Professional Documents
Culture Documents
3.1
Introduction
Particles with diameters in the range of 2 to 100 nm, so-called nanocrystalline materials, have become a major interdisciplinary area of research during recent decades.
In fact, since the seventeenth century, noble metallic nanomaterials, though not
understood, have been obtained and used to give rise to a brilliant rose color
throughout Europe in stained glass windows of cathedrals and by the Chinese in
coloring vases and other ornaments [1, 2]. The scientific preparation of nanoparticles dates back to the nineteenth century, with Faraday reporting the preparation of colloids of relatively monodispersed gold nanoparticles. The scientists who
major in nanoscience and nanotechnology should appreciate the inventors who
designed transmission electron microscopy (TEM). The high-resolution TEM
(HRTEM) and low-resolution (LRTEM) allow one to observe a substance at a
nanometer scale directly. The magic machines make it possible to investigate the
nanomaterials with respect to size, size distribution, shape evolution, and shape uniformity and even the structure. The past couple of decades have witnessed an exponential growth of activities in this field worldwide, driven both by the excitement of
understanding new science and by the potential hope for applications and economic
impacts. Indeed, many efforts have been devoted to investigation into the synthesis,
characterization, and application of nanomaterials. In general, nanomaterials can
be classified into three groups: zero-dimensional materials, so-called nanoparticles,
with variations in shape and diameter, one-dimensional materials, including
nanorod and nanowire, and two-dimensional materials, including nanobelts,
nanodisks, films, and nanosheets. Herein we focus on the nanoparticles (NPs), especially the metallic ones.
The intense interest in the metallic NPs derives from their unique chemical and
electronic properties arising from the small volume to big surface area ratio and the
separation in the electronic energy level. The change in the properties at this length
scale from their bulk counterparts is not a result of only scaling factors. It results
from different causes as far as different materials are concerned. In semiconductors,
it results from the further confinement of the electronic motion to a length scale that
is comparable to or smaller than the length scale characterizing the electronic
motion in bulk semiconducting material (called the electron Bohr radius, which is
43
44
45
3.2
46
past decades, many approaches have been utilized to produce the noble nanoparticles. Before moving into the idiographic synthesis methods, we discuss the particles growth in kinetics and the mechanism.
3.2.1.1
Introduction
The synthesis of colloidal gold has been studied intensively for a long time [18]. In
1994, Brust et al. [19] developed an effective method for the generation of
long-chain alkanethiolate-stabilized gold nanoparticles. The product from this system is easy to disperse in organic solvent and to reisolate as pure powders. Due to the
superb stability of the as-prepared particles, many groups have tried to follow or
47
develop the approach. Much work has been done based on this smart design to
modify the properties, such as the reactivity and solubility, of the nanoparticles
through changing the molecular structures of the thiolates on the particle surface
[2026]. Both long-chain thiolates containing an aromatic moiety [20], -substituted (cyano, bromo, vinyl, ferrocenyl) [21] and poly-hetero--functionalized [22]
long-chain alkanethiolates and small molecular thiolates such as ( -mercaptopropyl) trimethyloxysilane [23] and the rigid aromatic thiols of 4-mercaptobiphenyl [24] and p-mercaptophenol [25] have been utilized. Kunitake and
coworkers also showed that sodium 3-mercaptopropionate could be used for the
synthesis gold nanoparticles using citrate as reductant [26]. Since the diluted aqueous solution was needed in this method, it is difficult to separate the particles as
solid powders; therefore, the detailed characterization of the surface structure with
a variety of methods where solid powders are needed as the specimens is prevented.
In addition, regarding the application of the gold nanoparticles in biomedicine,
where the pH value is close to neutral, there still will be the challenge of developing
the approach from which the Au nanoparticles can be easily dispersed into aqueous
solution at a neutral pH value.
In 1999, Sihai Chen and Keisaku Kimura reported a new approach that can
avoid the disadvantages mentioned above [27]. The shining points of their work are
the standout elucidation of the surface structure of these particles and that the
as-synthesized nanoparticles are easily dispersible in water, a property that has not
been achieved using other kinds of methods where the thiolates are employed
[2026]. For biomedical purposes, the latter feature is a prerequisite because the
nanoparticles should combine with macromolecules in living or fixed cells or tissues
in aqueous solution.
Let us introduce the process briefly. In a typical preparation, aqueous solution
was at first mixed with mercaptosuccinic acid (MSA) in methanol to give a transparent solution. Under vigorous stirring a freshly prepared aqueous sodium
borohydride solution was then added at a certain rate. The solution turned dark
brown immediately but remained transparent (which indicates that the nanoparticles do not form) until enough reductant was added. The pH of the solution
increased gradually with the addition of reactant. A flocculent dark brown precipitate would be generated by further addition of the reductant, and finally the pH of
the solution was brought to 8.6. We can see that just as in the common synthesis in
aqueous solution, the pH must be adjusted. After being stirred for another duration,
the solvent was removed by decantation after the centrifugation. The precipitate
was washed, suspended in ethanol, and dried by rotary evaporation without
exceeding a temperature of 40C under vacuum and at last the product in powder
state was obtained.
They explained the whole procedure as follows:
8nHAuCL 4 + 8mR SH + 3nNaBH 4 + 12nH 2 O
8( Au ) n R S
(3.1)
The initial molar ratio between MSA and chloroauric acid plays a dominate role in
tailoring the relative rates of particle nucleation and (then) growth, and hence it
finally determined the particle size. It was found that the particle sizes decreased
48
with the increase in S/Au ratios. This result is similar to that reported by Leff et al. on
the long-chain alkylthiol-passivated gold nanoparticles [28], suggesting that the
thermodynamics are responsible for the particle stability.
One can see that H+ ions were released in reaction 1, while the final pH of the
reaction mixture increased with the addition of NaBH4 due to the strong hydrolysis
of NaBH4, which can be elucidated as follows:
NaBH 4 + 4H 2 O NaOH + 4H 2
(3.2)
Hence, the final pH is determined by the competition between reactions (3.1) and
(3.2). We keep our eyes on the pH value because it influences the dispersibility of the
gold nanoparticles, which is important in terms of biomedical applications. For
instance, the dispersibility of gold particles abruptly decreased when the pH of the
solution increased to around 5 during the addition of NaBH4. Furthermore, due to
these pH-dependent properties, gold particles can be easily separated from inorganic
and organic impurities by increasing the pH of the solution.
The notable property of the as-synthesized nanoparticles is that they are easily
dispersed in neutral water, since before this report gold nanoparticles were found to
be dispersible only in alkaline aqueous solution (pH 13) using other kinds of
thiolate such as p-mercaptophenol [29]. For the bioapplication purposes being considered, the pH value is too high. However, the size-dependent dispersibility of the
powder in water (that is, its increase as the particle size decreases) may be a shortcoming rather than an advantage because it suggests a poor MSA coating on the surfaces of larger gold nanoparticles, which means that the larger particles from this
method may be not suitable for bioapplications.
As mentioned in the introduction section, one problem is that experimental
measurements for the small nanoparticles are often degraded by a lack of size and
shape uniformity that renders comparison with theory questionable [30]. Hence, it
is almost impossible to unequivocally identify quantum size effects in the optical
spectra of metal nanoparticles prepared in macroscopic quantities, although such
effects are well known from experiments on metal-cluster beams and from conductance measurements on single-metal nanostructures [31].
As the saying goes, where there is a will, there is a way. The discovery of methods for preparing gold nanoparticles by Dubois and Nuzzo in 1992 [32] has overcome this obstacle successfully. Gold alkylthiolate assemblies that are a
nanometer-scale analog to the well-studied surface system were utilized. One notable point in this synthesis is that on the extended Au surfaces, one finds an extraordinary example of the uniform protection of the surface without modification of its
essential structural and electronic properties. N-Alkylthiolates and their derivatives
form compact, ordered monolayers where thiolates (RS-) or dialkyl disulfides
(RSSR) reversibly attach to various Au surfaces. Because of favorable (van der
Waals) interactions among the long, ordered alkane chains in concert with a weak
chemisorption of the sulfur head group to the metal surface (~1.2 eV per RSSR unit
desorbed versus ~10 eV surface energy per six Au atoms covered by this unit), the
highly ordered self-assembly occurs spontaneously. These extensively studied systems have recently begun to be well understood. For instance, a surface structure
49
Population (%)
Magnetic nanoparticles, especially those with diameters of less than 10 nm, are of
intense interest because the particle size is smaller than the characteristic dimensions
such as atomic or ionic diffusion length, electronic elastic and inelastic mean free
path length, and the magnetic domain size [3539]. Many new phenomena and
properties are expected to appear in the NP systems. The size and shape specificity
of nanoparticles naturally acts as building blocks of the self-assembled passivated
NP superlattices (SLs) or NP arrays [40]. In this emerging field, monodisperse magnetic NPs are a prerequisite to a broad range of applications, such as data storage
devices, biosensors, and drug delivery [4144]. The advances in preparing semiconductor and metallic NPs, specifically by the method of injecting molecular precursors into hot organic surfactant solution, have improved the NP samples
remarkably, with good size control, narrow size distribution, and good crystallinity
of disperse nanoparticles [45, 46].
Now we introduce the general approach though the synthesis of cobalt
nanoparticles. The reason we chose to begin with the Co system is that Co nanocrystals possess a wealth of size-dependent structural, magnetic, electronic, and
30
25
20
15
10
5
0
90 nm
3.0 3.5 4.0 4.5 5.0 5.5
Diameter (nm)
Figure 3.1 TEM image of a two-dimensional silver nanoparticle superlattice and (inset) the histogram of the nanoparticles. (Source: [34].)
50
51
52
presence of DiBAH have been reported by them [69]. In that work, they obtained Co
nanoparticles with small mean diameters of about 1.4 nm, by means of hydrogenation of Co(3C8H13)(4C8H12) companied by poly(vinyl pyrrolidone). A comprehensive study of their magnetic properties has shown that they are similar to those
expected for free-standing cobalt NPs, which were discussed in 1998 [70]. Briefly, if
one assumed that the polymer matrix has few or no interactions with the NPs, the
deduction is that surface metal atoms are undercoordinated and electronically
unsaturated. The noteworthy point is that the formation of -cobalt has been
reported to be promoted by a careful choice of the ligands used during the synthesis
[53, 54], which will be discussed latter. Hence, in the presence of trioctylphosphane
oxide (TOPO) or PR3 ligands, which are known to damp the magnetic properties of
NPs, cobalt NPs adopt the -cobalt structure, whereas an fcc structure is observed in
the absence of such ligands [70]. In sample 1, the electron deficiency is also important. On the contrary, NPs in sample 2 (when [CoN(SiMe3)2]2 is utilized as precursor; for detail, see [71]), the surfaces of which are surrounded by donating ligands,
adopt a compact structure displaying features of both fcc and hcp phases. It is worth
noting that the hcp structure is not often reported for such small sizes, especially for
NPs obtained at low temperature. It is suggested that coordination of a -donor
ligand at their surfaces favors this structure. The synthesis conditions also play an
important role on the structure of the product. In fact, sample 1 is obtained at room
temperature by a slow reaction process (48 hours), whereas the formation of sample
2 is fast at 50C. This is in contradiction with the expected kinetic quenching of a
metastable structure and further attests the influence of the coordinated ligand on
the structure adopted by the NPs.
Now let us come to the identification of the new phase of cobalt in detail. As was
mentioned, besides the well known structures including hcp and fcc, another phase
of cobalt has been found in cobalt nanomaterials. Though the insight was reported
by Hull [72] in 1921 and then Krainer [73] and Kajiwara et al. [50], it is from 1976
that the evidence for a new structure of cobalt present in small cobalt clusters produced by the decomposition of organometallic precursors is provided by Respaud et
al. [74]. However, the structure was not identified until Dinega and Bawendi
reported their work on this new structural cobalt in 1999 [53]. In the report, thermal
decomposition of octacarbonyldicobalt in solution in the presence of TOPO as a
coordinating ligand was employed to synthesize cobalt nanoparticles. This method
provides a clean route for the preparation of the material since elemental cobalt is
the only nonvolatile product of the reaction, which can be depicted as follows:
[Co
(3.3)
They claimed that the as-synthesized powder was highly susceptible to magnetic
fields generated by a small permanent magnet, suggesting that it consisted of magnetic materials without or with thimbleful of cobalt oxides. TEM showed that the
powder consisted of roughly spherical particles with an average diameter of 20 nm
and a 15% size distribution. Later experiments showed that the size of the crystals is
not limited to this range. In fact, injection at higher temperature yields regular
polyhedron-shaped faceted crystals as large as 0.3 m and possibly even larger
(Figure 3.3). These results are reasonable since we know that the temperature will
611
442
610
531
510
511
330
420
321
210
211
110
520
521
310
53
311
221
(a)
(b)
(c)
20
30
40
50
60
70
80
90
100
2/
Figure 3.2 (a) Experimental and (b) calculated X-ray powder diffraction patterns of the -cobalt
structure and (c) the sample shown in (a) after being heated to 500C. Peaks corresponding to fcc
cobalt are denoted by X. (From: [53]. 1999 Wiley-VCH Verlag GmbH, Weinheim, Germany.
Reprinted with permission.)
influence the nucleation and the growth process, as discussed previously. They
also found the presence of oxygen, which indicates surface oxidation of the cobalt
particles, corresponding to coverage of one to two monolayers of cobalt oxide for
particle sizes in the 20-nm range. The other elements, in their opinion, come from
residual organic solvents adsorbed on the particle surfaces. The absence of phosphorus in the elemental analysis indicates that TOPO was completely removed during washing of the particles. It is noted in this work that freshly synthesized cobalt
nanoparticles are extremely reactive toward oxidation. In fact, rapid exposure to air
results in immediate oxidation accompanied by a red glow. Therefore, even simply
washing the particles in organic solvents inevitably leads to surface oxidation by
residual moisture and dissolved air. The consideration is reasonable that the resulting oxide layer appears to passivate the surface of the particles and considerably
reduce the speed of further oxidation. It should be mentioned that the new -cobalt
phase appears to be metastable under normal conditions. Although stable at room
temperature for at least several months, heating the sample at 500C completely
transforms it to the known fcc phase. Furthermore, subsequent cooling does not
return the sample to its original -cobalt structure.
Figure 3.2 shows the X-ray powder diffraction patterns of the as-synthesized
product, which present the evolution of the nanoparticle structure with variation of
the molar ratio of TOPO/precursor, and one can see that this product takes a new
54
phase that does not correspond to either of the two known structures of cobalt, fcc
and hcp. In other words, neither the positions of the peaks nor their relative intensities correspond to any known cobalt phases. Based on the detailed analysis, the conclusion is that this pattern corresponds to a new structure of cobalt, so-called
-cobalt. This structure is cubic (space group P4132) with a unit cell parameter =
6.097 0.001 . The unit cell structure is similar to that of -manganese (a
high-temperature phase of manganese) [75]. Twenty cobalt atoms constitute one
unit cell, and they are divided into two types: 12 atoms of type I and 8 atoms of type
II. These two types of atoms differ in their local coordination. Unlike an ideal
close-packed structure, which has 12 nearest neighbors, -cobalt has only three
nearest neighbors for type I and two nearest neighbors for type II atoms. This phenomenonthe same atoms being divided into several types based on the local coordinationoften happens in compounds such as magnetite. This results in a structure
for -cobalt that is less dense than both the hcp and fcc structures. The calculated
density of -cobalt is 8.635 g(cm)3, compared to 8.788 g(cm)3 and 8.836 g(cm)3
for the fcc and hcp structures, respectively [53].
Hence, the structure of -cobalt was identified for a certainty. Figure 3.3 shows
the TEM images of the as-obtained nanoparticles. However, as far as the uniformity
of the size and shape is concerned, which we continue to emphasize, the system
needs to be developed for the uniformity of the morphology.
Figure 3.4 shows models of the unit cell for -cobalt with respect to different lattice direction projections. In addition, for the purpose of applications, it should be
mentioned that the magnetic moment per atom in -cobalt, measured as 1.70 B, is
similar, within experimental error, to those of the two known structures (1.75 and
1.72 B for bulk fcc and hcp cobalt) [53, 76].
It was also claimed that the use of TOPO seized the key role here because of the
fact that the same preparation method but without the presence of TOPO yields
nanocrystals with a pure fcc phase. Hence, the influence of TOPO on the structure
of the as-obtained nanoparticles should be clarified, and it was indeed done. Tight
coordination of ligand molecules (TOPO) around the growing crystal and around
solubilized cobalt atoms is responsible for changing the energetics of growth in favor
of the new, less dense phase. However, the particular synthesis of this paper is not
unique in producing -cobalt. Sun and Murray [54] have confirmed this assignment
0.1
0.1m
Figure 3.3 TEM image of cobalt crystals with the new -cobalt structure. (From: [53]. 1999
Wiley-VCH Verlag GmbH, Weinheim, Germany. Reprinted with permission.)
(a)
55
(b)
Figure 3.4 The unit cell of -cobalt: (a) unit cell cube filled with 8 atoms of type I (dark) and 12
atoms of type II (light) and (b) 111 projection of the same cube showing threefold symmetry along
its main diagonal. (From: [53]. 1999 Wiley-VCH Verlag GmbH, Weinheim, Germany. Reprinted
with permission.)
of the structure of -cobalt in nanocrystals obtained by reducing cobalt salts in solution in the presence of alkylphosphanes serving as the coordinating ligands. Due to
its metastability, -cobalt may be accessible only by solution-phase approaches,
rather than more common techniques, such as electrochemical deposition or chemical vapor deposition and so on. As far as the uniformity of nanoparticle morphology is concerned, remarkable achievements have been obtained by Sun and
Murrays work. In their work, briefly, the particle generation began with the injection of dioctylether superhydride solution into a hot (200C) CoCl2 dioctylether
solution in the presence of oleic acid (octadec-9-ene-1-carboxylic acid,
CH3(CH2)7CH=CH(CH2)7COOH) and trialkylphosphine (PR3, R=n-C4H9, or
n-C8H17). Reduction occurred instantly upon injection, leading to the simultaneous
formation of many small metal clusters serving as nuclei. Continued heating at
200C allowed steady growth of these clusters into nanometer-sized, single crystals
of cobalt. The steric bulk of the alkylphosphine controlled the rate of particle
growth and consequently the morphology. Short-chain alkylphosphines (e.g.,
tributylphosphine) allowed faster growth, leading to the generation of large particles, while bulkier species (e.g., trioctylphosphine) reduced particle growth and
favored production of smaller nanocrystals (NCs). However, it is worth mentioning
that the influence of the organic coatings steric hindrance on the particles size, size
distribution, and shape must play some essential role in such cases. These organically stabilized cobalt NCs were readily dispersed in aliphatic, aromatic, and chlorinated solvents and could be precipitated with the addition of short-chain alcohols,
facilitating a size-sorting procedure that isolated nearly monodisperse samples [77].
In Sun and Murrays work, average particle size was coarsely controlled by
varying the type of alkylphosphine used in combination with oleic acid during the
growth. For instance, bulky P(C8H17)3 limited growth to producing particles with
diameters of 2 to 6 nm, and less bulky P(C4H9)3 led to larger particles (7 to 11 nm in
diameter). A size selection procedure is needed for the purpose of fine tuning the
particle size. It should be pointed out that the size sorting procedure is the disadvantage of such a system in view of the cost for application in industry. And some
groups developed approaches that lead to nanoparticles with tunable size and narrow size distribution without the size selection procedure. Based on the fact that
trialkylphosphine reversibly coordinates neutral metal surface sites, slowing but not
56
stopping the particles growth, it cannot prevent the particles from eventually growing to undispersible aggregates at high temperature when used alone. If oleic acid is
employed alone, though it is an excellent stabilizing agent, it binds so tightly to the
particle surface during synthesis that it greatly impedes the particle growth. The
combination of trialkylphosphine and oleic acid produced a tight ligand shell that
allowed the particles to grow steadily while it protected them from aggregation and
oxidation.
It is well known that a temporal separation of the nucleation and growth stages
is needed to produce a monodisperse colloid. Ideally, a large number of critical
nuclei should be formed in a short interval of time followed by the simultaneous and
steady growth of those nuclei. In Sun and Murrays approach [54], the reducing
agent (superhydride) was injected to induce rapid reduction of cobalt (II) chloride in
hot (200C) media. This procedure provided the temporally discrete homogeneous
nucleation desired. Growth of the nuclei was continued by addition of cobalt-containing species to the surface of the particles and was halted by cooling to below
100C naturally. From the TEM characterization (Figure 3.5), one can see the standout results obtained by this means. However, the size selection procedure is needed.
This may lead to a high cost in terms of economics when it is realized in industry.
In 2003, a report from Gaos group presented the advancement in the control of
size, size distribution, and the monolayer and multilayer self-assembly based on the
narrow size distribution [78, 79]. In their work, they employed the approach in
which thermal decomposition of octacarbonyldicobalt was utilized. This method
provides a clean route for the preparation of the material since the cobalt is the
only nonvolatile product during the whole and the reaction can be described by (3.3)
[53, 80]:
A combination of surfactants in the presence of stabilizing ligands triphenylphosphine and oleic acid was employed for controlling the particle growth, stabilizing the particles, and preventing oxidation. Because the phenyl can provide greater
steric hindrance than that of the straight-chain alkyl, they adopted the triphenylphosphine, instead of the tributylphosphine or trioctylphosphine, to synthesize
cobalt NPs with smaller size.
48nm
Figure 3.5 TEM image of a two-dimensional assembly of 9-nm cobalt nanocrystals. Inset:
High-resolution TEM image of a single particle. (From: [54]. 1999 American Institute of Physics.
Reprinted with permission.)
57
3.3
In the following we will locate space to discuss the properties of metal nanoparticles, especially those with organic coating, because these underlie the application of the nanoparticles in all kinds of purposes, and in most cases the ligands
influence the properties of the nanoparticles, more or less. According to the backbone of this issue, let us begin with those of noble metal nanoparticles.
At a fundamental level, optical absorption spectra provide information on the
electronic structure of small metallic particles. The absorption spectra of many
metallic nanoparticles are characterized by a strong broad absorption band that is
58
absent in the bulk spectra. Classically, this giant dipole (or surface plasmon) band is
ascribed to a collective oscillation of the conduction electrons in response to optical
excitation. The optical spectra of gold particles with diameters in the range of 1 to
3.4 nm in aqueous solution are shown in Figure 3.7 [27]. For the entire samples one
can find the continuous background rising toward higher energy. The Mie scattering
from the nanoparticle suspension is considered to be responsible for the particles
smaller than 2 nm in the solutions presenting orange-brown, and no additional
absorption peak appears around 500 nm. This result can also be found in the works
of Fauth et al. [81] and Duff et al. [82].
The spectra are normalized arbitrarily at 800 nm and offset vertically. The solutions are all made in distilled water at a concentration of 86 g/mL.
When the particle size increases to 2 nm [Figure 3.7(d)], superimposed on the
background, a broad surface plasmon band around 500 nm occurs that is attributed
to a collective oscillation of conduction electrons in response to optical excitation
[83]. This band becomes more apparent for the particles with average diameters of
3.4 nm [Figure 3.7(e)]. A similar phenomenon has been reported by Robert L.
Whettens group. Based on an optical study using long-chain thiolate-modified gold
nanoparticles in a similar size range (from 1.4 to 3.2 nm), they reported that the surface plasmon band was essentially unidentifiable for crystallites of less than 2.0 nm
effective diameter [84].
(a)
100 nm
(b)
80 nm
Figure 3.6
59
1.6
Absorbance
1.2
0.8
e
d
c
b
a
0.4
0.0
300
400
500
Wavelength
600
700
800
Figure 3.7 UV-visible spectra of Au particles of different sizes: (a) 10.2, (b) 10.8, (c) 12.8, (d)
19.4, and (e) 33.6 . (From: [27]. 1999 American Chemical Society. Reprinted with permission.)
60
be prepared quantitatively by at least two distinct methods [19, 32], effective across
the ~1.4- to 3.5-nm core diameter range (~70 to 1,000 atoms). Their surface properties are determined by the surfactant tail group (usually methyl, CH3), and hence
the interactions with external agents (e.g., solvents, strong acids, or bases) and with
each other is weakened; the surfactant groups (RS) initially adsorbed can later be
exchanged with excess thiol (RSH) in solution [8587]. A small metal particle surrounded by a condensed dielectric medium [88] governed the electronic properties.
It is notable that the raw, as-synthesized samples have been shown by mass spectrometry to contain enhanced abundances at certain sizes, which is probably attributable to the filled structural shells of Au atoms. The components of these mixtures
can be fractionated from one another according to size and then handled as purified
molecular substances in various manners [87, 89]. Realistic simulations [90] and
X-ray diffraction patterns obtained on selected samples provided considerable
insight into the structure of these assemblies and their interactions, which is consistent with patterns calculated from theoretically generated structures. Whetten et al.
[91] in 1997 found methods to produce and isolate smaller nanocrystals (1.4 to 1.7
nm) in large quantities, which allows one to obtain a rather complete picture of the
evolution of the optical properties of gold nanoparticles.
Based on the achievement mentioned above, the scientists are able to investigate the properties, especially the optical ones, at will. The spectra shown in Figure
3.8 are arbitrarily normalized at the high-energy end and are presented superimposed, without risk of confusion, based on the fact that the relative spectral intensity
at 2.5 eV decreases monotonically with mass (core size) for each series shown. Furthermore, Figure 3.9 suggests that chain lengths for a given core mass have nearly no
influence on the spectra, while at smaller sizes (~1.7 nm) they can be very sensitive to
sample purity [Figure 3.9(b)].
Cursorily, the evolution of the optical spectrum of Au particles with diameters
ranging from 1.4 to 3.2 nm is so simple that it could even be described as that which
is already known about Au from spectroscopic investigations of unfractionated
samples of nanometer-scale particles in glasses, in solution, and from the smaller
cluster compounds [92, 93]. However, the use of well-fractionated samples of gold
nanocrystals, grown in a strongly etching environment [89], allows one to distinguish certain features that had not been noticed previously:
There is an onset of strong absorption located near 1.7 eV, where it behaves
increasingly distinctly at smaller nanocrystal sizes. These show a clean break
from the preonset behavior occurring at the energy of ~1.6 eV, which can be
attributed to the interband (d sp) absorption edge, as shown in the semilogarithmical scale in Figure 3.10.
A weak, steplike structure occurs, beginning with the onset mentioned earlier,
in the spectra of all the smallest clusters studied. It is clearly evidenced in the
differential (or log-differential) spectral intensities plotted in Figure 3.11 (and
its inset), which transforms the weak steps into a series of distinct peaks and
shoulders. These also bring out the onset structure (1.7 eV) as a first peak.
For these effects to be mastered and understood, one should acquaint oneself with
the elementary facts of the electronic structure of close-packed Au, as are confirmed
61
(a)
C6
0
(b)
C12
0
(c)
C18
0
Figure 3.8 Optical absorption spectra for dilute solutions of several purified gold molecules passivated by (a) hexyl-, (b) dodecyl-, and (c) octadecylthiolates. The spectra are scaled to unity at 4 eV
and are compared to an aqueous solution of commercial colloidal Au particles (dotted lines) of
9-nm mean size. The peak amplitude (near 2.5 eV) descends with the metallic core diameters: (a)
3.2, 2.5, 2.4, 2.2, 2.0, and 1.7 nm (SC6 passivant); (b) 2.5, 2.4, 2.2, 2.1, 2.0, and 1.7 nm (SC12
passivant); and (c) 1.7 and 1.4 nm (SC18 passivant). (From: [84]. 1997 American Chemical Society. Reprinted with permission.)
62
(equivalent to a dielectric constant near 10) results in a second effect, which is the
large redshift of the collective resonance to its observed location (2.4 eV), with the
unusual results that its intensity is dominated by the interband transitions, its existence by the intraband transitions, and its location by the latter but modified
strongly by the polarizable Au+ ion cores. It seems that the breadth and proximity of
this resonance compared to the interband transition onset will make it more difficult
to distinguish the interband absorption edge. Fortunately, Taylor et al. [94] gained
the photoemission spectra on mass-selected cold AuN- beams, thereby demonstrating that this interband energy, which is located at just below 2 eV, was defined
clearly and exactly for clusters of N = 21 to more than 200 atoms. As a result, its
location is probably constant for all close-packed Au clusters, in spite of the fraction
of atoms at the surface. Based on these facts, the onset feature observed near 1.7 eV
can be ascertained without ambiguity to the first interband transitions, that is, to
transitions from the highest occupied d orbitals to the lowest unoccupied level(s) of
the conduction band.
Because no such structure has previously been ascertained in condensed Au, it
is not easy to propose an interpretation of the discrete steplike structure (the peaks in
Figure 3.11). However, by elimination methods, the evidence points toward the reason that it results from the pattern of unoccupied energy levels located just above the
Fermi level, that is, from the quantum size effect in the conduction band. Each step
would then represent a newly accessible channel in the sparse conduction bandlevel
structure for transitions from the quasicontinuum of d-band levels.
Scientists prefer to address the applicability of various size-dependent corrections to the optical properties derived from bulk Au in a quantitative analysis of the
optical spectra of passivated gold nanocrystals. The effects of surface scattering
(mean free path correction), surfactant effects on the core electron density, and
(a)
(b)
3.5
Figure 3.9 Optical spectra comparing nanocrystals of like metallic cores. (a) Spectra (normalized
at 4 eV) from 2.5-nm-diameter clusters passivated with hexyl (solid line) and dodecyl (dashed). (b)
Spectra (offset for ease of viewing) of 1.7-nm clusters passivated by (top to bottom) SC6, SC12,
and SC18. (From: [84]. 1997 American Chemical Society. Reprinted with permission.)
63
quantum size effects are taken into account. Now it is time to discuss the basic Mie
theory.
The essential assumption of Mies theory of (linear) optical absorption by
small particles is that the particle and its surrounding medium are each homogeneous and describable by their bulk optical dielectric functions. Briefly, by analyzing Maxwells equations for this geometry one can obtain an expression for the
absorption cross section, which is a sum over electric and magnetic multipoles
(spherical vector harmonics and Legendre polynomials). In the case that the size of a
particle is much smaller than the wavelength of the exciting radiation, the absorption is dominated by the dipole term, with a cross section () given by [96]
( ) = 9 m3 / 2 V0
2 ( )
c ( ) + 2 2 + 2 ( )
m
1
2
(3.4)
In this expression, is the frequency and c is the speed of light, m is the dielectric
constant of the embedding medium (assumed to be frequency independent over the
spectral range of interest), V0 is the volume of the absorbing particle, and 1() and
2() are the real and imaginary parts of the frequency-dependent dielectric constant of the absorbing solid.
According to the bulk optical constants [97] extracted by cubic spline fit,
Whetten calculated the bulk dielectric function for Au and the corresponding
absorption spectrum based on (3.4). The spectrum (Figure 3.12), calculated assuming m = 1, shows a giant dipole resonance peaking near 2.3 eV, which is consistent
with the vanishing of 1() + 2m . From (3.4) one can see that the spectral features
10.00
1.00
0.10
0.01
1.0
1.5
2.0
3.0
2.5
Energy (eV)
3.5
4.0
Figure 3.10 Optical spectra (offset for ease of viewing) of diverse selected fractions replotted on a
semilogarithmic scale (top to bottom): 9-nm colloidal gold and 3.2-, 2.5-, 1.7-, and 1.4-nm
nanocrystal gold molecules. The vertical dashed line marks the onset of the interband absorption
just above 1.6 eV. (From: [84]. 1997 American Chemical Society. Reprinted with permission.)
64
dA/d
0
1.5
1.5
2.0
2.5
Energy (eV)
2.0
2.5
3.0
3.5
Figure 3.11 Derivative (with respect to wavelength , dA/d ) of the optical absorption of diverse
selected fractions (top to bottom): 2.6-, 2.4-, 2.2-, 2.0-, 1.7-, and 1.4-nm-diameter nanocrystal
1
gold molecules. The inset shows the logarithmic derivative (A dA/d ) for the smallest of these.
(From: [84]. 1997 American Chemical Society. Reprinted with permission.)
are size independent and the particle dimension acts only as a volumetric scaling factor. To obtain a size dependence for metal particles, one must decompose the dielectric function into two terms: an interband (IB) contribution, accounting for the
response of 5d electrons, and a free-electron contribution (Drude, D) [98] from the
electrodynamics of the nearly free conduction electrons [99]:
1 ( ) = 11B ( ) + 1D ( ),
1 ( ) = 1
2p
2 + 20
2 ( ) = 21B ( ) + 2D ( )
2D ( ) =
2p 0
2 + 20
(3.5)
(3.6)
(3.7)
which corresponds energetically to 8.89 eV for gold and 9.08 eV for silver, respectively [98]. The term 0, which usually takes a magnitude on the order of hundredths of an electron volt, acts as the deputy of the frequency of inelastic collisions
(electron-phonon coupling, defects, impurities) of free electrons within the metal.
From Figure 3.12 we can see the intuitive separation of susceptibilities of bulk fcc
Au. The variation in the 1() is free electron in nature below 2.4 eV, although offset
by the large, positive, and near-constant interband contribution. A significant inflection appears near 2.4 eV. The 2() function is dominated by the free-electron term
65
10
11B
ID
10
20
(a)
30
40
50
6
21B
5
4
(b)
3
2
2
1
2D
0
1
1.5
2.5
3.5
4.5
Energy (eV)
Figure 3.12 Decomposition of the experimental dielectric functions (solid lines) of bulk fcc Au
into free-electron (Drude) and interband (5d f 6sp) contributions: (a) real and (b) imaginary parts.
The onset of significant interband absorption [dotted curve in (b)] can be seen near 1.8 eV. (From:
[84]. 1997 American Chemical Society. Reprinted with permission.)
below 1.7 eV and by the interband contribution above 2.4 eV. To introduce size
effects, one should assume that as the size of the particle diminishes, the rate of scattering from the particle surface (s) begins to greatly exceed the bulk scattering rate
0 [99]. The surface scattering rate is described in terms of the Fermi velocity (F =
8
1.410 cm/s for gold and silver) and particle radius [98]:
s = A F / R
(3.8)
This expression can be explained as a limitation on the mean free path of the free
electrons by the particle sizes. The coefficient A is of the order of unity, and its
meaning is not sufficiently identified [100]. It takes the value unity if the scattering is
assumed to be isotropic, three-fourths if diffusive, or zero if elastic. When other factors such as electron density at the surface, the effect of the interface, the anisotropy
of particle, and quantum mechanical computations are incorporated into the theory, values from 0.1 to above 2 can be deduced and theoretically justified [6]. To
identify dielectric functions of the particles, one assumes that the interband contribution (Figure 3.12) is the same as that of the bulk, but that free-electron contributions for small particles use s in place of 0 in (3.6).
66
Assuming that the scattering rate in nanomaterials is the same as that in the
bulk, the researchers calculated the size evolution of the optical absorption spectrum
and compared the results with those obtained using this mean-free path correction
(A = 1). The conclusion can be described as follows.
The damping and redshift of the giant dipole band are clearly evident. The
redshift in the position of the plasmon band is the result of the correction to 1,
which is not negligible at smallest sizes [99].
Whetten and coworkers attempt to fit the measured absorption coefficients
recorded on SC12 surfactant with Mie theory and simple MFP correction (3.4) to the
Drude terms in the dielectric function. The fits failed to describe the broadening of
the surface plasmons and their positions. Nor did they observe a relative increase in
low-energy absorption, accompanying the quenching of the surface plasmon, which
is demanded by the functional behavior in the mean-free path model (Figure 3.13).
Enlightened by Hengleins work [101], Whetten and coworkers take that
adsorbates can significantly affect the metals electronic properties, both for bulk
surfaces and in small particles as the second correction. The dependence in (3.7) of
the plasmon frequency (p) on the free-electron density N allows for a simple
phenomenological description of the adsorbate effect. It is possible that they provide
or withdraw additional electron density at the interface based on the facts that
thiol(ate)s are intimately bound to the particle surface. In the case of flat gold surfaces with self-assembled thiol monolayers on them, approximately 0.2 nm2 of surface supports one thiol [102], and the coverage density is increased in small particles
because of curvature introduced by edges and vertices [90]. When correction is
0
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Energy (eV)
Figure 3.13 Absorption spectra predicted by Mie theory for particles of decreasing diameter (top
to bottom): 40 (bulk), 3, 2, and 1 nm. All curves were normalized to unity at 4.1 eV. The size
dependence results entirely from mean-free path corrections to both real and imaginary dielectric
functions. (From: [84]. 1997 American Chemical Society. Reprinted with permission.)
67
When the magnetic properties are talked about in terms of magnetic metallic
nanoparticles, the most noteworthy point of the origin is that the volume is small
and the surface/bulk ratio is big. Because of the small volume, the thermal fluctuation energy of the magnetic moment can compete with the magnetic barrier energy
and even overcome it. Hence, the magnetic moment can rotate freely despite the
magnetic barrier energy. In the most cases, the saturation magnetization is smaller
than the bulk counterpart, and it is attributed to the surface/volume ratio and the
situation of the surface.
In general, there are two basically points in the nanoparticles that are different
from those of their bulk counterpart. First, and also most important, when the particle size is smaller than a certain critical value, they present very different behavior
from their bulk counterpart. Concretely, at room temperature, they exhibit zero
coercive field and no remanence, but they do saturate and the slope of the M-H
68
curve varies continually before it reaches the saturation state. Second, the measured
average moment per molecule (atom, in the case of a metallic nanoparticle) usually
is smaller than the one in bulk materials. The loss in magnetization corresponds to a
partial quenching of the magnetic contribution of the surface. The fact that the magnetization of the surface layer is not completely quenched is explained by a partial
coverage of the surface, in good agreement with the bulkiness of the ligands. Before
the discussion of the magnetic properties, we introduce some terms that are needed
when we talk about the magnetic properties of nanoparticles.
Following are the achievements obtained to date. As always, the statements are constructed based on the outstanding work from different groups.
In the work reported by Gaos group [79], magnetic properties of the ~7-nm
cobalt nanoparticles deposited on highly oriented pyrolytic graphite (HOPG) substrate are measured by SQUID using a standard airless procedure. The magnetization as a function of the temperature in a 10-Oe field between 5K and 300K
determines the blocking temperature using a zero-field cooling (ZFC) procedure.
Figure 3.14(a) shows the typical result for magnetic NPs. One can see that below the
critical size at which a particle becomes a single-domain magnet and is small
enough, the nanoparticles display superparamagnetism [106] at high temperature.
From the concrete measurement data, the blocking temperature (Tb) is 92K. The
broad transition from superparamagnetism to ferromagnetism shown in Figure
3.14(a) around 92K is probably due to the magnetostatic particle interactions in the
close-packed arrays. The blocking temperature should roughly satisfy the
relationship
Tb = KV / 30kB
(3.9)
69
M (emu, normalized)
M (emu/g)
20
100
150
Temperature (K)
(a)
50
(b)
200
250
(c)
10
0
10
250K
20
60
(d)
10K
(e)
M (emu/g)
40
20
0
20
40
10K
10K
60
20
10
H(kOe)
10
10
10
20
H(kOe)
Figure 3.14 (a) ZFC magnetization versus temperature of cobalt NPs showing a blocking temperature (Tb) of 92K. Hysteresis loop of powder of cobalt NPs compacted into a capsule obtained at (b)
250K and (c) 10K. Hysteresis loop obtained at 10K: (d) diluting cobalt NPs with wax and (e) cobalt
NPs deposited on HOPG and dried under N2 to prevent oxidation. (Source: [79].)
magnetization Mr is about 1.5 emu/g, the coercive field Hc is 163 Oe, and the magnetization at saturation (Ms) is estimated to be only 14.0 emu/g (the estimation is
based on an extrapolation of curves of H/M versus H).
The noteworthy point of this work is the measurement conducted in different
states of the cobalt nanoparticles. Yang et al. diluted the Co NPs in wax with a mass
ratio of cobalt nanoparticles:wax equal to 1:4. Figure 3.14(d) shows the hysteresis
loop of diluted particle powder, and a clear change in the shape of the hysteresis
loop can be found. The Mr reaches 7.3 emu/g, Ms reaches 59.6 emu/g, and Hc
increases from 163 to 600 Oe in comparison with the cobalt nanoparticles in powder states. The hysteresis loop of ordered arrays of cobalt nanoparticles on HOPG
substrate is shown in Figure 3.14(e). The Mr reaches 12.6 emu/g and Hc increases to
790 Oe. However, the improvement of Ms is not obvious compared to that of the
particles diluted with wax, which is 61.6 emu/g. These values are lower than those
70
obtained from the bulk phase. Because TEM images taken over large areas of the
sample show no evidence of the coalescence, the observed changes cannot be attributed to the coalescence of the cobalt NPs. They suggested that the exchange coupling between adjacent particles should account for explanations for the change in
magnetic properties. The dipole coupling enhancements are attributed to the longrange order of the two-dimensional lattice and collective flips of the magnetic
dipoles.
In the work of Catherine Amiens [6467], she and coworkers discussed the magnetic properties of the cobalt NPs they synthesized. The most striking result is the
high differential susceptibility observed in high fields and leading to a very low magnetization at 5T in a certain sample (namely, sample 1). The fact that the NPs adopt
structures very different from the bulk ones, even if it may lead to strong changes in
magnetic anisotropy, should not account for this strong decrease in magnetization.
Indeed, body-centered-cubic (bcc), fcc, and hcp phases present very close values of
magnetic moment per atom (1.6 to 1.7 B/Co). It is also the same in -Co, where the
value determined is close to 1.7 B/Co, or in polytetrahedral arrangements, where no
magnetization reduction could be evidenced, whereas the atom packing is very different from the bulk one [52, 53]. Either partial oxidation of the NP surface or a
strong damping of the magnetic moment of the surface atoms introduced by coordination of the ligands or other chemical species at their surface could then result in
this magnetization damping. Either through a strong alteration of surface anisotropy or via the formation of a diamagnetic surface layer, a strong coordination
effect would also be responsible for the fact that the magnetization is difficult to saturate. Oxidation can be ruled out by the given XANES and EXAFS spectra. This
sample was synthesized under a di-hydrogen pressure, whereas another sample
(namely, sample 2) was not. In fact, surface hydrides are expected to be present at
the surface of the particles, as recently demonstrated for Ru NPs [104]. One can thus
question the effect of hydrogen adsorption on the magnetic properties. However, in
Selwoods book one can find experiments showing that hydrogen adsorption
induces only a small decrease of magnetization in the case of cobalt [107]. Furthermore, as Amiens et al. demonstrated in 1998, NPs prepared in the same conditions
of hydrogen pressure but stabilized by poly-(vinylpyrrolidone), displayed magnetic
properties similar to those of free-standing Co NPs [52]; hence, the possible role of
coordinated hydrides in the low Ms value found for sample 1 was ruled out. However, the ligand used during the synthesis of sample 1 can strongly interact with
hydrides coordinated to transition metals [24]. In this particular case, the formation
of hydrido aluminate species at the NP surface should take place, which could then
transform into tricoordinated surface alkyl aluminum species. Formation of surface
alkyl radicals may also be envisaged, in agreement with the short distance observed
by EXAFS. Hence, the surface of the NPs should then be regarded as electronically
depleted. As is well known, electron-withdrawing ligands such as carbon monoxide
deprive the surface atoms of their magnetic properties [108]. In the assumption of
Amiens et al., a similar phenomenon takes place in sample 1 and accounts for the
observed low magnetization.
The loss in magnetization corresponds to a partial quenching of the magnetic
contribution of the surface. The fact that the magnetization of the surface layer is
71
3.4
72
(a)
Br
(b)
N
=N= N
=N= N
(c) =
R
N
=N= N
=
R
O
Figure 3.15 (a) Br(CH2)11SH in DCM, 60 hours at room temperature; (b) 0.25 M NaN3 in
DCM/DMSO solution, 48 hours; and (c) R = propyn-1-one derived compounds as in scheme 2, 24
to 96 hours in dioxane or 1:1 hexane/dioxane. (From: [119]. 2006 American Chemical Society.
Reprinted with permission.)
73
48 hours, they added water and the black organic layer was isolated. They performed reaction of azide-functionalized Au nanoparticles with alkynyl derivatives
using a procedure that can be described as follows: N3-functionalized particles and
the alkynyl-modified compound (Figure 3.16) were codissolved in dioxane or hexane/dioxane and stirred for 24 to 96 hours. Under vacuum, the solvent was removed
and ethanol was utilized to remove any unreacted alkynyl derivative. Finally, the
particles were dried and then redissolved in DCM. Additionally, Au nanoparticle
samples could be decomposed using standard disulfide-forming reactions. For testing the functionalization approach, reactions involving triazole ring formation were
conducted using small Au particles.
Because its length sufficiently stabilizes the particles from aggregation but is not
so sterically hindered as to prevent ligand exchange, decanethiol served as the surface ligand in the synthesis of monolayer-protected Au clusters.
One can see in the representative TEM image in Figure 3.17 that the Au nanoparticles are spherical and have an average diameter of about 1.8 nm. For the purpose of replacing a fraction of these ligands with Br-terminated undecanethiol
ligands, the decanethiol-stabilized Au particles were then stirred in a solution containing BrC11H22-SH. By means of NMR and FTIR spectra, the replacement was
confirmed. In the reaction of the Br termini, in the way that nucleophilic substitution with NaN3 was used to append azide functionalities to the Au nanoparticles, as
shown in Figure 3.17(b), the size and shape of the resultant particles were not
affected. According to the FTIR and NMR spectra, the ligand replacement with
BrC11H22-SH and subsequent reaction with NaN3 and the resultant Au
nanoparticles containing mixed monolayers that are 44% CH3- and 52% N3-terminated alkanethiol ligands were confirmed.
In Williamss work, they also conducted the triazole functionalization of Au
nanoparticles, and further particle functionalization through 1,3-dipolar cycloaddition reactions (i.e., click chemistry), by fusing ethynyl- and azide-bearing
O
O
Fe
NO2
Fc
NB
Pyr
O
O
CH3
3
HN
O
An
PEG
Ani
Figure 3.16 Scheme 2. Propyn-1-one compounds for attachment via triazole ring formation.
(From: [119]. 2006 American Chemical Society. Reprinted with permission.)
74
# of Particles
160
120
80
40
0
0
# of Particles
50
40
30
20
10
0
0 1 2 3 4 5 6
Size (nm)
(a)
2
4
Size (nm)
(b)
Figure 3.17 (a) Representative TEM images of synthesized C10H21SH-modified Au particles and (b)
azide-functionalized Au nanoparticles. Insets contain particle size distributions. The scale bar is 50
nm. (From: [119]. 2006 American Chemical Society. Reprinted with permission.)
75
76
3.4.1.2
Ag Nanoparticles
First, we introduce the useful technology that utilizes localized surface plasmon resonance (LSPR) nanosensors in detecting biological molecules. This novel, nanoscale
development is a refractive indexbased sensing device that relies on the extraordinary optical properties of noble (Ag, Au, Cu) metal nanoparticles [131135].
Briefly, it is based on the nanoscale limit of surface plasmon resonance sensors. In
other words, LSPR, which refers to the ability of the conduction electrons in the
nanoparticle to oscillate collectively [136], induces electromagnetic fields surrounding the nanoparticle, which determine the sensing volume in which refractive
indexbased sensing can occur [135, 137]. Nanoparticles exhibit selective photon
absorption, which can easily be monitored using ultraviolet-visible, since it is well
known that the conduction electrons oscillate collectively to only specific wavelengths of light (UV-visible) spectroscopy. It is well established that the maximum
extinction wavelength max of the LSPR is dependent upon the composition, size,
shape, and interparticle spacing of the nanoparticles. And the dielectric properties of
their local environment (i.e., substrate, solvent, and surface-confined molecules)
[135, 136] also play an important role.
The first nonmodel application of the LSPR nanosensor was reported by Haes et
al. in 2004 [135]. In their work, the LSPR nanosensor underlain by the optical properties of Ag nanotriangles was shown to aid in the understanding of the interaction
between amyloid derived diffusible ligands (ADDL) and the anti-ADDL antibody,
molecules possibly involved in the development of Alzheimers disease. By varying
the concentration of anti-ADDL antibody, a surface-confined binding constant of
3.0 107 M1 for the interaction of ADDLs and anti-ADDLs was measured. Influences of Cr, the nanoparticle adhesion layer, will be shown to be the limiting factor
in the sensitivity of this assay.
In fact, in 2002, they presented the work on the employment of Ag nanoparticles
together with Au nanoparticles in the LSPR procedure. Briefly, it is based on the fact
that triangular silver nanoparticles (~100 nm wide and 50 nm high) have remarkable optical properties and the peak extinction wavelength max of their LSPR spectrum is unexpectedly sensitive to nanoparticle size, shape, and local (~10 to 30 nm)
external dielectric environment. This sensitivity of the LSPR max to the
nanoenvironment has provided the opportunity for developing a new class of
nanoscale affinity biosensors. Utilizing the well-studied biotin-streptavidin system,
the essential characteristics and operational principles of these LSPR nanobiosensors were demonstrated. In this work, a 27.0-nm redshift in the LSPR max
occurred, after the exposure of biotin-functionalized Ag nanotriangles to 100-nM
streptavidin (SA). They measured the LSPR max shift, R/Rmax, versus the [SA]
15
6
response curve over the concentration range 10 < [S] < 10 M. By comparing
the data with the theoretical normalized response expected for 1:1 binding of a
ligand to a multivalent receptor with different sites but invariant affinities, one got
approximate values for the saturation response. The values are Rmax = 26.5 nm
together with the surface-confined thermodynamic binding constant Ka,surf = 1011
1
M for the actual limit of detection (LOD) for the LSPR nanobiosensor, which is in
the low-picomolar to high-femtomolar region. Hence, a strategy for amplifying the
response of the LSPR nanobiosensor with employment of biotinylated Au colloids
and thereby further improving the LOD was carried out. Several control
77
experiments were conducted to define the LSPR nanobiosensors response to nonspecific binding as well as to demonstrate its response to the specific binding of
another protein. They then drew the conclusion that the factors determine the LSPR
procedure could be summarized as follows:
1. Electrostatic binding of SA to a nonbiotinylated surface;
2. Nonspecific interactions of prebiotinylated SA to a biotinylated surface;
3. Nonspecific interactions of bovine serum albumin to a biotinylated surface;
4. Specific binding of antibiotin to a biotinylated surface.
The LSPR nanobiosensor provides a pathway to ultrasensitive biodetection experiments with extremely simple, small, light, robust, and low-cost instrumentation
that will greatly facilitate field-portable environmental or point-of-service medical
diagnostic applications.
In 2003, Riboh et al. published their sequential results in the Journal of Physical Chemistry B [134]. In the later work, the width of the silver nanoparticles was
narrowed to 90 nm. The target was the probe of the interaction between a surfaceconfined antigen, biotin (B), a solution-phase antibody, and antibiotin (AB). An
about 38-nm redshift in the LSPR max took place after exposure of biotin7
6
functionalized Ag nanotriangles to 7 10 M < [AB] < 7 10 M. The experimental normalized response of the LSPR max shift, (R/Rmax), versus [AB] was mea10
6
sured over the concentration range 7 10 M < [AB] < 7 10 M. The similar
comparison as mentioned above led to values for the saturation response, Rmax =
7
38.0 nm, the surface-confined thermodynamic binding constant, Ka,surf = 4.5 10
1
10
M , and the limit of detection (LOD) < 7 10 M. The experimental saturation
response was explained in terms of a closest packed structural model for the surface
B-AB complex in which the long axis of AB, lAB = 15 nm, is oriented horizontally
and the short axis, hAB = 4 nm, is oriented vertically to the nanoparticle surface.
Based on this model, a quantitative response for the saturation response, Rmax =
40.6 nm, is obtained, which is in good agreement with experiment, Rmax = 38.0
nm. By means of atomic force microscopy (AFM), this interpretation was confirmed. Furthermore, they also described the major improvements in the LSPR
nanobiosensor. A series of LSPR nanobiosensor substrates was utilized such as
glass, mica, and a surfactant, Triton X-100, in the nanosphere lithography fabrication procedure. These changes increased the adhesion of the Ag nanotriangles by a
factor of 9 as determined by AFM normal force studies. The improved adhesion of
Ag nanotriangles allows the study of the B-AB immunoassay in a physiologically
relevant fluid environment as well as in real time. These results represent important
new steps in the development of the LSPR nanosensor for applications in medical
diagnostics, biomedical research, and environmental science.
In addition, the Ag nanoparticles can form the nanoscale bioconjugated superstructures together with semiconductor one-dimensional nanostructures, and the
system shows a twofold enhancement of luminescence intensity. Theoretical calculations of the electric field in the cylindrically organized NPs and experimental data
indicate that the enhancement in emission is based on the increase in absorption of
the Ag NP shells in the regime of the collective plasmon resonance. This situation is
78
3.4.2
79
Antibody
Tumor
Before
10 min
6h
10 h
24 h
Figure 3.18 Sketch map of the magnetic nanoparticles serving as biodetection agents. (Courtesy
of Professor Mingyuan Gao, Institut of Chemistry, Chinese Academy of Science.)
80
d
25ms
Figure 3.19 Left: Photograph of 5% polyacrylamide separating gel stained by coomassie blue.
The four lanes from left to right are filled with Fe3O4-(rch 24 mAb) conjugates, a mixture of Fe3O4
and rch 24 mAb, Fe3O4 nanocrystals, and rch 24 mAb, respectively. Right: T2-weighted MR images
of CEA-positive LS 180 cells treated with (a) Fe3O4-(rch 24 mAb) conjugates, (b) Fe3O4-irrelevant
antibody conjugates, and (c) Fe3O4@PEG-COOH nanocrystals. (d) Untreated LS 180 cells shown for
comparison. The echo time used to acquire the images was (left to right) 25, 75, 125, and 175 ms.
(Courtesy of Professor Mingyuan Gao, Institute of Chemistry, Chinese Academy of Science.)
81
Based on the success of in vitro experiments, the rch 24 mAb conjugates were
used in the subsequent in vivo experiments aiming at detecting human colon carcinoma xenograft tumors implanted in nude mice at their proximal thigh region.
The conjugates were injected through a tail vein into a group of mice carrying
tumors, and the magnetic nanocrystals were intravenously injected into another
group of tumor-bearing, randomly selected mice. T2- and T2*-weighted MR
images acquired before and at different times after injection are shown in Figure
3.20. One can see that part of the tumor exhibits hypointensity 24 hours after the
injection of the conjugates; hence, a successful bond of magnetic nanocrystals to the
tumor via rch 24 mAb was established. More detailed evolvement with time has
been studied.
The T2 values gradually decreased and then reached a maximum of approximately 10% 24 hours after injection. In principle, these results suggest that the signal decrease should be caused mainly by the enrichment of the magnetic nanocrystals on the cancer cells mediated by rch 24 mAb. One can find the inhomogeneous distribution of the Fe3O4 nanocrystals, which was observed by Huh et al.
[153]. Their work is worth mentioning. In 2005, they reported an in vivo diagnosis
of cancer, with employment of a well-defined magnetic nanocrystal probe system
with multiple capabilities, such as small size, strong magnetism, high
biocompatibility, and the possession of active functionality for desired receptors.
The magnetic nanocrystals were conjugated to a cancer-targeting antibody, herceptin. The subsequent utilization of these conjugates as MRI probes was successfully
demonstrated for the monitoring in vivo of selective targeting events of human cancer cells implanted in live mice. Further conjugation of these nanocrystal probes
with fluorescent dyelabeled antibodies enables both in vitro and ex vivo optical
detection of cancer as well as in vivo MRI, which are potentially applicable for an
Before
24 h
Before
10 min
6h
10 h
24 h
(b)
(a)
(c)
Figure 3.20 (a) T2-weighted (upper row) and T2*-weighted (lower row) MR images acquired
before and after the injection of Fe3O4-(rch 24 mAb) conjugates. (b) T2*-weighted MR images
acquired before and at different times after injection of Fe3O4-(rch 24 mAb) conjugates (upper row)
and Fe3O4@PEG-COOH nanocrystals (lower row). (c) Histochemical analysis of a tumor tissue slice.
(Courtesy of Professor Mingyuan Gao, Institute of Chemistry, Chinese Academy of Science.)
82
83
Retention of HPPH, %
100
80
60
40
20
0
0
10
20
30
40
Time, hrs
50
60
70
Figure 3.21 The retention of HPPH in PEG-2000-PE magnetic polymeric micelles (10% magnetic
material) in PBS, pH = 7.4 at 37C. (From: [160]. 2006 American Chemical Society. Reprinted
with permission.)
laser scanning microscopy. They claimed that the loading efficiency of HPPH was
almost unaffected upon coloading with the magnetic nanoparticles, and its phototoxicity was retained. The magnetic response of the nanocarriers was demonstrated
by their magnetically directed delivery to tumor cells in vitro. The magnetophoretic
control on the cellular uptake provides enhanced imaging and phototoxicity.
Figure 3.22 indicates that the cells took up the HPPH entrapped in PEG-PE
micelles successfully. Localization of HPPH inside the cells [Figure 3.22(c)] is similar
to that of fluorescent micelles [Figure 3.22(b)], indicating that PEG-PE micelles keep
HPPH in cells after the treatment with a duration of 90 minutes. Hence, the similarity
of the distribution patterns obtained for fluorescence from fluorescein-labeled
micelles and those from encapsulated HPPH gives the opportunity to draw the conclusion that micellar nanocarriers sustained enough stability to retain HPPH during
incubation with cells in medium and after cellular uptake. In addition, the cells were
found to be in good condition even after staining overnight. In the same report, one
can see the remarkable achievement in magnetophoretic control and guidance of
cellular uptake by an external magnetic field [160].
3.5
84
(a)
(b)
(c)
Figure 3.22 Confocal microscopy images of Hela cells after 90 minutes of incubation with fluorescent PEG-2000-PE micelles loaded with HPPH: (a) transmission, (b) fluorescence from
fluorescein-labeled micelles, and (c) fluorescence from HPPH. (From: [160]. 2006 American
Chemical Society. Reprinted with permission.)
and Onushchenko [161] and Efros and Efros [162] in 1982. The physical properties
and applications of QDs have been extensively investigated in many physics and
engineering laboratories since the early 1980s [163165]. The size and shape can be
precisely controlled by duration, temperature, and the ligant molecules used in the
synthesis [166]. At these small sizes, materials behave differently, giving quantum
dots unprecedented properties and enabling never before seen applications in science and technology.
The optical and electronic qualities of traditional semiconductors are difficult to
adjust because their bandgap is fixed, and as a consequence they have fixed emission
frequencies. In contrast with conventional semiconductors, QDs have the ability to
tune the bandgap and hence the emission wavelength. The tunable bandgap of QDs
derives from the physical dimension of QDs, which by definition is smaller than the
exciton Bohr radius. In this case, the electron energy levels must be treated as discrete, and this gives rise to a unique phenomenon known as quantum confinement.
Quantum confinement, which refers to the spatial confinement of charge carriers
(i.e., electrons and holes) within a material, gives QDs unique optical and electronic
properties not observed in larger crystallites. With QDs, the size of the bandgap is
controlled simply by adjusting the size of the dot, and therefore it is possible to control the output wavelength of a dot with extreme precision. They exhibit a sizedependent color, as shown in Figure 3.23(a). However, it has been demonstrated
that the size differences of multicolor QDs could be addressed by using alloyed semiconductor QDs, where tuning of the emission wavelength is achieved via nanocrystal composition rather than size [Figure 3.23(b)].
QDs display advantages in their absorptive properties too, since they display
tunable absorption patterns. The absorption spectrum for quantum dots appears as
a series of overlapping peaks that get larger at shorter wavelengths. The dots will
absorb light that has a wavelength shorter than that of the first exciton peak, also
referred to as the absorption onset. The wavelength of the first exciton peak (and all
subsequent peaks) is a function of the composition and size of the dot.
In QDs the emission pattern is also tunable. The peak emission wavelength has a
Gaussian shape and occurs at a slightly longer wavelength than the absorption
85
onset. This energy separation is what is referred to as the Stokes shift. An interesting
characteristics of QDs is that the peak emission wavelength is independent of the
wavelength of the excitation light, assuming that it is shorter that the wavelength of
the absorption onset. Many sizes of nanocrystals may therefore be excited with a
single wavelength of light, resulting in many colors that may be detected simultaneously. The absorption onset and the emission maximum shift to higher energy
with decreasing size of nanocrystals [166]. By varying the material used and different sizes for the nanocrystals, a spectral range from 400 nm to 2 m for emission
was obtained [167170]. The bandwidth of the emission spectra denoted as the full
width at half maximum (FWHM) has typical values of 20 to 30 ns in the visible
region of the spectrum and large extinction coefficients in the visible and ultraviolet
range. Narrower size distributions yield smaller FWHM. For CdSe a 5% size distribution corresponds to ~30-nm FWHM while in PbSe a 5% size distribution corresponds to ~100-nm FWHM [171].
The ratio of absorbed and emitted photons is called quantum yield (QY). QY is
controlled by the existence of nonradiative transitions of electrons and holes
between energy levelstransitions that produce no electromagnetic radiation.
Nonradiative recombination largely occurs at the dots surface and is therefore
greatly influenced by the surface chemistry. Passivation is the chemical process by
which the surface atoms of a QD are bonded to another material of much larger
bandgap. These (core) shell structures achieved efficient surface passivation as well
as carrier confinement within the QD core, which decreased the likelihood of
recombination via nonradiative pathways involving surface electronic states. In this
way QDs are coated with several atomic layers of an inorganic wideband semiconductor, and this results in a significant increase in the QY [172174]. The shell can
be designed carefully to obtain a QY close to 90% [175]. By bonding appropriate
molecules to this shell surface like thiol, amine, nitrile, phosphonic acid, carboxylic
acid, the QDs can be dispersed or dissolved in nearly any solvent or incorporated
into a variety of inorganic and organic films. In addition to the surface chemistry
Quantum dot composition
2.1 nm
400 450
3.2 nm
7.5 nm
(a)
Fluorescence (a.u.)
Fluorescence (a.u.)
700
750
CdSe
550 600
CdTe
CdSexTe1-x
900
(b)
Figure 3.23 QD emission wavelength, tunable by changing the nanoparticle size and composition. (a) The emission spectrum of a CdSe QD may be adjusted within the visible spectrum (450 to
650 nm) by selecting the QD size between 2 and 7.5 nm. The relative sizes of these particles are
shown schematically below the fluorescence spectrum. (b) While keeping the QD size constant
(5-nm diameter) and varying the composition of the ternary alloy CdSexTex, the maximum emission wavelength may be tuned between 610 and 800 nm. (From: [170]. 2004 Elsevier. Reprinted
with permission.)
86
molecular coupling can be used to effectively alter the properties of the quantum dot
including brightness and electronic lifetime. Of the semiconductor families investigated to date, IIVI materials have shown the most promise, and consequently their
use in biological applications has predominated [170].
3.6
Traditionally, the favored materials used to label microscopic structures have been
organic dyes, which can be chemically engineered to adhere to a diverse variety of
cellular structures. After the dye comes into contact with the proper cellular structures, they become fluorescent if excited with a certain wavelength of light. The
emitted radiation has a peak wavelength controlled by the chemical nature of the
organic dye being used. Unfortunately, many disadvantages exist with this technique, most of them a result of the extremely limited absorptive and emissive capabilities of organic dyes. For example, the peak emission of organic dyes is fixed, with
each dye having a different preset emission wavelength (color) that is set by its
nature. Therefore, applications that make use of light frequencies that do not correspond to the emission peaks of preexisting organic dyes cannot be performed.
Another disadvantage is the narrow absorption spectrum of organic dyes. Dyes have
uneven absorption and emission peaks. The shape of their emission and absorption
peaks shows shoulders, which is a major disadvantage in applications that require
a Gaussian-type emission spectrum. They tend to display absorption peaks that are
87
Biological coating
Core
Figure 3.24
Shell
not always in convenient regions of the spectrum, making the excitation of various
organic dyes challenging and costly. The last disadvantage is that of stability. The
lifetime of organic dyes varies but is generally low relative to that of other tagging
mechanisms.
The fact that cell autofluorescence is in the visible spectrum and can mask signals from labeled molecules with dyes and the requirement of long observation
times have created the need to look for other fluorescent biological labels. The QDs
are nanometer-sized semiconductor nanocrystals with superior fluorescent properties. In comparison with organic dyes, which have fixed emissions, QDs can be
made to emit light at any wavelength in the visible and infrared ranges and can be
inserted almost anywhere, including liquid solution, dyes, paints, epoxies, and
sol-gels. QDs can be attached to a variety of surface ligands and inserted into a variety of organisms for in vivo research. They have unique technical capabilities to
enable all new standards for fluorescent tagging, eliminating the shortcomings of
dyes.
3.6.2
QDs have distinctive optical and spectral properties that provide unique properties
and benefits for a rich variety of biomedical life science applications. Bruchez et al.
[178] and Chan and Nie [179] and their coworkers first highlighted the great promise of QDs in biological applications:
Higher brightness As biological probes, QDs are extremely bright, with large
quantum yields. Studies comparing the brightness of single CdSe/ZnS
core/shell QDs to that of single rhodamine 6G molecules indicate that the fluorescent photon flux in QDs is tenfold to a hundredfold greater.
Broadband absorption. In contrast to organic fluorophores that have narrow
absorption spectra, QDs have large absorption cross sections over a wide
bandwidth. They display the unique property that all light that has a shorter
wavelength than the emission wavelength can be absorbed, with increasing
strength at shorter wavelengths. Therefore, the peak emission wavelength is
88
effectively independent of the excitation source. The effective Stokes shift can
be tens to hundreds of nanometers. In the visible range of the spectrum, the
peak emission wavelength is shifted from the absorption onset by 15 nm.
Tunable and narrow emission bandwidth. QDs have peak emission wavelengths dependent upon their composition and size. Currently, peak emission
wavelengths are available from 490 to 620 nm in 20-nm increments. Peak
emission wavelengths from 350 to 480, 640 to 720, and 800 to 2,300 nm have
been demonstrated [171]. QDs emit light within a narrow Gaussian spectrum
without any shoulders and centered at the peak wavelength.
Higher resistance to photo bleaching. QDs are based on inorganic particles
that are inherently more photostable than organic molecules, and as such they
can survive orders of magnitude longer than organic fluorescent dyes under
intense illumination.
Long fluorescence lifetime. The fluorescent lifetimes (electronic lifetime) of
visibly emitting quantum dots have been measured to be 15 to 20 ns, independent of the emission wavelength. The fluorescence lifetime is orders of
magnitude longer than typical autofluoresence lifetimes and many multiples
of typical organic dye lifetimes.
Large two-photon absorption. QDs have large two-photon absorption cross
sections that allow for narrowband visible light to be emitted when
long-wavelength IR lasers are focused on the quantum dots.
Improved detection sensitivity. The optical properties of QDs lead to
improved detection sensitivity and simplification in experiment and instrument design. Due to their inorganic semiconductor-based composition, they
can be also be visualized with electron microscopes.
3.7
350
89
Absorbance of fluorescence (a.u.)
FITC
400
450
500
550 600
Wavelength (nm)
650
350
400
450
500
550
Wavelength (nm)
600
650
Figure 3.25 Absorption and fluorescence emission spectra of an organic dye (fluorescein
isothiocyanate, FITC, left) and a CdSe QD (right) with identical emission wavelengths. The relative
sizes of the dye and nanoparticle are also shown. (From: [170]. 2004 Elsevier. Reprinted with
permission.)
overlap between the quantum dot emission wavelength and the absorption spectra
of the dyes. Because the emission characteristics of the QDs can be continuously
tuned, it is possible to create a FRET donor for any number of organic dyes that
emit between approximately 510 and 640 nm [171, 187]. QDs can act as highly efficient donors with other acceptors as other QDs and metallic nanoparticles [188].
The high photostability of QDs relative to organic dyes allows for long-term
tracking of biological processes. Due to the unprecedented photostability and color
multiplexing properties, continuous observations using fluorescence and confocal
microscopes of multiple cellular components and cellular metabolism/transport can
be made over a period of hours to days [189, 190]. An impressive demonstration of
this capability was illustrated by microinjecting QD bioconjugates into Xenopus
embryos and monitoring their partitioning into various cells during tadpole development over a period of several days [191].
The long-term stability and brightness of QDs make them ideal candidates for
live cell and animal targeting and imaging.
Water-stabilized quantum dots have been introduced into large multicellular
animals (mice) to preferentially stain vascular and lymphatic systems tumors, showing that higher contrast and imaging depth can be obtained at a lower excitation
power than with organic dyes. In [192], imaging of targeted QD delivery in live
mice was achieved by intravenously injecting them with QDs functionalized with
antibodies to prostate-specific membrane antigen. Many studies have shown the
great potential of using quantum dots as new probes in vivo [193, 194].
QDs have been shown to operate well in Western blot assays and can serve to
replace the radioactive labels and expensive detection equipment that are currently
in use. Conjugated to secondary antibodies, they have been successful applied to
protein electrophoretic gels that were prepared with primary antibodies. QD-based
Western blot technology was used for ultrasensitive detection of tracer proteins
[195].
QD fluor dyes are ideally suited for flow cytometry applications as recorders
and standards. For example, QDs 525 and 605 are useful for analyzing antigenic
expression. Flow cytometry (FCM), which is well adapted to detect fluorescence
90
91
References
[1] Trindade, T., P. OBrien, and N. L. Pickett, Nanocrystalline Semiconductors: Synthesis,
Properties, and Perspectives, Chem. Mater., Vol. 13, 2001, p. 3843.
[2] Burda, C., et al., Chemistry and Properties of Nanocrystals of Different Shapes, Chem.
Rev., Vol. 105, 2005, p. 1025.
[3] Bohren, C. F., and D. R. Huffman, Absorption and Scattering of Light by Small Particles,
New York: Wiley, 1983.
[4] Creighton, J. A., and D. G. Eadon, New Mixed Quantum/Semiclassical Propagation
Method, J. Chem. Soc. Faraday Trans., Vol. 87, 1991, p. 3881.
[5] Ostwald, W., Theoretical and Applied Colloid Chemisty, New York: Wiley, 1922, p. 200.
[6] Kreibig, U., and L. Genzel, Optical Absorption of Small Metallic Particles, Surf. Sci., Vol.
156, 1985, p. 678.
[7] Fragstein, C. V., and U. Z. Kreibig, Studies of Intersystem Crossing Dynamics in Acetylene, Phys., Vol. 224, 1969, p. 306.
[8] Doremus, R. H., Optical Properties of Small Silver Particle, J. Chem. Phys., Vol. 42,
1965, p. 414.
[9] Doyle, W. T., Absorption of Light by Colloids in Alkali Halide Crystals, Phys. Rev., Vol.
B111, 1958, p. 1067.
[10] Huffman, D. R., in An Introduction to Physiscs and Technology of Thin Films,
Wagendristerl, A., and Wang, Y., (eds.), New Jersey: World Scientific, 1988.
[11] Ralph, D. C., C. T. Black, and M. Tinkham, Spectroscopy of the Superconducting Gap in
Individual Nanometer-Scale Aluminum Particles, M. Phys. Rev. Lett., Vol. 74, 1995, pp.
32413244.
[12] Peng, X., J. Wickham, and A. P. Alivisatos, Kinetics of II-VI and III-V Colloidal Semiconductor Nanocrystal Growth: Focusing of Size Distributions, J. Am. Chem. Soc., Vol.
120, 1998, p. 5343.
[13] Murray, C. B., C. R. Kagan, and M. G. Bawendi, The Width of the Absorption Band,
Annu. Rev. Mater. Sci.,Vol. 30, 2000, p. 545.
[14] Pamplin, B. R., Crystal Growth, New York: Pergamon Press, 1975.
[15] Jiang, Y., Forced Hydrolysis and Chemical Co-Precipitation, in Handbook of
Nanophase and Nanostructured Materials, Z. L.Wang, Y. Liu, and Z. Zhang, (eds.), New
York: Kluwer Academic, 2003, p. 59.
[16] Burda, C., et al., Chemistry and Properties of Nanocrystals of Different Shapes, Chem.
Rev., Vol. 105, 2005, p. 1025.
[17] Alivisatos, A. P., Naturally Aliened Nanoparticles, Science, Vol. 289, 2000, p. 736.
[18] Hayat, M. A., (ed.), Colloidal Gold: Principles, Methods and Applications, San Diego, CA:
Academic Press, Vols. 1 and 2, 1989.
[19] Brust, M., et al., Ultrafast Optical Kerr Effect Spectroscopy of Water Confined in
Nanopores of the Gelatin Gel, Chem. Soc. Chem. Commun., 1994, p. 801.
[20] Johnson, S. R., et al., Alkanethiol Molecules Containing an Aromatic Moiety Self-Assembled onto Gold Clusters, Langmuir, Vol. 13, 1997, p. 51.
[21] Hostetler, M. J., et al., Monolayers in Three Dimensions: Synthesis and Electrochemistry
of Functionalized Alkanethiolate-Stabilized Gold Cluster Compounds, Am. Chem. Soc.,
Vol. 118, 1996, p. 4212.
[22] Ingram, R. S., M. J. Hostetler, and R. W. J. Murray, Poly-HeteroFunctionalized
Alkanethiolate-Stabilized Gold Cluster Compounds, Am. Chem. Soc., Vol. 119, 1997, p.
9175.
[23] Buining, P. A., et al., Preparation of Functional Silane-Stabilized Gold Colloids in the
(Sub)nanometer Size Range, Langmuir, Vol. 13, 1997, p. 3921.
92
93
[47] Kitakami, O., et al., Size Effect on the Crystal Phase of Cobalt Fine Particles, Phys. Rev.
B, Vol. 56, 1997, p. 13849.
[48] Hull, A. W., X-Ray Crystal Analysis of Thirteen Common Metals, Phys. Rev., Vol. 17,
1921, p. 571.
[49] Krainer, E., and J. Robitsch, New Diffraction in Special Bulk Cobalt Surfaces, 2
Metal//kd, Vol. 65, 1974, p. 729.
[50] Kajivara, S., et al., Computational Study of the Melting-Freezing Transition in the Quantum Hard-Sphere System, Phil. Mag. Lett., Vol. 55, 1987, p. 215.
[51] Leslie-Pelecky, D. L., et al., Using High-Temperature Chemical Synthesis to Produce
Metastable Nanostructured Cobalt, Chem. Mater., Vol. 10, 1998, p. 3732.
[52] Respaud, M., et al., Surface Effects on the Magnetic Properties of Ultrafine Cobalt Particles, Phys. Rev. B, Vol. 57, 1998, p. 2925.
[53] Dinega, D. P., and M. G. Bawendi, A Solution-Phase Chemical Approach to a New Crystal Structure of Cobalt, Angew. Chem. Int. Ed., Vol. 38, No. 12, 1999, pp. 17881791.
[54] Sun, S., and C. B. Murray, Synthesis of Monodisperse Cobalt Nanocrystals and Their
Assembly into Magnetic Superlattices, J. Appl. Phys., Vol. 85, 1999, p. 4325.
[55] Gomez, S., et al., Water-Gas-Shift Reaction on Metal Nanoparticles and Surfaces, Chem.
Commun., 2001, p. 1474.
[56] Gomez, S., et al., Theoretical Study of the Electronic Structure of Silver Nanoparticles,
Chem. Commun., 2000, p. 1945.
[57] Dinega, D. P., and M. G. Bawendi, A Solution-Phase Chemical Approach to a New Crystal Structure of Cobalt, Angew. Chem. Int. Ed., Vol. 38, 1999, p. 1788.
[58] Zhang, P., and T. K. Sham, Tuning the Electronic Behavior of Au Nanoparticles with Capping Molecules, Appl. Phys. Lett., Vol. 81, 2002, p. 736.
[59] Selwood, P. W., Chemisorption and Magnetization, New York: Academic Press, 1975.
[60] Puntes, V. F., et al., Synthesis of hcp-Co Nanodisks, J. Am. Chem. Soc., Vol. 124, 2002,
p. 12874.
[61] Bradley, J. S., et al., Surface Chemistry on Colloidal Metals: Spectroscopic Study of
Adsorption of Small Molecules, Faraday Discuss., Vol. 92, 1991, p. 255.
[62] Pan, C., et al., Ligand-Stabilized Ruthenium Nanoparticles: Synthesis, Organization, and
Dynamics, J. Am. Chem. Soc., Vol. 123, 2001, p. 7584.
[63] Choukroun, R., et al., H2-Induced Structural Evolution in Non-Crystalline Rhodium
Nanoparticles, New J. Chem., Vol. 25, 2001, p. 525.
[64] Rodriguez, A., et al., Synthesis and Isolation of Cuboctahedral and Icosahedral Platinum
Nanoparticles. Ligand-Dependent Structures, Chem. Mater., Vol. 8, 1996, p. 1978.
[65] Osuna, J., et al., Chemical Control of Structural and Magnetic Properties of Cobalt
Nanoparticles, Chem. Mater., Vol. 17, 2005, p. 107.
[66] Pery, T., et al., Direct NMR Evidence for the Presence of Mobile Surface Hydrides on
Ruthenium Nanoparticles, Angew. Chem., Int. Ed., Vol. 6, 2005, p. 605.
[67] Cordente, N., et al., Synthesis and Magnetic Properties of Nickel Nanorods, Nano Lett.,
Vol. 1, 2001, p. 565.
[68] Bnnemann, H., et al., A Size-Selective Synthesis of Air Stable Colloidal Magnetic Cobalt
Nanoparticles, Inorg. Chim. Acta, Vol. 350, 2003, p. 617.
[69] Dassenoy, F., et al., Experimental Evidence of Structural Evolution in Ultrafine Cobalt
Particles Stabilized in Different PolymersFrom a Polytetrahedral Arrangement to the
Hexagonal Structure, J. Chem. Phys., Vol. 112, 2000, p. 8137.
[70] Petit, C., A. Taleb, and M.-P. Pileni, Self-Organization of Magnetic Nanosized Cobalt
Particles, Adv. Mater., Vol. 10, 1998, p. 259.
[71] Andersen, R. A., et al., Synthesis of bis[bis(trimethylsilyl)amido]iron(II). Structure and
Bonding in M[N(SiMe3)2]2 (M=Manganese, Iron, Cobalt): Two-Coordinate Transition-Metal Amides, Inorg. Chem., Vol. 27, 1988, p. 1782.
94
95
[96] Ostwald, W., Theoretical and Applied Colloid Chemisty, New York: Wiley, 1922, 200.
[97] Johnson, P. B., and R. W. Christy, Monodispersed Particles of Gold, Phys. Rev., 1972,
Vol. B6, 1972, pp. 43704379.
[98] Ziman, J. M., Theory of Solids, Cambridge, U.K.: Cambridge University Press, 1979, p.
278.
[99] Fang, M., et al., Magnetic Bio/Nanoreactor with Multilayer Shells of Glucose Oxidase
and Inorganic Nanoparticles, Langmuir, Vol. 18, 2002, p. 6338.
[100] Hovel, H. S., et al., Raman Scattering of 4-Aminobenzenethiol Sandwiched between
Ag/Au Nanoparticle and Macroscopically Smooth Au Substrate, Phys. Rev., Vol. B48,
1993, p. 18178.
[101] Henglein, A., Physicochemical Properties of Small Metal Particles in Solution:
Microelectrode Reactions, Chemisorption, Composite Metal Particles, and the
Atom-to-Metal Transition, J. Phys. Chem., Vol. 97, 1993, p. 5457.
[102] Camillone III, et al., Substrate Dependence of the Surface Structure and Chain Packing of
Docosyl Mercaptan Self-Assembled on the (111), (110), and (100) Faces of Single Crystal
Gold, J. Chem. Phys., Vol. 98, 1993, p. 4234.
[103] Linnert, T., P. Mulvaney, and A. J. Henglein, Surface Chemistry of Colloidal Silver: Surface Plasmon Damping by Chemisorbed Iodide, Hydrosulfide (SH-), and Phenylthiolate,
Phys. Chem., Vol. 97, 1993, p. 679.
[104] Kreibig, U., et al., in Large Clusters of Atoms and Molecules, T. P. Martin, (ed.), Dordrecht:
Kluwer, 1996, p. 477.
[105] Cleveland, C. L., et al., Structural Evolution of Larger Gold Clusters, Phys. D, Vol. 47,
1999, p. 468.
[106] Leslie-Pelecky, D. L., and R. D. Rieke, Magnetic Properties of Nanostructured Materials, Chem. Mater., Vol. 8, 1996, p. 1770.
[107] Selwood, P. W., Chemisorption and Magnetization, New York: Academic Press, 1975.
[108] Ruitenbeek, J. M. V., D. A. V. Leeuven, and L. J. D. Jongh, Magnetic Properties of Metal
Cluster Compounds, in Physics and Chemistry of Metal Cluster Compounds, L. J. D.
Jongh, (ed.), Dordrecht: Kluwer Academic, 1994, p. 277.
[109] Haes, A. J., et al., Detection of a Biomarker for Alzheimers Disease from Synthetic and
Clinical Samples Using a Nanoscale Optical Biosensor, J. Am. Chem. Soc., Vol. 127,
2005, p. 2264.
[110] Kohler, N., G. E. Fryxell, and M. Zhang, A Bifunctional Poly(ethylene glycol) Silane
Immobilized on Metallic Oxide-Based Nanoparticles for Conjugation with Cell Targeting
Agents, J. Am. Chem. Soc., Vol. 126, 2004, p. 7206.
[111] Hu, A., G. T. Yee, and W. U. Lin, Magnetically Recoverable Chiral Catalysts Immobilized
on Magnetite Nanoparticles for Asymmetric Hydrogenation of Aromatic Ketones, J. Am.
Chem. Soc., Vol. 127, 2005, p. 12486.
[112] Roucoux, A., J. Schulz, and H. Patin, Reduced Transition Metal Colloids: A Novel Family
of Reusable Catalysts? Chem. Rev., 2002, Vol. 102, p. 3757.
[113] Ingram, R. S., M. Hostetler, and R. W. Murray, Poly-HeteroFunctionalized
Alkanethiolate-Stabilized Gold Cluster Compounds, J. Am. Chem. Soc., Vol. 119, 1997,
p. 9175.
[114] Morrison, R. T., and R. N. Boyd, Organic Chemistry, 6th ed., Englewood Cliffs, NJ:
Prentice-Hall, 1992.
[115] Carey, F. A., and R. J. Sundberg, Advanced Organic Chemistry: Part A: Structure and
Mechanism, 3rd Ed., New York: Penum Press, 1990, p. 476.
[116] Templeton, A. C., et al., Gateway Reactions to Diverse, Polyfunctional MonolayerProtected Gold Clusters, J. Am. Chem. Soc., Vol. 120, 1998, p. 4845.
[117] Lwowski, W., in 1,3-Dipolar Cycloaddition Chemistry, A. Padwa (Ed.), New York: Wiley,
1984.
96
97
[138] Braude, A. I., The Optical Properties of Metal Nanoparticles and the Bio Application, N.
Engl. J. Med., Vol. 258, 1958, p. 1289.
[139] Klein, H. G., et al., Complementary Bioconjugates Based on Antibody-Antigen with
Nanoparticles, J. Nemo, Transfusion, Vol. 37, 1997, p. 95.
[140] Lee, C. K., et al., Detectable Signal in the Luminescence Image in Bio-Nanoparticles, Vox
Sang., Vol. 82, 2002, p. 204.
[141] Goodnough, L. T., et al., Surfactantless Synthesis of Multiple Shapes of Gold
Nanostructures and Their Shape-Dependent SERS Spectroscopy, N. Engl. J. Med., Vol.
340, 1999, p. 438.
[142] Puckett, A., et al., Capacitance Measurements of Antibody-Antigen Interactions in a Flow
System, J. Clin. Pathol., Vol. 45, 1992, p. 155.
[143] McDonald, C. P., et al., Detection of a Biomarker for Alzheimers Disease from Synthetic
and Clinical Samples Using a Nanoscale Optical Biosensor, Transfus. Med., Vol. 6, 1996,
p. 61.
[144] Anderson, K. C., et al., Magnetic Resonance Imaging of the Brain in Williams-Beuren Syndrome, Am. J. Med., Vol. 81, 1986, p. 405.
[145] Gu, H. W., et al., Using Biofunctional Magnetic Nanoparticles to Capture VancomycinResistant Enterococci and Other Gram-Positive Bacteria at Ultralow Concentration, J.
Am. Chem. Soc., Vol. 125, 2003, No. 15, p. 702.
[146] Gu, H. W., et al., A Nanoscale Optical Biosensor Based on Localized Surface Plasmon
Resonance, Chem. Commun., 2003, p. 1966.
[147] Sun, S. H., et al., Monodisperse FePt Nanoparticles and Ferromagnetic FePt Nanocrystal
Superlattices, Science, Vol. 287, 2000, 1989.
[148] Performance Standards for Antimicrobial Susceptibility Testing, National Committee
for Clinical Laboratory Standards, Villanova, PA, 1998.
[149] Zhao, Q., et al., Carbon Nanotube-Metal Cluster Composites: A New Road to Chemical
Sensors, Nano. Lett., Vol. 5, 2005, p. 847.
[150] See, E. G., in Metal Nanoparticles: Synthesis, Characterization and Applications, C. A.
Foss, Jr. and D. L. Feldheim (Eds.), New York: Marcel Dekker, 2002.
[151] Hu, F., et al., Preparation of Biocompatible Magnetite Nanocrystals for In Vivo Magnetic
Resonance Detection of Cancer, Adv. Mater., Vol. 19, 2006, p. 2553.
[152] Such, C. C., A. S. Berry, and G. Curtis, Functionalisation of Magnetic Nanoparticles for
Applications in Biomedicine, J. Phys. D: Appl. Phys., Vol. 36, 2003, p. R198.
[153] Huh, Y.-M., et al., In Vivo Magnetic Resonance Detection of Cancer by Using Multifunctional Magnetic Nanocrystals, J. Am. Chem. Soc., Vol. 127, 2005, p. 12387.
[154] Zhao, M., et al., Silica-Encapsulated Nanomagnetic Particle as a New Recoverable Biocatalyst Carrier, Nat. Med., Vol. 7, 2001, p. 1241.
[155] Weissleder, R., et al., Long-Circulating Iron Oxides for MR Imaging, Adv. Drug Delivery Rev., Vol. 16, 1995, p. 321.
[156] Gupta, A. K., and M. Gupta, Synthesis and Surface Engineering of Iron Oxide
Nanoparticles for Biomedical Applications, Biomaterials, Vol. 26, 2005, p. 3995.
[157] Meldrum, F. C., B. R. Heywood, and S. Mann, Magnetoferritin: In Vitro Synthesis of a
Novel Magnetic Protein, Science, Vol. 257, 1992, p. 522.
[158] Bulte, J. W. M., et al., MRI Angiography of the Carotid Artery, Magn. Reson. Imaging,
Vol. 4, 1994, p. 497.
[159] Li, Z., L. Wei, M. Y. Gao, and H. Lei, One-Pot Reaction to Synthesize Biocompatible
Magnetite Nanoparticles, Adv. Mater., Vol. 17, 2005, p. 1001.
[160] Cinteza, L. O., et al., Diacyllipid Micelle-Based Nanocarrier for Magnetically Guided
Delivery of Drugs in Photodynamic Therapy, Molecular Pharmaceutics, Vol. 3, No. 4,
2006, pp. 415423.
98
99
[185] Tran, P. T., et al., Use of Luminescent CdSe-ZnS Nanocrystal Bioconjugates in Quantum
Dot-Based Nanosensics, Phys. Stat. Sol. B, Vol. 229, 2002, p. 427.
[186] Willard, D. M., and A. Van Orden, Quantum Dots: Resonant Energy-Transfer Sensor,
Nat. Mater., Vol. 2, 2003, p. 575.
[187] Clapp, A. R., et al., Fluorescence Resonance Energy Transfer Between Quantum Dot
Donors and Dye-Labeled Protein Acceptors, J. Am. Chem. Soc.,Vol. 126, 2004, p. 301.
[188] Wargnier, R., et al., Energy Transfer in Aqueous Solutions of Oppositely Charged
CdSe/ZnS Core/Shell Quantum Dots and in Quantum DotNanogold Assemblies Nano
Lett.,Vol. 4, 2004, p. 451.
[189] Bocsi, J., et al., Automated Four-Color Analysis of Leukocytes by Scanning Fluorescence
Microscopy Using Quantum Dots, Cytometry A., Vol. 69, No. 3, March 2006, pp.
131134.
[190] Le Gac, S. I. Vermes, and A. van den Berg, Quantum Dots Based Probes Conjugated to
Annexin V for Photostable Apoptosis Detection and Imaging, Nano Lett., Vol. 6, No. 9,
September 2006, pp. 18631869.
[191] Dubertret, B., et al., In Vivo Imaging of Quantum Dots Encapsulated in Phospholipid
Micelles, Science, Vol. 298, 2002, p. 1759.
[192] Gao, X. Y. et al., In Vivo Cancer Targeting and Imaging with Semiconductor Dots,
Nature Biotechnol., Vol. 22, No. 968, 2004.
[193] So, M. K., et al., Self-Illuminating Quantum Dot Conjugates for In Vivo Imaging, Nature
Biotechnol., Vol. 24, No. 3, March 2006, pp. 339343.
[194] Zimmer, J. P., et al., Size Series of Small Indium Arsenide-Zinc Selenide Core-Shell
Nanocrystals and Their Application to In Vivo Imaging, J. Am. Chem. Soc., Vol. 128, No.
8, March 1, 2006.
[195] Bakalova, R., et al., Quantum Dot Based Western Blot Technology was Used for
Ultrasensitive Detection of Tracer Proteins, J. Am. Chem. Soc., Vol. 127, No. 26, July 6,
2005, pp. 93289329.
[196] Kahn, E., et al., Analysis of CD36 Expression on Human Monocytic Cells and
Atherosclerotic Tissue Sections with Quantum Dots: Investigation by Flow Cytometry and
Spectral Imaging Microscopy, Anal. Quant. Cytol. Histol., Vol. 28, No. 1, February
2006, pp. 1426.
[197] Swift, J. L., R. Heuff, and D. T. Cramb, A Two-Photon Excitation Fluorescence CrossCorrelation Assay for a Model Ligand-Receptor Binding System Using Quantum Dots,
Biophys. J., Vol. 90, No. 4, February 15, 2006, pp. 13961410.
[198] Dahan, M., et al., Time-Gated Biological Imaging by Use of Colloidal Quantum Dots,
Opt. Lett., Vol. 26, 2001, p. 825.
[199] Kirchner, C., et al., Cytotoxicity of Colloidal CdSe and CdSe/ZnS Nanoparticles, Nano
Lett., Vol. 5, No. 2, 2005, pp. 331338.
[200] Derfus, A. M., et al., Probing the Cytotoxicity of Semiconductor Quantum Dots, Nano
Lett., Vol. 4, No. 1, 2004, pp. 1118.
[201] Hardman, R., A Toxicologic Review of Quantum Dots: Toxicity Depends on
Physicochemical and Environmental Factors, Environ. Health Perspect., Vol. 114 , No. 2,
February 2006, pp. 165172.
[202] Michalet X., et al., Quantum Dots for Live Cells, In Vivo Imaging, and Diagnostics, Science, Vol. 307, No. 5709, Jan. 28, 2005 pp. 53844.
[203] Gao, X., and S. Nie, Molecular Profiling of Single Cells and Tissue Specimens with Quantum Dots, Trends in Biotechnology, Vol. 21, No. 9, 2003, pp. 371373.
[204] Chan, W. C., Bionanotechnology Progress and Advances, Biol. Blood Marrow Transplant., Vol. 12, No. 2, Suppl. 1, January 2006, pp. 8791.
[205] Alivisatos, A.P., W. Gu, and C. Larabell, Quantum Dots as Cellular Probes, Annu. Rev.
Biomed. Eng., Vol. 7, 2005, pp. 5576.
100