I. 1. What are nanoparticles?

Nanoscience and nanotechnology are recent, revolutionary development in Science and Engineering that are evolving at a very fast pace. It is driven by the desire to fabricate materials with novel and improved properties that is likely to impact virtually all areas of the physical and chemical sciences, biological sciences and health sciences. The word nano derives from the Greek word nanos, which means dwarf or extremely small and mathematically, a nanometer is a billionth of a meter or 10-9 m. Particles whose sizes fall in the size range of 1-100 nm is called a nanoparticle, whether it is dispersed in gaseous, liquid or solid medium. Nanoparticles are a number of atoms or molecules bonded together (these particles usually contain 106 atoms or less) and intermediate in size between individual atoms and aggregates large enough to be called bulk material.1,2 The dimensionality of different systems is shown in Figure 1 As the nanoparticles are larger than individual atoms and molecules but are smaller than bulk solid, those obey neither absolute quantum chemistry nor the laws of classical physics and therefore, materials in the nanometer size regime show the behavior which is intermediate between that of a macroscopic solid and that of an atom or molecular system. There are two major phenomena those are responsible for these differences.

Figure 1. Dimension of different objects in the length scale.

First, as the size of a particle is reduced, the number of atoms at the surface of the particle increases compared to their bulk i. e., surface-to-volume ratio increases. A cluster with a radius of 1 nm contains ~25 atoms, but most of them are on the surface of the cluster. Therefore, properties, which are usually determined by the molecular structure of the bulk lattice, now become increasingly dominated by the structure of the surface. Second, as the size of material becomes equal to or falls below the nanometer length scale that characterizes the motion of its electrons and thus its properties, become sensitive to the size of the particles. The effect specific to the small size of nanoparticles is called the quantum size effect3 (QSE) that opens up the opportunity to create materials and devices with entirely new properties.

I. 2. Historical background Although the extraction of gold started in the 5th millennium B.C. in Bulgaria, soluble gold appeared around the 5th century B.C. in Egypt and China.4 Until the Middle Ages, the soluble gold was used to disclose fabulous curative powers for various diseases, such as heart and venereal problems, dysentery, epilepsy, and tumors, and for diagnosis of syphilis. The optical properties of gold sols have attracted interest ever since the Romans began to use them as colorants in glasses, which were known as Purple of Cassius. But the scientific research on gold sol was started by the revolutionary work of Michael Faraday.5 In 1857, Faraday reported the formation of deep red solutions of colloidal gold by reduction of an aqueous solution of chloroaurate (AuCl4-) using phosphorus in CS2 (a two-phase system). He investigated the optical properties of thin films prepared from dried colloidal solutions and observed reversible color changes (from bluish-purple to green) of the films upon mechanical compression. The term colloid (from the French, colle) was coined shortly thereafter by Graham, in 1861. In 1908, G. Mie6 first came forward to explain the red color of the gold nanoparticles in solution. In fact, it was the color variation of colloidal gold with size that motivated Mie to apply the general theory of light extinction to small particles.

Mies idea of 1908 about the study of the optical properties of gold clusters with sizes below the resolution limit of light microscopes took about 50 years for its realization. His idea has led to the development of new field of cluster science and has become the subject of immense investigation as the time progresses.

I. 3. Some special aspects of nanoparticles

I. 3. 1. Electronic structure

The electronic properties of nanostructures of metals and insulators (and semiconductors) have been widely investigated as they originate interesting transport and optical phenomena in these materials. Bulk metals posses a partially filled electronic band and their ability to conduct electrons is due to the availability of a continuum of energy levels above the Fermi energy, EF. However, when bulk metal is broken down, the continuum of electronic states breaks down and ultimately, the material becomes insulating.7,8 In a bulk semiconductor, the EF is in between the filled and unfilled levels. Thus, excitation induces the formation of an electron and hole, which are delocalized over a number of atoms or ions constituting the material. Due to the delocalization, along with the high dielectric constant of the material, their binding energy is relatively small. Such separated electron and hole are bound by weak Columbic attraction which is also termed as Wannier exciton. The distance between the hole and the electron is known as Bohr radius and is on the nanometer length scale. The minimum amount of energy required for the formation of these charge carriers is the band gap energy of the semiconductor. It is the energy difference between the top of the valence band and the bottom of the conduction band. When the size of the semiconductor material becomes smaller than Bohr radius, it would decrease the space in which the charge carriers move and thus confine their motion. The situation is equivalent to the motion of an electron in a box.

Figure 2. Schematic representation of the density of states of a semiconductor nanocrystal as compared to a molecule or bulk material.

The solution to Schrdinger equation of an electron in a box of length L gives the Eigen function:
( x ) = 1 n sin( kx ), k n = 2L L (1)

(1)

whose corresponding energy Eigen values are given by

En = 2 2 n 2 2mL2

(2)

where, m is the effective mass. As the size of the box is reduced, the energy level spacing increases since it is inversely proportional to L2. The separation energy between the valence band and conduction band increases upon confinement. Nanocrystals exhibiting quantum confinement are called quantum dots. Figure 2 shows how the electronic structures of semiconductor nanocrystals differ from those of bulk and isolated molecules. 5

I. 3. 2. Surface plasmon resonance

Metal nanoparticles exhibit a strong absorption band in UV-visible region, which can be assigned to the surface plasmon resonance of the metal clusters. When a small spherical metallic nanoparticle is irradiated by light, the oscillating electric field causes the conduction electrons to oscillate coherently. This is schematically presented in Figure 3. When the electron cloud is displaced relative to the nuclei, a restoring force arises from Coulomb attraction between electrons and nuclei that results in oscillation of the electron cloud relative to the nuclear framework. The oscillation frequency is determined by four factors: the density of electrons, the effective electron mass, and the shape and size of the charge distribution.

Figure 3. Schematic representation of plasmon oscillation for a sphere, showing the displacement of the conduction electron charge cloud relative to the nuclei as a result of electromagnetic field as shown by Kelly et al. J. Phys. Chem. B 2003, 107, 668.

The collective oscillation of the electrons is called the dipole plasmon resonance of the particle. Higher modes of plasmon excitation can occur, such as the quadrupole mode where half of the electron cloud moves parallel to the applied field and half moves antiparallel. For a metal like silver, the plasmon frequency is also influenced by other electrons such as those are in d-orbital, and this prevents the plasmon frequency from being easily calculated using electronic structure calculations. However, it is not hard 6

to relate the plasmon frequency to the metal dielectric constant, which is a property that can be measured as a function of wavelength for bulk metal.9 The frequency and width of the surface plasmon absorption depend on the size and shape of the metal nanoparticle as well as on the dielectric constant of the metal itself and of the medium surrounding it.10-12 Figure 4 shows the surface plasmon band of gold nanoparticles of different size. The plasmon resonance is strongest and shifted into the visible part of the electromagnetic spectrum for the noble metals [copper, silver, and gold],13 which is the reason why the noble metals have historically fascinated scientists dating back as early as Faraday.5 Most other transition metals only show a broad and poorly resolved absorption band in the ultraviolet. 13 This difference can be attributed to the strong coupling between the plasmon transition and the interband excitation and to the fact that the conduction-band electrons of the noble metals can be well approximated by the Drude free-electron model.14 This model assumes that the conduction-band electrons can be treated independently from the ionic background and can move freely, whereas the ions only act as scattering centers. This in turn gives the electrons in the noble metals a higher polarizability, thereby shifting the plasmon resonance to lower frequencies and also giving rise to a sharp bandwidth.

Figure 4. UV-visible absorption spectra of 9, 22, 48, and 99 nm gold nanoparticles in water as reported by El-Sayed et al., J. Phys. Chem. B 1999, 103, 4212.

I. 3. 3. The Mie theory

The linear optical properties like extinction and scattering of small spherical metal particles accounting for the surface plasmon resonance were first time explained theoretically by the groundbreaking work of Mie in 1908.6 Mie solved Maxwells equation for an electromagnetic light wave interacting with a small sphere having the same macroscopic, frequency-dependent material dielectric constant as the bulk metal. With the appropriate boundary conditions for a spherical object, his electrodynamic calculations gave a series of multipole oscillations (dipole, quadrupole, etc.) for the extinction and scattering cross section of the particles as a function of the particle radius. For large nanoparticles (>20 nm), the absorption spectrum is then composed of the sum of absorption and scattering modes, each of which has a contribution that depends on the particle size. Higher-order modes become more dominant with increasing particle size causing the plasmon absorption band to red shift while the bandwidth increases. Physically, this can be explained by the fact that for larger particles the light cannot polarize the nanoparticles homogeneously and retardation effects lead to the excitation of higher-order modes.1 As the optical absorption spectra depend directly on the size of the nanoparticles, this is regarded as an extrinsic size effect.1 In this size regime, Mies theory and experimental spectra agree well until for bulk metals the normal incidence absorption no longer shows a plasmon resonance. For nanoparticles much smaller than the wavelength (<20 nm) of the interacting light only the dipole oscillation contributes significantly to the extinction cross section. Mies theory then reduces to the following expression (dipole approximation)1,10-12: (3) Where V is the particle volume, is the angular frequency of the exciting light, and c is the speed of light.
m

and ( ) = 1( ) + 2( ) are the dielectric functions of

the surrounding medium and the metal, respectively. For the metal, the dielectric

ext
8

9. V. m3/2 = c

function is complex and depends on the frequency. The resonance condition is fulfilled roughly when 1( ) = -2
m,

if

is small or weakly dependent on .

In general, Mies theory is only valid for very low concentrations of the nanoparticles in a solvent or solid matrix. It is assumed that the individual particles are non-interacting and separated far enough that the electric field created around one particle by the excitation of a surface plasmon resonance is not felt by other surrounding particles. If the interparticle distances become smaller than the spherical particle dimensions, or even when aggregation occurs, the plasmon resonance red shifts, and often a second absorption peak at longer wavelength is observed.1,15-17 This second band could be regarded as a longitudinal resonance similar to the nanorods in case of chain-like aggregation of the individual nanodots. In the case of particle aggregation, and for composite materials like densely packed nanoparticles in a transparent host medium or nanoparticle assemblies, effective medium theories are better suited to explain their optical absorption spectra.1,10-12 Mies theory was developed for particles of spherical shape only. For cylindrical or oblate nanoparticles, Gans18 extended Mies theory within the dipole approximation. The particles are usually characterized by their aspect ratio, which is defined by the ratio between the length and the width of the particle. The plasmon resonance for nanorods splits into two bands. As the aspect ratio increases, the energy separation between the resonance frequencies of the two plasmon bands increases.19 The high-energy band corresponds to the oscillation of the electrons perpendicular to the major rod axis and is referred to as the transverse plasmon absorption. The other absorption band, which is red shifted to lower energies, is caused by the oscillation of the electrons along the major rod axis and is known as the longitudinal surface plasmon absorption. The transverse plasmon absorption band is relatively insensitive to the nanorod aspect ratio and coincides spectrally with the surface plasmon oscillation of the nanodots. The linear spectral dependence of the longitudinal surface plasmon absorption can be reproduced using Gans theory.19,20 Maxwell-Garnett theory,21 an effective medium theory, treats materials in which particles are isolated from each other by a layer of an insulating dielectric. MaxwellGarnett theory is based on the Clausius-Masotti equation, and it assumes that it is

justified to describe the composite material containing metal nanoparticles embedded in an inert host medium by an effective complex dielectric constant, which in turn is related to the optical refractive index and the absorption coefficient. Martin and coworkers22-23 have used and extended Maxwell-Garnett theory in order to explain the optical absorption spectra of needle-like and pancake-like gold nanoparticles in a porous alumina membrane. It has become necessary, however, to account for the optical absorption properties and the shifts in the plasmon resonance for any arbitrary shape of the particles. Schatz24 has developed and applied the finite element discrete dipole approximation (DDA) theory to nonspherical particles in particular nanoparticle arrays with varying size, shape, substrate, and solvent coverage. DDA methods divide the particle into a large number of polarizable cubes, and their induced dipole polarizations are determined self consistently. From these polarizations the overall extinction cross section can be determined. DDA theory has also been recently applied to nanoprisms25 produced by photoinduced irradiation of silver nanospheres and was able to reproduce the spectral shape of the nondegenerate plasmon resonances.

I. 4. Nanoparticle synthesis strategies: top-down and bottom-up approaches The technology for making nanoparticles are categorized into top-down and bottom-up approaches.26 Top-down (extreme miniaturization) approach is that nanoparticles are produced by breaking larger materials into fine particles, while bottom-up (building blocks) approach is that nanoparticles are built up from atoms or molecules. In the synthesis of metal colloids, controlling the size or shape is done by adjusting the ratio of the concentration of the chemicals making the nanoparticle to that of the selected capping material (e.g., polymers, micelles, surfactants, or dendrimers). This type of synthesis is termed as the bottom-up method. In the top-down method, the bulk is machined down to the nanometer length scale by lithographic or laser ablation-condensation techniques. A schematic presentation of the top-down (physical method) and bottom-up (chemical method) approaches is shown in Figure

10

5. From the viewpoint of mass production of metal nanoparticles, the chemical methods are more effective than the physical ones.

Bulk Metal Nanoparticles

Sub Division

Metal atoms Nanoparticles

Aggregation

Phys a ic

Figure 5. Schematic representation of the top-down (above) and bottom-up approach (below).

I. 4. 1. Chemical methods

Metallic nanoparticles can be prepared in chemical routes by iterative (single step) and non-iterative growth (seed-mediated) method. The simplest and the most commonly used bulk-solution synthetic method is the chemical reduction of metal salts. The particle synthesis usually makes use of a soluble metal salt, a reducing agent and a stabilizing agent. Reducing agents such as sodium borohydride27,28 and alcohols29 are commonly used for this purpose. The stabilizing agent caps the particle and prevents further growth or aggregation. The seeding-growth procedure (non-iterative) is another popular technique that has been used for a century. The sizes can be

11

manipulated by varying the ratio of seed to metal salt. The step-by-step particle enlargement is more effective than a one-step seeding method to avoid secondary nucleation and therefore, to control nanoparticle size distribution.30-32

12