in these hybrid species as intracellular reduction of the conjugate can facilitate the intracellular separation of both species. For example, Nitin and co- workers tested both reducible and non- reducible linkers when linking molecular beacons to cell delivery peptides for the purpose of monitoring mrna expression in living cells [23]. In our case, apart from some specific intracellular delivery applications, reduction of the disulfide linkage on a QD- DNA conjugate could be a liability in both in vivo and in vitro experiments as it would dissociate the QD- attached biomolecule or probe. We thus focused here on developing a modified chemistry that provides almost all the benefits of this polyhistidine- driven QD self- assembly bioconjugation approach but removed the reducible disulfide linkage. Use of a reactive iodoacetyl- group allowed targeting of the same thiols while shortening the peptide linker as compared to the previous by one residue in length. Discussion of the reducibility of disulfide bonds and the strength and long- term stability of iodoacetyl functional groups and thioether bonds can be found in the Refs. [14, 24]. It is worth noting that identical His6 sequences can be purchased commercially from vendors along with the OSu- iodoacetamide (3). Such commercially obtained His6- sequences would only contain one primary amine allowing it to be easily modified with the iodoacetamide to provide the same functional reactive linker. As thiol groups occur naturally in many proteins including antibodies, can be engineered into cloned proteins recombinantly, inserted into nascent peptides or added to DNA during synthesis, the chemistry demonstrated here can provide a simple method for self- assembling a variety of stable QD bioconjugates. Although a noncovalent conjugation strategy is used here, the ratio of His6 peptide and thus acceptor DNA acceptor per QD is also easily controlled through the molar ratio added for self- assembly with the QD as repeatedly demonstrated in Refs. [8, 9, 16, 18, 25-29] and discussed in detail in references [8, 16, 18]. The ability to switch the functional reactive group on the His6 peptide to those that target amines, for example, succinimidyl esters or isothiocyanates, or alternatively periodate 93

100 groups that target the sugar residues on carbohydrates can allow for multiple types of biomolecules to be targeted for self- assembly with the QDs [24]. Further, combining these types of linkers in one experiment can allow different orthogonally- targeted biomolecules to be self- assembled with the same QD to create multi- functional probes. Alternatively, switching the peptide sequences to those that recognize other metals or semiconductors can potentially allow self- assembly of similar biomolecules with a variety of other nanoparticle surfaces [30]. Lastly, it is also worth mentioning that several recent studies have suggested that it is possible to self- assemble polyhistidine modified proteins and peptides with commercial QD preparations [31, 32]. This can extend the choice to many different QD- biomolecule pairs and allow greater flexibility in incorporating them into more complex functional probes and nanostructures [33]. References [1] S. Bruckenstein, M. Shay Electrochemica Acta, 1985, 3(10), [2] K.K. Kanazawa, J.G. Gordon, Analytica Chimica Acta, 1985, 175, 99. [3] P.L. Konash, G.J. Bastians, Anal. Chem., 1980, 52, [4] Z.A. Talib, Z. Baba, S. Kurosawa, H.A.A. Sidek, A. Kassim, W.M.M. Yunus, American Journal of Applyed Science, 2006, 3(5), [5] T. Nomura, M. Okuhara, Analytica Chimica Acta, 1982, 1-12, 281. [6] M.L. Caiazzo et al., in preparation. [7] K.E. Sapsford, T. Pons, I.L. Medintz, S. Higashiya, F.M. Brunel, P.E. Dawson, H. Mattoussi, J. Phys. Chem. C, 2007, 111, [8] I.L. Medintz, A.R. Clapp, F.M. Brunel, T. Tiefenbrunn, H.T. Uyeda, E.L. Chang, J.R. Deschamps, P.E. Dawson, H. Mattoussi, Nature Materials, 2006, 5, 581. [9] E. Goldman, I.L. Medintz, J. Whitley, A. Hayhurst, A.R. Clapp, H. Uyeda, J.R. Deschamps, M. Lassman, H. Mattoussi, J. Am. Chem. Soc., 2005, 127, [10] B.I. Ipe, C.M. Niemeyer, Chem. Int. Ed., 2006, 45, 504. [11] W. Liu, M. Howarth, A.B. Greytak, Y. Zheng, D.G. Nocera, A.Y. Ting, M.G. Bawendi, J. Am. Chem. Soc., 2008, 130, [12] I.L. Medintz, L. Berti, T. Pons, A.F. Grimes, D.S. English, A. Alessandrini, P. Facci, NanoLett., 2007, 7,

101 [13] H. Mattoussi, J.M. Mauro, E.R. Goldman, G.P. Anderson, V.C. Sundar, F.V. Mikulec, M.G. Bawendi, J. Am. Chem. Soc., 2000, 122, [14] G.T. Hermanson, Bioconjugate Techniques, Academic Press: San Diego, [15] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer: New York, [16] T. Pons, I.L. Medintz, X. Wang, D.S. English, H. Mattoussi, J. Am. Chem. Soc., 2006, 128, [17] A.R. Clapp, I.L. Medintz, J.M. Mauro, B.R. Fisher, M.G. Bawendi, H. Mattoussi, J. Am. Chem. Soc., 2004, 126, 301. [18] T. Pons, H.T. Uyeda, I.L. Medintz, H. Mattoussi, J. Phys. Chem. B, 2006, 110, [19] B.O. Dabbousi, J. Rodriguez- Viejo, F.V. Mikulec, J.R. Heine, H. Mattoussi, R. Ober, K.F. Jensen, M.G. Bawendi, J. Phys. Chem. B, 1997, 101, [20] I.L. Medintz, A.R. Clapp, F.M. Brunel, T. Tiefenbrunn, H.T. Uyeda, E.L. Chang, J.R. Deschamps, P.E. Dawson, H. Mattoussi, Nat. Mater., 2006, 5, 581. [21] K. Susumu, H.T. Uyeda, I.L. Medintz, T. Pons, J.B. Delehanty, H. Mattoussi, J. Am. Chem. Soc., 2007, 129, [22] N. Venkatesan, B.H. Kim, Chem. Rev., 2006, 106, [23] N. Nitin, P.J. Santangelo, G. Kim, S.M. Nie, G. Bao, Nucl. Acids Res., 2004, 32, e57. [24] R.P. Haugland, The Handbook: A Guide to Fluorescent Probes and Labeling Technologies; Invitrogen: San Diego, [25] M.C. Daniel, D. Astruc, Chem. Rev., 2004, 104, 293. [26] I. Medintz, H. Uyeda, E. Goldman, H. Mattoussi, Nat. Mater., 2005, 4, 435. [27] K.E. Sapsford, T. Pons, I.L. Medintz, H. Mattoussi, Sensors, 2006, 6, 925. [28] K.E. Sapsford, C. Bradburne, J.B. Detehanty, I.L. Medintz, Mater. Today, 2008, 11, 38. [29] I.L. Medintz, A.R. Clapp, H. Mattoussi, E.R. Goldman, B. Fisher, J.M. Mauro, Nat. Mater., 2003, 2, 630. [30] M. Sarikaya, C. Tamerler, D.T. Schwartz, F.O. Baneyx, Ann. Rev. Mat. Res., 2004, 34, 373. [31] A.M. Dennis, G. Bao, Nano Lett., 2008, 8, [32] H. Yao, Y. Zhang, F. Xiao, Z. Xia, J. Rao, Angew. Chem. Int. Ed. 2007, 46, [33] I. Medintz, Nat. Mater., 2006, 5,

102 Chapter 6. Magnetic force microscopy (MFM) characterization 6.1 Atomic force microscopy (AFM) The atomic force microscope (AFM) [1] belongs to the broad family of scanning probe microscopes in which a proximal probe is exploited for investigating properties of surfaces with subnanometre resolution. The AFM, initially developed to overcome the limitations of its predecessors, the scanning tunnelling microscope (STM) [2], in imaging non- conducting samples, immediately attracted the attention of the biophysical community [3 10]. At the beginning the emphasis was mainly on the improved imaging resolution compared to that of optical microscopy, but, soon after, it became clear that AFM was much more than just a high- resolution microscope. The possibilities of spectroscopic analysis, surface modification and molecular manipulation gave rise to a real breakthrough in the realm of AFM use. In biological applications, the most appealing advantage of the AFM as a high- resolution microscope in comparison with other techniques such as SEM and TEM, is that it allows measurements of native biological samples in physiological- like conditions [11, 12], avoiding complex sample preparation procedures and artefacts connected to them. The use of mild imaging conditions opened the way to dynamic studies in which conformational changes and molecular interactions could be followed in real time at single- molecule level. The set of samples of biological interest studied by AFM ranges nowadays from the smallest biomolecules, such as phospholipids, proteins, DNA, RNA, to subcellular structures (e.g. membranes), all the way down to living cells and tissues. Not only structural properties can be investigated, but also mechanical or chemical and functional properties are the focus of many AFM applications. 96

103 6.2 Fundamental elements of the atomic force microscope Several excellent reviews can be found in the literature on issues related to instrumental aspects of the AFM [13 14] and we direct the reader to them for a more exhaustive review. In what follows we will just briefly recall the operating principles of an AFM [15]. Figure 6.1 Scheme of an AFM coupled with an inverted optical microscope How does the AFM work? The AFM works by scanning, in a raster fashion, a very tiny tip mounted at the end of a flexible microcantilever in gentle touch with the sample. This relative motion is performed with sub- Angstrom accuracy by a piezoelectric actuator (usually a tube, sometimes a tripod). Interacting with the sample the cantilever deflects and the tip sample interaction can be monitored with high resolution exploiting a laser beam impinging on the back of the cantilever [16]. The beam is reflected towards a split photodetector configuring an optical lever which amplifies cantilever deflections. In almost all operating modes, a feedback circuit, connected to the cantilever deflection sensor, keeps tip sample interaction at a fixed value controlling the tip sample distance. The amount of feedback signal, measured at each scanning point of a 2D matrix, concurs to form a 3D reconstruction of the sample topography which is usually displayed as an image. In figure 6.1 a scheme of an AFM is reported in a typical configuration for biological applications, where it is coupled with an 97

104 optical microscope to simultaneously acquire an optical image and the surface topography with the AFM Cantilevers and tips The scanning probe is the heart of the AFM, as for every scanning probe- based technique. The most common cantilevers used nowadays are realized by exploiting silicon micromachining technology [17 19]. In figure 6.2(a) a scanning electron microscope image of a silicon nitride microfabricated cantilever chip, usually used for contact AFM, is shown. At the end of a cantilever a tiny pyramid, the tip, is integrated (figure 6.2(b)). Depending on the imaging mode adopted, different types of cantilevers and tips may be used. When the AFM is operated in the static contact mode, the stiffness of the cantilever should be as low as possible (less than the interatomic spring constant of atoms in a solid), whereas in dynamic operation modes (see below) higher values for the spring constant help us reduce noise and instabilities. Figure 6.2(c) reports an image of a silicon cantilever and tip, usually used in dynamic operation modes, and a magnified view of the integrated silicon tip is shown in figure 6.2(d). Typical spring constants for AFM cantilevers range from 0.01 N m 1 to 100 N m 1, enabling a force sensitivity down to N. The limit in force sensitivity is related to an interplay between thermal, electrical and optical noise. The use of carbon nanotubes as AFM tips has represented a great breakthrough in terms of resolution [20 23]. Carbon nanotube tips possess a high aspect ratio, mechanical robustness, small diameter and a well- defined surface chemistry. They can be chemically functionalized, and appear to be the ideal probe for biological applications of the AFM requiring high resolution, particularly in the case of structural biology. The high- resolution imaging capabilities of carbon nanotube tips have been demonstrated on a variety of biological samples, such as DNA, IgG, IgM, GroES, SWI/SNF. The resolution attainable with carbon nanotube tips is comparable with that of other ultimate resolution imaging techniques such as cryogenic electron microscopy, but carbon nanotubes offer also the possibility to be functionalized, exposing, thus, a well- defined chemical group or chemisorbed biomolecule. This opportunity can be exploited to study the spatial distribution of chemical functional groups or complementary biomolecules in a sample. 98

105 Figure 6.2 Scanning electron microscope images of microfabricated AFM cantilevers and tips. (a) Silicon nitride cantilevers and (b) high magnification of the tip with oxide- sharpened apex; (c) silicon cantilever and (d) high magnification of the silicon tip. Courtesy of Veeco Instruments Imaging modes The classification of the possible operation modes is strictly related to the region of the force field between the tip and the sample spanned by the tip during imaging. Considering a non- linear force field composed of a repulsive short range force (<1 nm) and an attractive long range one (van der Waals force, <100 nm), depending on the force experienced by the tip cantilever ensemble, three working methods for the atomic force microscope can be defined. If the force between the tip and the sample is always repulsive and the tip is constantly in contact with the sample the microscope works in the contact mode. If the tip experiences only an attractive force with the sample and it never touches the sample, the usually called true- non- contact mode is being used. If the tip experiences both the attractive and the repulsive force with the sample, the intermittent- contact mode is used. The first imaging mode which has been developed is the contact mode, in which the tip is constantly in gentle touch with the sample surface [1]. In this imaging mode the applied force, hence cantilever deflection, is kept constant by the feedback system while the tip scans the surface. Images are created by recording the piezo- vertical position required to keep the force constant. This mode of operation can be used under aqueous environments allowing a reduction of the interaction force between the tip and the sample with respect to operation in air. The interaction force reduction comes from the removal of capillary forces 99

106 due to the presence of a thin water layer on surfaces in air [24]. This mode is the one of choice as far as flat and rather rigid samples are involved (e.g. reconstituted membranes, 2D protein crystals) enabling the highest resolution level. Sometimes, image artefacts, such as tilt of the surface or drift of the scanner perpendicular to the surface, dominate the topography image over real topographical features, especially when imaging is performed over large areas with slow scan rate. To avoid low- frequency artefacts the error mode can be used [25], in which the cantilever deflection signal is recorded keeping the feedback response time as fast as possible. The error signal gives a measure of how well the feedback system is maintaining the desired deflection setpoint. Due to the finite response time of the feedback loop, high- frequency signals cannot be completely compensated, and the cantilever will not be always at the same deflection value, especially in the case of high- frequency signals, generating a compensation error, which can be exploited to obtain an image rich in information. Moreover, tilt and vertical drift of the scanner generally have a low frequency and do not contribute to the error- mode image. When the tip is scanned in contact with the sample, lateral forces arise which cause a torsion of the cantilever. The torsion can be monitored by the signal from lateral segments of the photodiode [26]. Recording the torsion of the cantilever, surface distribution of different chemical functionalities, which result in different friction or adhesion properties with the tip, can be mapped. This operation mode is usually referred to as lateral force mode or, when chemically modified tips are used to measure differences between areas of distinct chemical properties, the technique is referred to as chemical force microscopy [27]. Figure 6.3 Schematic representation of an AFM tip operating in the intermittent contact mode. A drawback of the contact mode is the development of dragging forces associated with the lateral movement of the tip in contact with the sample. This problem is particularly evident 100

107 in the case of biological samples, which are usually loosely bound to the substrate and easily damageable. To overcome this problem another mode of operation, the intermittent contact mode, has been developed (figure 6.3) [28]. In this case the cantilever is oscillated at a frequency near its resonance and the oscillation amplitude is monitored. Starting from a free oscillation amplitude, when the cantilever approaches the sample and starts to hit its surface, the oscillation amplitude is damped. By recording the feedback signal required to keep the amplitude constant, the topography of the sample surface can be obtained. The tip being only intermittently in contact with the sample, the dragging forces during scanning are greatly reduced [29]. In the intermittent contact mode the tip goes through both the attractive and the repulsive regions of the tip sample force field during oscillation. The operation setpoint can be chosen so as to make attractive forces the dominant ones reducing damage to the sample and sometimes increasing resolution [30]. Technically, the cantilever can be oscillated by two methods [31]: acoustically or magnetically. In the first case the cantilever is oscillated by a piezo- actuator in contact with the cantilever supporting chip, whereas in the second case the cantilever is oscillated by means of an alternating magnetic field which acts on a magnetically susceptible film deposited on the backside of the cantilever [32]. Furthermore, important information on the viscoelastic properties of the sample can be retrieved by monitoring the phase difference between the cantilever driving signal and the output oscillation signal [33]. In particular, the phase shift is strictly related to the amount of energy dissipated in the tip sample contact and a mapping of the phase shift on a sample surface allows us to identify regions of different interaction properties [34, 35]. The development of intermittent contact modes of operation in liquid configured an important breakthrough for the application of the AFM in biology, especially in the case where single loosely immobilized molecules have to be imaged [36, 37]. Imaging at high resolution (about 1 nm) by intermittent contact mode AFM in liquid on biological samples has been reported [38], allowing us also to retrieve information about the interaction forces from the phase signal [39]. Dealing with the intermittent contact mode in liquid, a new technique which has been recently introduced for controlling in a better way the force applied by the tip is active quality factor control (active- Q) [40, 41]. This is a technique which allows us to increase the otherwise low quality factor for the oscillating cantilever in liquid and correspondingly to decrease the force applied by the tip on the sample. 101

108 6.3 Magnetic force microscopy In this study we investigate the ability of magnetic force microscopy (MFM), an atomic force microscopy (AFM)- based technique, to detect and localize superparamagnetic nanoparticles. MFM has been proven to be a useful technique to localize and characterize macroscale magnetic domains in materials and more recently for ferromagnetic nanoparticles [42, 43]. However, the capability of MFM to detect a signal from nanoscale para- or superparamagnetic particles has not been fully explored [44]. There have been some reports on the possible uses of MFM for detection of such particles occurring naturally in biological systems. These include detection of iron compounds in neurological disorders [45], magnetic domains in magnetotactic bacteria [46], and iron deposits in Hepatitis B- diseased livers [47]. Limited studies exist on localizing and detecting magnetic nanoparticles in vitro [48] or in cell- based systems [49]. A systematic study of the applicability of MFM for characterizing superparamagnetic nanoparticles (SP- MNPs) in air and in liquid is lacking and for this we report here the use of MFM to detect probe sample interactions from SP- MNPs in air and in liquid under ambient atmospheric conditions. By using both magnetic and nonmagnetic probes in dynamic lift- mode imaging and by controlling the direction of the external magnetic field applied to the tip we demonstrate that it is possible to detect and identify the presence of SPNs in the sample. A recent publication discussed the application of MFM to detection and localization of SP- MNPs [50], comparing the response of nonmagnetic nanoparticles with that from magnetic nanoparticles, and thanks to results presented, MFM images of mixed systems, (i.e. those showing magnetic nanoparticles in the presence of other features), can be correctly interpreted. MFM employs a magnetic probe, which is brought close to the sample and interacts with the magnetic fields near the surface. MFM detects local magnetic interaction by measuring deflections of the tip due to tip sample magnetic interaction as it scans across the sample. Unlike in AFM, the probe in MFM is magnetized. Thus, in addition to the forces measured in AFM, in MFM the magnetic interactions between the fields of the probe and those of the sample are measured. In order to be able to measure these interactions, the probe is lifted a certain distance from the sample surface, otherwise short- range forces, which can be much stronger than magnetic ones (such as van der Waals forces) will make measurement of the 102

109 234+' magnetic forces unfeasible. The most commonly used MFM implementation is the so- called lift mode [51] in which a two- pass technique is used each line in the image is measured M twice, the first line measuring only sample topography, the second line being used to Themeasure the magnetic fields at a fixed distance from the surface at each point, taking into bioconjugation was performed between MNPs-APTES and insulin by means of an account topography nformation from tthe he first scan. This is important because even at large omobifunctional linker,iglutaraldehyde. reaction is based on imino-group formation lift heights, long- range forces (other than (C=N) magnetic ones) might act on the probe. By!"#$ %&'(#!)*$+#,-.$&!$ &$ +/#-$-).!&(*#$012%$!"#$.310&*#$&!$ #&*"$ 42)(!5$!&6)('$ &**23(!$ maintaining the probe sample distance constant, it is assumed that such )(!2$ nonmagnetic 6. Receptor Mapping!242'1&4"7$ )(021%&!)2($ 012%$ +1.!$.*&(8$ 9").$ ).$ )%421!&(!$ #;#($ &!$,&1'#$,)0!$ forces will remain constant, and!"#$ thus only magnetic forces will be :#*&3.#$ measured [52]. "#)'"!.5$,2('<1&('#$ 021*#.$ =2!"#1$!"&($ %&'(#!)*$ 2(#.>$ %)'"!$ &*!$ 2($!"#$ 412:#8$?7$ For SP- MNPs there have been some puzzling results published, such as lift mode images of %&)(!&)()('$!"#$412:#@.&%4,#$-).!&(*#$*2(.!&(!5$)!$).$&..3%#-$!"&!$.3*"$(2(%&'(#!)*$021*#.$ The purpose of bioconjugation is to allow such MNPs to interact with surface cell re magnetite nanoparticles which show contrast that is not correlated with magnetic A),,$1#%&)($*2(.!&(!5$&(-$!"3.$2(,7$%&'(#!)*$021*#.$A),,$:#$%#&.31#-$BCCD8 properties of the sample, and which are described simply as fake MFM images, since the E21$FG<HIG.$!"#1#$"&;#$:##($.2%#$43JJ,)('$1#.3,!.$43:,)."#-5$.3*"$&.$,)0!$ %2-#$)%&'#.$20$ contrast seen in these images could not be properly explained in terms of magnetic %&'(#!)!#$(&(24&1!)*,#.$A")*"$."2A$*2(!1&.!$!"&!$).$(2!$*211#,&!#-$A)!"$%&'(#!)*$4124#1!)#.$ The advantages of this approach are: Higher resolution with respect to other techniques such as Fl Unfortunately, while phase shift derived of from lift mode images is commonly used to!"#.#$)%&'#.$*23,-$(2!$:#$4124#1,7$#/4,&)(#-$)($!#1%.$20$%&'(#!)*$)(!#1&*!)2(bckd8 site-specific magnetic manipulation Availability interaction [53]. 20$!"#$.&%4,#5$&(-$A")*"$&1#$-#.*1):#-$.)%4,7$&.$0&6#$HEH$)%&'#.5$.)(*#$!"#$*2(!1&.!$.##($)($ L(021!3(&!#,75$ A"),#$nanostructures 4"&.#$.")0!$ -#1);#-$ 012%$there,)0!$ %2-#$ )%&'#.$ ).$ a *2%%2(,7$ 3.#-$!2$ of characterize magnetic [45, 54], has never been systematic study *"&1&*!#1)J#$%&'(#!)*$(&(2.!13*!31#.$BCM5$CND5$!"#1#$"&.$(#;#1$:##($&$.7.!#%&!)*$.!3-7$20$!"#$ the response of this technique to the magnetization of the sample, and comparison of this 1#.42(.#$ 20$!").$!#*"()O3#$!2$!"#$ %&'(#!)J&!)2($ 20$!"#$.&%4,#5$ &(-$ *2%4&1).2($ 20$!").$ response with that obtained with nonmagnetic samples. For this reason, the interpretation 1#.42(.#$A)!"$!"&!$2:!&)(#-$A)!"$(2(%&'(#!)*$.&%4,#.8$E21$!").$1#&.2(5$!"#$)(!#141#!&!)2($20$ 2 Separazione del is contributo di topografia of the fake magnetic images referred to above difficult [50]. MAGNETIC FORCE MICROSCOPY!"#$P0&6#$%&'(#!)*$)%&'#.Q$1#0#11#-$!2$&:2;#$).$-)R*3,!$BSD8$ LIFT MODE Photodetector Magnet Contact IN LIQU Cantilever of the sam Cantilever During fi Cantilever image is mode; Lift height In second Sample Sample retracted t mode) i Piezo topogra Piezo separation is due to l Figure 6.4 MFM!"!#$%&'(#$(%)&%*+#,-&#+./0+#&)#,1(#+/*(#/%(/2#34%506#7%+,#+./0#,&$&6%/$18#5*/6(#5+#&',/50(9# probe performs two scans of the same area. During first scan topography image is obtained in!"#$%&'()*t$ such as ma contact mode; In second scan cantilever is elevated to a certain height (lift mode) in order to have topography Photodetector 7. Preliminary Results 50#.&0,/.,#*&9(:#;0#+(.&09#+./0#./0,5<(=(%#5+#(<(=/,(9#,&#/#.(%,/50#1(561,#><5),#*&9(?#50#&%9(%#,&#1/=(#,&$&6%/$18# contribution separation. Cantilever deflection is due to long- range interactions such as magnetic interactions..&0,%5'4,5&0#+($/%/,5&02#@/0,5<(=(%#9(a(.,5&0#5+#94(#,&#<&06b%/06(#50,(%/.,5&0+#+4.1#/+#*/60(,5.#50,(%/.,5& U($!").$A216$ Control:&$.!3-7$20$!"#$1#.42(.#$20$ HEH$!2$ %&'(#!)*$&(-$ (2(%&'(#!)*$(&(24&1!)*,#.V &1#&$).$1#421!#-8$ hard-disk MNPs-APTES

110 In this work a study of the response of MFM to magnetic and nonmagnetic nanoparticles/area is reported. 6.4 Materials and methods AFM and MFM characterization All magnetic force microscopy was carried out with a commercial AFM system (Veeco Multimode with Nanoscope IIIa controller) under ambient conditions. All the results reported in this work were carried out with non- conductive silicon nitride cantilevers for contact mode and TESP sharp silicon cantilevers for tapping mode (resonance frequency ~ 300 khz). All cantilevers were coated with a chromium/cobalt film; Cr (5 nm/3.5 nm, tungsten filament); Co (30 nm/13 nm, tungsten boat) evaporated in vacuum chamber through thermal evaporation. The probe was magnetized by a permanent magnetic holder. Images were collected in lift mode controlled by the software, and values of phase shift extracted from the images. All samples were deposited from solution and thoroughly dried before analysis. Magnetic nanoparticles were exposed to ultrasounds before deposition to reduce clustering. In some experiments, we examined the same nanoparticles with different probes, to see the response without magnetized probes, or to repeat the experiment to determine the variability between different probes. In order to do this, it was necessary to navigate back to exactly the same position on the sample, which is not a trivial task. Average size: 50.2 ± 13.6 nm Figure 6.5 AFM image and size distribution of the Insulin- MNPs. The red line represents the best fit of a normal size distribution to the data. 104

111 In this work APTES- MNPs, Insulin- MNPs, IgG constructs obtained as described in Chapter 5, were used. These MNPs after were deposited on the gold surface treated with protein A and IgG anti- Insulin (Caiazzo et al, in preparation) and used for MFM characterization. Figure 6.6 Construct on gold surface uniformly oriented formed by Protein A and IgG anti- insulin. MNPs- insulin are conjugated with IgG anti- insulin and used in liquid MFM. 6.5 Preliminary results in MFM Tests of the magnetized tip Figure 6.7 AFM and MFM images in lift mode of hard disk. 105

112 One of the most crucial parts for the image formation process in a MFM is the magnetization of tip. Artefacts because of the usually unknown magnetic state of the tip, its unknown behaviour in the sample's stray field and its influence on the sample magnetization may lead to image perturbations and misinterpretations. An image of a hard disk acquired in MFM mode is shown in figure LIFT mode in liquid on MNPs- APTES Magnetic traces in lift mode images are well visible (D Agostino et al., in preparation) as can be seen in figure 6.8. An image taken with a magnetic tip contains information about both the topography and the magnetic properties of a surface. Which effect dominates depends upon the distance of the tip from the surface, because the magnetic force persists for greater tip- to- sample separations than the van der Waals force. If the tip is close to the surface, in the region where standard non- contact AFM is operated, the image will be predominantly topographic. As one increases the separation between the tip and the sample, magnetic effects become apparent. We collect a series of images at different tip heights to separate magnetic from topographic effects and we obtain an image in liquid where magnetic trace in lift mode image are well visible and this is a good result for future application in receptor mapping of cells. Figure 6.8 LIFT mode of MNPs- APTES in liquid. Magnetic trace are well visible as indicated by arrow 106

113 6.5.3 LIFT mode with reversed field of tip The magnetic behaviour of MNPs- insulin deposited on construct on gold surfaces were checked. We confirm black magnetic traces using reversed magnetic field of tip. After inversion, we obtain white traces and this confirm that black traces of first image are magnetic. Figure 6.9 LIFT mode of MNPs- insulin in liquid. Magnetic trace are well visible (upper) and were confirmed by reversed magnetic field of tip (lower) Larger MNPs embedded in PMMA (poly(methyl methacrylate)) in order to ensure the independence from topography In order to ensure the independence from topography we worked with MNPs embedded in PMMA. If traces in lift mode are present, they will necessary magnetic because there is not topographic contribution. 107

114 Sample preparation MNPs are added in PMMA solution and spinned with a home- made spinner. We obtained a sample where MNPs are inside PMMA layer. Figure 6.10 Schematic diagram of magnetic nanoparticles embedded in PMMA. Figure 6.11 TEM images of magnetic nanoparticles embedded in PMMA. The particles are uniformly distributed in all surface and this confirm the trend of magnetic traces obtained with MFM. MFM characterization Figure 6.12 MFM images of magnetic nanoparticles embedded in PMMA. In areas where there is not topography contribution however are present magnetic traces as indicated by arrows. 108

115 We operate AFM in air, using tip magnetically coated with Cr:Co 5:30 nm and in contact mode. Magnetic traces are well visible despite the MNPs are embedded in PMMA and this confirms the magnetic character of the spot Comparison of magnetic and not magnetic area to access the true magnetic character of the imaged features In this work a study of the response of MFM to magnetic and nonmagnetic nanoparticles/area is reported. In order to clearly distinguish between magnetic and nonmagnetic NPs, we analyzed mixed systems, consisting of depositions of colloidal solutions of un- coated MNPs in specific area, with alterning magnetic and not magnetic areas. Also in this case, only where MNPs are present we can see magnetic traces as can be seen in figure 6.8 and 6.9. In the oscillating lift mode used here, the contrast is generated by the change in effective spring constant of the cantilever caused by the interaction between the fields above the sample with the field from the probe. This causes a shift in the frequency of oscillation which is most commonly detected by the effect it has on the phase of oscillation [55, 56]. Other authors have observed a negative phase shift over small magnetic domains (usually represented in MFM images by a darker shade), due to attractive interactions between the probe and the magnetic domain [57]. In addition, when the probe- magnetic domain distance is small, rather than being a unidirectional effect, the magnetically induced phase shift can have both positive (at the edges) and negative (in the centre) regions over a single magnetic domain [58]. Figure 6.13 (a) shows an image obtained at 15 nm lift height. It is clear that both not magnetic and magnetic areas with MNPs can be imaged by lift mode MFM under these conditions. Comparison of parts A and B shows that the magnetite particles gave a far stronger response than the not magnetic areas. Furthermore, it can be seen that the not magnetic area gave an almost entirely positive phase shift with no appreciable negative phase shift. On the other hand, the large cluster of magnetite particles showed large negative shifts. Various sample preparations were examined, and the same sample examined with different probes. Overall, the results were similar, but we found that the magnitude of the response varied greatly from one probe to another. Despite using probes from the same batch, and magnetizing them using the same technique their responses varied enormously. Presumably, manufacturing variability leads to probes that have considerably different magnetic properties, as well as variations in probe geometry. 109

116 Unfortunately, measurement of the magnetization of MFM probes in order to take into account this variability is not simple [57]. In this and most reported works, this parameter is unknown and one of the great limiting factors for quantitative MFM [50]. In order to standardize the response of the probes, we normalized the response on a particular magnetite particle, at 15 nm lift height (the lift height showing the greatest response), assuming that the difference found here was simply due to magnetization of the probe. The results of this analysis, reporting the response with and without applied field are shown in figure 6.7 and 6.8. Figure 6.13 MFM phase shift images at lift heights of 15 nm showing the same area. (a) The first image is topography of the sample, the second and third images are phase shift of the sample in presence and in absence, respectively, of magnetized tip. (b, c) Line profiles through phase and height images of alterned magnetic areas. 110

117 Figure 6.14 MFM phase shift images at lift heights of 15 nm showing the same area. (a) The first image is topography of the sample, the second and third images are phase shift of the sample in presence and in absence, respectively, of magnetized tip. (b, c) Line profiles through phase and height images of alterned magnetic areas. The third image represent a case of fake image [53] because the week contrast present is not correlated with magnetic properties of the sample (as confirmed by line profile) [53]. These values were obtained by extracting the phase information measured over all area. The curves in figure 6.7 and 6.8 clearly show that the response from the not magnetic areas gave a positive phase shift, while that from the magnetite particles was negative. Based on the results reported here, we believe that the positive phase shift response seen on nonmagnetic nanoparticles, was due to a repulsive electrostatic interaction between equally charged probe and surfaces [59]. On the other hand, since the probe could magnetize the samples, magnetic nanoparticles exhibited a mostly attractive interaction with the probe, exhibiting a negative phase shift over all the particles. 111

118 6.5.6 Conclusion In conclusion, we have confirmed that MFM is sensitive to the magnetic fields of magnetic nanoparticles. We showed a direct comparison of the response of lift mode MFM to nonmagnetic areas with that to magnetic areas with magnetite. A response that might be mistaken for a magnetic interaction can be detected from nonmagnetic nanoparticles of different materials and sizes. The nonmagnetic interactions gave positive phase shifts, indicative of repulsive interactions with the MFM probe. Magnetic interactions were characterized by negative phase shifts over the particles. This difference in phase shift response can be used to characterize the nature of the interactions with the sample, and thus the nature of the sample itself. A previous work has compared the response of nonmagnetic nanoparticles with that from magnetic nanoparticles, it is only with the new results presented here, that MFM images of mixed systems, (i.e. those showing magnetic nanoparticles in the presence of other features), can be correctly interpreted. Systems such as those studied here do not require an external magnetic field in order to be studied greatly simplifying the experimental setup. Therefore, MFM is capable of discriminating between magnetic and nonmagnetic nanoparticles, but researchers who wish to show the magnetic response from their nanoparticles using MFM must ensure they show negative phase shifts attractive interactions, and not only positive ones. Considering these points, magnetic particles can be easily distinguished from nonmagnetic particles, and thin surface coatings make no measurable difference to their response. References [1] G. Binnig, C.F. Quate, C. Gerber, Phys. Rev. Lett., 1986, 56, 930. [2] G. Binnig, H. Rohrer, E. Gerber Cand Weibel, Phys. Rev. Lett., 1982, 57. [3] O. Marti, V. Elings, M. Haugan, C.E. Bracker, J. Schneir, B. Drake, S.A. Gould, J. Gurley, L. Hellemans, K. Shaw, J. Microsc., 1988, 803. [4] A. Engel, Ann. Rev. Biophys. Chem., 1991, [5] M. Radmacher, R.W. Tillmann, M. Fritz, H.E. Gaub, Science, 1992, 257, [6] S.M. Lindsay, Biophys. J., 1994, 67, [7] C. Bustamante, D. Keller, Phys. Today, 1995, December, 32. [8] Z. Shao, J. Yang, Q. Rev. Biophys., 1995, [9] Z. Shao, J. Yang, A.P. Somlyo, Ann. Rev. Cell. Dev. Biol., 1995,

119 [10] Z. Shao, J. Mou, D.M. Czajkowsky, J. Yang, J.- Y. Yuan, Adv. Phys., 1996, 45, 1. [11] B. Drake, C.B. Prater, A.L. Weisenhorn, S.A. Gould, T.R. Albrecht, C.F. Quate, D.S. Cannell, H.G. Hansma, P.K. Hansma, Science, 1989, 243, [12] C. Bustamante, C. Rivetti, D.J. Keller, Curr. Opin. Struct. Biol., 1997, 7, 709. [13] D. Sarid, Scanning Force Microscopy 2nd edn., Oxford University Press, New York, [14] R.J. Colton, A. Engel, J.E. Frommer, H. Gaub, A.A. Gewirth, A. Guckenberger, J. Rabe, W.M. Heckl, B. Parkinson, Procedures in Scanning Probe Microscopy, Wiley, New York, [15] A. Alessandrini, P. Facci, Meas. Sci. Technol., 2005, 16, R65. [16] G. Meyer, N.M. Amer, Appl. Phys. Lett., 1988, 53, [17] T.R. Albrecht, S. Akamine, T.E. Carver, C.F. Quate, J. Vac. Sci. Technol. A, 1990, 8, [18] S. Akamine, R.C. Barrett, C.F. Quate, Appl. Phys. Lett., 1990, 57, 316. [19] O. Wolter, T. Bayer, J. Greschner, J. Vac. Sci. Technol. A, 1991, 9, [20] H. Dai, J.H. Hafner, C.M. Lieber, Nature, 1996, 384, 147. [21] J.H. Hafner, C.- L. Cheung, A.T. Woolley, C.M. Lieber, Prog. Biophys. Mol. Biol., 2001, 77, [22] S.S. Wong, J.D. Harper, P.T. Lansbury, C.M. Lieber, J. Am. Chem. Soc., 1998, 120, 603. [23] S.S. Wong, E. Joselevich, A.T. Woolley, C.- L. Cheung, C.M. Liber, Nature, 1998, 394, 52. [24] D.A. Grigg, P.E. Russell, J.E. Griffith, J. Vac. Sci. Technol. A, 1992, [25] C.A. Putman, K.O. van der Werf, B.G. de Grooth, N.F. van Hulst, J. Greve, P.K. Hansma, Scanning Probe Microsc., 1992, 1693, 198. [26] C.M. Mate, M. McClelland, R. Erlandsson, S. Chiang, Phys. Rev. Lett., 1987, 59, [27] C.D. Frisbie, L.F. Rozsnyai, A. Noy, M.S. Wrighton, C.M. Lieber, Science, 1994, 265, [28] Q. Zhong, D. Inniss, V.B. Kjoller Kand Elings, Surf. Sci. Lett., 1993, 290, L688. [29] J. Tamayo, R. Garcia, Langmuir, 1996, 12, [30] A. San Paulo, R. Garcia, Biophys. J., 2000, 78, [31] I. Revenko, R. Proksch, J. Appl. Phys., 2000, 87, 526. [32] W. Han, S.M. Lindsay, T. Jing, Appl. Phys. Lett., 1996, 69, [33] S.N. Magonov, V. Elings, M.- H. Whangbo, Surf. Sci., 1997, 375, L385. [34] J.P. Cleveland, B. Anczykowski, A.E. Schmid, V.B. Elings, Appl. Phys. Lett., 1998, 72, [35] J. Tamayo, R. Garcia, Appl. Phys. Lett., 1997, 71, [36] P.K. Hansma et al., Appl. Phys. Lett., 1994, 64,

120 [37] C.A.J. Putman, K.O. van der Werf, B.G. de Grooth, N.F. van Hulst, J. Greve, Appl. Phys. Lett., 1994, 64, [38] C. Möller, M. Allen, V. Elings, A. Engel, D.J. Müller, Biophys. J., 1999, 77, [39] M. Stark, c. Möller, D.J. Müller, R. Guckenberger, Biophys. J., 2001, 80, [40] J. Tamayo, A.D. Humphris, M.J. Land Miles, Appl. Phys. Lett., 2000, 77, 582. [41] J. Tamayo, A.D.L. Humphris, R.J. Owen, M.J. Miles, Biophys. J., 2001, 81, 526. [42] D.G. Ramlan, S.J. May, J.G. Zheng, J.E. Allen, B.W. Wessels, L.J. Lauhon, NanoLett., 2006, 6, 50. [43] Y. Takamura, R.V. Chopdekar, A. Scholl, A. Doran, J.A. Liddle, B. Harteneck, Y. Suzuki, NanoLett., 2006, 6, [44] M. Rasa, A.P. Philipse, J. Magn. Magn. Mater., 2002, 252, 101. [45] S. Schreiber, Small, 2008, 4(2), 270. [46] J. Dobson, FEBS Lett., 2001, 496, 1. [47] C.E. Diebel, R. Proksch, C.R. Green, P. Neilson, M.M. Walker, Nature, 2000, 406, 299. [48] A.L. Martinelli, A.B. Filho, R.F. Franco, M.H. Tavella, L.N. Ramalho, S. Zucoloto, S.S. Rodrigues, M.A. Zago, J. Gastroenterol. Hepatol., 2004, 19, [49] Y. Amemiya, T. Tanaka, B. Yoza, T. Matsunaga, J. Biotechnol., 2005, 120, 308. [50] M. Rasa, B.W. Kuipers, A.P. Philipse, J. Colloid Interface Sci., 2002, 250, 303. [51] R. Giles et al., Appl. Phys. Lett., 1993, 63, 617. [52] A. Schwarz, R. Wiesendanger, Nano Today, 2008, 3(1-2), 28. [53] M. Rasa, A.P. Philipse J. Magn. Magn. Mater., 2002, 252, 101. [54] J. Pacifico et al., Nanotechnology, 2009, 20 (9), [55] U. Hartmann, Annu. Rev. Mater. Sci., 1999, 29, 53. [56] B. Letal, J. Appl. Phys. Lett., 1997, 70, [57] Y. Amemiya et al., J. Biotechnol., 2005, 120(3), 308. [58] S. Foss et al., Appl. Phys. Lett., 1996, 69, [59] C. S. Neves et al., Nanotechnology, 2010, 21,

121 Chapter 7. Conclusion Magnetic nanoparticles have been synthesized using aqueous solution (Massart method modified) and organic phase synthesis (thermo- decomposition method). The advantages of the first method is facility of synthesis; the disadvantages, instead, are size dispersion, aggregation and crystallinity rather poor. For this reason, the MNPs obtained with this procedure are not monodispersed in size and their shape is not very regular. Through thermo- decomposition method we obtained nanoparticles capped with nonpolar endgroups (oleic acid) on their surface, approximately 1o nm in diameter, spherical shape and narrow size distribution, but insoluble in water. This can be an inconvenient when the particles have to be functionalized with biomolecules, for use in a biosensor or receptor mapping. To render MNPs dispersible in water, the hydrophobic ligands were exchanged for hydrophilic ones. Ligand exchange is a well- known method for improving the surface properties of nanoparticles; toward that aim 3- aminopropyltriethoxysilane (APTES) and dopamine were used as ligands and reaction parameters preventing aggregation were established. MNPs were biofunctionalized with insulin and characterized by QCM and MFM. With MFM we obtained preliminary results to discriminate MNPs and not magnetic nanoparticles in liquid, data confirmed by reversing tip magnetization. Furthermore, we obtained images of magnetic traces independently of topography using a system formed by MNPs embedded in PMMA. The purpose of bioconjugation is to allow such MNPs to interact with surface cell receptors enabling MFM measurements. The advantages of this approach are 1) higher resolution with respect to other techniques such as Fluorescence detection; 2) availability of site- specific magnetic manipulation. MFM imaging could also be adopted to study protein dynamics in live cells, since AFM was previously used to study the surface topography of living cells in solution (see figure 7.1). This offers a potential advantage of AFM/MFM imaging over electron microscopy, which delivers a comparable resolution but cannot be used in live cells. 115

122 The ability to positively identify magnetic nanoparticles with this technique is expected to be extremely useful for present and future application of MNPs in biological systems. QDs also were biofunctionalized with thiol- reactive iodoacetyl hexahistidine peptidic linker attached to thiolated- DNA oligomers (QDs) for subsequent applications. The reactive haloacetyl chemistry demonstrated here results in a stable thioether bond linking the DNA to the peptide which can withstand strongly reducing environments such as the intracellular cytoplasm. Future applications of this construct could be in highly luminescent multilabeled hybridation probes tests or on cultured cells QDs- MB construct to discriminate between different sequences of DNA. Figure 7.1 Melanoma cells observed with atomic force microscopy (AFM) in physiological environment. Application of MNPs in cancer cells to obtain surface cell receptor mapping. 116