Vous êtes sur la page 1sur 19




Atomic clusters of transition metals embedded into polymeric matrices repre-

sent a very attractive class of materials which combine properties belonging
to the nanometer-sized metal phase (such as certain magnetic, electronic, opti-
cal, or catalytic properties) and the polymer phase (such as processability and
lm-forming properties) (1). By the careful selection and combination of both
components, tailor-made functional materials exhibiting unique characteristics
(2) can be obtained, with the advantage of using the well-established, low cost
technologies available for polymer processing, such as printing, spraying, or
Polymeric dispersions of nanometer-sized metal particles offer the possibil-
ity of functionalizing the polymer by properties coming from the large number
of surface atoms (3) and the quantum-size effects (4). Nanometric metals show
properties that differ signicantly from that of bulk metals, which makes these
nanocomposite systems intriguing for scientic study and potentially useful for a
number of technological applications (512). The control of nanoparticle morphol-
ogy becomes a very important aspect, since morphology profoundly inuences the
material performance. As a long-term goal the ability to control and vary particle
size, distributions, shapes, and composition independently from one another is
very desierable, in order to allow the tuning of nanocomposite properties. A broad
area of nanoparticle features, ranging from nanosized single crystals to somewhat
larger (in the range of about 100200 nm) yet well-stabilized nonagglomerates,
gain signicant technological importance.
Polymer-embedded nanostructures are potentially useful for a num-
ber of technological applications, especially as advanced functional materials
(eg, high energy radiation shielding materials, microwave absorbers, optical lim-
iters, polarizers, sensors, hydrogen storage systems) (512). In addition to the
intrinsic nanoscopic material properties, the presence of a very large interface area
in these polymer-based nanocomposites can affect signicantly polymer charac-
teristics (eg, glass-transition temperature, crystallinity, free volume content, igni-
tion temperature), allowing the appearance of further technologically exploitable
mechanical and physical properties (eg, re-resistance, low gas diffusivity).

Historical Background

The fundamental knowledge on the preparation and nature of metal/polymer

nanocomposites looks back at a long history which is connected to the names
of many illustrious scientists (2).
The oldest technique for the preparation of metal/polymer nanocomposites
that can be found in the literature was described in detail in an abstract in 1835
(13). The original article appeared in 1833 (J. Erdmann, p. 22). In an aqueous
solution, a gold salt was reduced in the presence of gum-arabic, and subsequently
a nanocomposite material was obtained in the form of a purple solid simply by
coprecipitation with ethanol.
Encyclopedia of Polymer Science and Technology. Copyright John Wiley & Sons, Inc. All rights reserved.

Around 1900, widely forgotten reports indicated the preparation of polymer

nanocomposites with uniaxially oriented inorganic particles, and their remark-
able optical properties (14,15). Dichroic plants and animal brils (eg, linen, cotton,
spruce, or chitin, amongst others) were prepared by impregnation with solutions
of silver nitrate, silver acetate, or gold chloride, followed by reduction of the cor-
responding metal ions under the action of light (16).
Dichroic lms were also obtained using gold chloride treated gelatin which
was subsequently drawn, dried, and nally exposed to light (17). Similar results
were obtained when gelatin was mixed with colloidal gold before drying and draw-
ing (18). The gold or silver content in such systems amounted typically to ca 1%
by weight (19).
In 1904, Kirchner and Zsigmondy (Nobel Laureate in Chemistry, 1925) re-
ported that nanocomposites of colloidal gold and gelatin reversibly changed color
from blue to red upon swelling with water (20). In order to explain the mechanism
of nanocomposite color change, they suggested that the material absorption must
also be inuenced by the distance between the embedded particles. In addition,
around the same time, the color of gold particles embedded in dielectric matrices
was subject of detailed theoretical analyses by Maxwell Garnett who explained
the color shifts upon variation of particle size and volume fraction in a medium
During the following three decades, dichroic bers were prepared with many
different elements (ie, Os, Rh, Pd, Pt, Cu, Ag, Au, Hg, P, As, Sb, Bi, S, Se, Te,
Br, I) (2327). The dichroism was found to depend strongly on the employed ele-
ment, and optical spectra of dichroic nanocomposites, made of stretched poly(vinyl
alcohol) lms containing gold, silver, or mercury, were presented in 1946 (however,
the preparative scheme used is not really clear) (28). It was assumed already in
the early reports that dichroism was originated by the linear arrangement of small
particles (29) or by polycrystalline rod-like particles (30) located in the uniaxially
oriented spaces present in the bers. An electron micrograph depicted in 1951
showed that tellurium needles with typical dimensions of ca 5 50 nm were
present inside a dichroic lm made of stretched poly(vinyl alcohol), however also
in this case, just a limited number of details were given about the technique used
for sample preparation (31).
In 1910, Kolbe was the rst to prove that dichroic nanocomposite samples
based on gold contained the metal indeed in its zero-valence state. Such afrma-
tion was conrmed a few years later by X-ray scattering. In particular, it was
shown that zero-valence silver and gold were present in the respective nanocom-
posites made with oriented ramie bers, and the ring-like interference patterns
of the metal crystallites showed that the individual primary crystallites were not
oriented (32). Based on Scherrers equation, which was developed just in this pe-
riod, the average particle diameter of silver and gold crystallites was determined
in bers of ramie, hemp, bamboo, silk, wool, viscose, and cellulose acetate to be
between 5 and 14 nm (33).

Basic Concepts

A nanocomposite is a material made of two or more phases one of which has at

least one dimension in a nanometric size range. Metal/polymer nanocomposites

Fig. 1. Structure of atomic clusters of metals: (a) elemental cluster, (b) small cluster, (c)
large cluster, (d) thiol-derivatized cluster, and (e) alloyed cluster.

are made of a continuous polymeric matrix embedding nanosized metal do-

mains. Thermoplastic polymers, elastomers, and thermosetting resins can be po-
tentially used as matrix, but amorphous linear polymers, like optical plastics
(eg, polystyrene, poly(methyl methacrylate)) and conductive polymers are the most
frequently used. About the metallic ller, very small clusters and nanoparticles
with a pseudospherical shape are frequently employed (see Fig. 1). The use of
nanorods and nanowires has also been described (2).
In these materials, the polymer has the function of protecting the nanostruc-
tures and allowing their manipulation, whereas the nanosized ller provides the
polymeric matrix with unique properties coming from small-size effects (ie, meso-
scopic properties). To allow the appearance of mesoscopic properties, the nanopar-
ticle size is required to be very small; usually a dimension less than 30 nm is
necessary for most metallic materials. In particular, the metal domain size, where
electrons move, needs to be comparable to the critical lengths of physical phe-
nomena (eg, wavelength of the electrons at the Fermi edge, mean free paths of
electrons or phonons, coherency length, screening length). Such phenomenon is
named electron connement (4). In addition, on this size scale, also phenomena
related to the high percentage of surface atoms appear in the material (3).
Different topologies can result in the preparation of polymer-embedded metal
clusters. To prepare materials characterized by the properties of surface atoms or
with characteristics coming from the connement effect, a contact-free disper-
sion of clusters must result in the polymer matrix, since only in this case the
large amount of surface atoms present allows the surface properties of matter to
prevail on that of bulk. Both regular (eg, uniaxially oriented pearl-necklace type
of arrays of nanoparticles) and irregular metal cluster distributions are used in
technological applications.
The properties of nanometric particles strictly depend on their microscopical
structure (ie, chemical composition, shape, size, percentage of defects, microstrain
concentration, etc). For example, the characteristic surface plasmon absorption of
a system of metal nanoparticles dispersed into a dielectric matrix is related to the
particle shape and size (34). To prepare a color lter, identical particles should be
used, otherwise the material will appear black. The presence of a single type of
microscopic structure allows each particle to provide the same contribution to the
composite properties. From a theoretical point of view, an ideal nanostructured
composite should be made of identical metal domains uniformly dispersed into
the polymeric matrix. However, since it is very difcult to prepare a sample of

identical metal grains, the only practically possible situation is a system of single
crystals with a narrow size distribution.
Because of the high surface chemical reactivity, metal clusters can be easily
oxidized or contaminated by nucleophilic molecules. The surface reactivity of small
clusters is so high that even noble metals (eg, Pt, Pd, Ag) can be oxidized by air.
Since small molecules present in the air (eg, oxygen, water, sulfur dioxide) are
dissolved in polymers and can diffuse through it, the polymeric matrix cannot
prevent metal from surface reactions. Finally, the polymer matrix has only the
function to freeze particles, avoiding their diffusion and aggregation by sintering.
For such a reason the nanoparticles need to be passivated before their embedding
in polymers. The surface treatment (eg, thiol-derivatization, see Fig. 1d) prevents
particle aggregation by short-range steric repulsion and stabilizes the metal core
principally by electronic effects. Surface passivation offers also the possibility to
disperse ller into most of liquid monomers and organic solvents where polymers
are soluble.


Nanocomposites are biphasic materials that can be classied on the basis of their
microstructure by the self-connectivity concept, that is the number of space di-
rections (X, Y, Z) in which each phase inside the composite physically contacts
itself. Composite classication based on self-connectivity has been proposed for
the rst time by Newnham in 1978 (35). Because of the discontinuous nature of
most nanocomposite materials, the self-connectivity concept needs to be applied
to a limited but representative composite portion (ie, a local self-connectivity def-
inition should be used). In particular, like in macro- and microcomposite systems,
each phase in a nano-composite material can be locally self-connected in zero, one,
two, or three dimensions. It is natural to conne the attention to three perpen-
dicular axes since all property tensors are generally referred to such a system. In
general, for a n-phase system the number of possible self-connectivity patterns
is given by (n + 3)!/3!n!. Therefore, inside a biphasic system (n = 2) there are
10 different self-connectivity patterns: (00) , (10) , (20) , (30) , (11) , (21) ,
(31) , (22) , (32) , and (33) . The rst number in the notation represents the
physical connectivity of one of the two phases and the second number refers to
connectivity of the other one. A schematic representation of these 10 types of self-
connectivities is given in Figure 2, using a cube as the basic building block. Arrows
are used in Figure 2 to indicate the connected directions, and an asterisk has been
introduced in the symbol to indicate the local connectivity.

Properties of Nanosized Metals

Nanosized metals are characterized by novel thermodynamic, chemical, catalytic,

optical, magnetic, and transport properties which are much different from those
of corresponding massive metals. For this reason a 3-D periodic table of ele-
ments has been frequently proposed by chemists working in this material science
eld (36).

Fig. 2. The connectivity patterns of a biphasic solid (35).

The thermodynamic properties of matter are classically described as natural

constants; however, they change signicantly when the dimension approaches a
few nanometers (37). The rst small-size effect which has been observed for met-
als is the change in the melting point (3843). When a solid is heated, it melts, and
the melting temperature is normally considered one of the most important charac-
teristics of a material. Nanometer-size crystals melt at much lower temperatures
than extended ones. This can be understood in two ways. First, the cohesive en-
ergy of a crystal arises by the sum of all pairwise interactions between the atoms.
In a very small crystal the number of surface atoms is large, and therefore the co-
hesive energy per atom has not yet converged on the bulk value. A second picture
is more thermodynamic: as a solid is heated, melting takes place at the temper-
ature where the chemical potentials of solid and liquid are equal. In addition to
the usual term in the chemical potential, a very small crystal also has a term in
it for the surface energy. In general, the surface energy of liquids is less than that
of solids, because a liquid can readily assume the lowest surface area shape, a
sphere. As a consequence, the smaller the crystal the lower is the melting temper-
ature, and the reduction in the melting temperature is proportional to the surface
to volume ratio, or inversely proportional to the nanocrystal radius. This scaling
law for melting temperature reduction has been veried in many metals (37).
Several interesting chemical properties arise as the grain size of a metal is
decreased to a nanometer-size range (eg, enhanced reactivity, stoichiometric be-
havior of heterogeneous reactions, new reaction routes). Theoretical computations

(37) showed that the fraction of metal atoms residing on the surface of a 100-nm
grain increases from 12% up to 8595% for grain sizes in the range of 12 nm.
The importance of this observation is that the chemistry of such surfaces is much
different from that of the same atoms contained within larger grains. If the size of
a cubic crystal is decreased from 100 nm down to the 35 nm range, the fraction
of atoms contained in edge sites as compared to those contained in basal planes
increases to 70%. The consequence of this observation is that as the crystal size
decreases, the fraction of atoms located in low coordination sites increases sharply,
which inherently imparts a higher chemical reactivity to such materials (activated
metals). High surface areas and intrinsically high surface reactivities of hyper-
ne reactants allow surface reactions to approach stoichiometric conversion. With
the development of hyperne powders, the heterogeneous phase reaction can be
carried out much faster and at lower temperatures (44,45). Finally, gassolid re-
actions and liquidsolid reactions take on a new dimension, and hyperne solids
enter now in the chemists arsenal as novel chemical reagents.
The properties of nanosized metals in heterogeneous catalysis are well es-
tablished, since heterogeneous catalysts represent one of the pioneering elds of
nanotechnology (3). The increasing portion of surface atoms with decreasing parti-
cle size, compared with bulk metals, makes small metal particles as highly reactive
catalysts, since surface atoms are the active centers for catalytic elementary pro-
cesses. Among the surface atoms, those sitting on the crystal edges and corners
(see Fig. 3) are more reactive than those on basal planes. The percentage of edge

Fig. 3. SEM micrograph of 3-D self-organized poly(methyl methacrylate) particles. If

beads are considered as atoms, the picture can be used to visualize the different catalytic
sites on the surface of a metal crystal.

and corner atoms also increases with decreasing size and this is the reason for
the enhanced catalytic activity (ie, supercatalytic effect) and different selectivity
of very small metal catalysts.
If metal particles become very small, reaching the nanometer-size scale, a
color may occur. This is a typical phenomenon of nanometric metals. Actually, op-
tical absorption may result in the ultraviolet or visible part of the spectrum, and
this arises from a surface plasmon resonance. This is due to a collective electron
plasma oscillation (plasmon) that is coupled to an external transverse electro-
magnetic eld through the particle surface. It is possible to quantitatively relate
the absorption coefcient to the wavelength of the exciting radiation by the Mie
theory for spherical inclusions in a dielectric matrix (34). Far-IR luminescence is
another optical phenomenon frequently observed with nanosized metals (46).
New magnetic properties appear in metals when the size approaches the
nanometric regime (47). The so-called oddeven effect is a phenomenon observed
when diamagnetic metals are reduced to a nanometric-size regime. In particular,
diamagnetic materials have only spin-paired electrons. However, in practice it
cannot be assumed that a macroscopic piece of a diamagnetic metal does not
have one or more unpaired electrons. This cannot be measured because of the
effectively innite number of atoms and electrons. However, if the particle size
is small enough to make one unpaired electron measurable, the oddeven effect
should become visible. Among small diamagnetic metal particles there should be
an equal percentage of odd and even numbers of electrons.
For magnetic materials such as Fe, Co, and Ni, the magnetic properties are
size-dependent (47). In particular, the coercivity force H c needed to reverse an
internal magnetic eld within the particle changes with particle size and it is
maximum for single-domain particles. Further, the strength of the internal mag-
netic eld of a single particle can be size-dependent.
Giant magnetostriction, magnetoresistivity, and magnetocaloric effects rep-
resent further examples of new properties arising from the small size of magnetic
domains (48).
Also transport properties, which are strictly related to the electronic struc-
ture of a metal particle, critically depend on size (37). For small particles, the
electronic states are not continuous, but discrete, because of the connement of
electron wave function. Consequently, also properties like electrical and thermal
conductivity may exhibit quantum-size effects.
Most of these unique chemical and physical characteristics of nanosized met-
als can be used for the functionalization of plastic materials.

Preparation Methods

A limited number of methods have been developed for the preparation of

metal/polymer nanocomposites. Usually, such techniques consist of highly spe-
cic approaches, which can be classied as in situ and ex situ methods. In the in
situ methods two steps are needed: rstly, the monomer is polymerized in solu-
tion, with metal ions introduced before or after polymerization. Secondly, metal
ions in the polymer matrix are reduced chemically, thermally, or by UV/ - irradia-
tion. In the ex situ processes, the metal nanoparticles are chemically synthesized,

and their surface is organically passivated. The derivatized nanoparticles are dis-
persed into a liquid monomer which is then polymerized.
In situ Methods. A simple, direct, and versatile in situ method was pro-
posed by Watkins and McCarthy (49). In this method, an organometallic precursor
is dissolved in supercritical uid (SCF) carbon dioxide and infused into a solid
polymer as SCF solution. Chemical or thermal reduction of the precursor to the
zero-valence metal either in presence of SCF or subsequent to its removal pro-
duces metal domains within the solid polymer matrix. Such technique has been
used in the synthesis of nanoscale platinum clusters embedded into poly(4-methyl-
1-pentene) and poly(tetrauoroethylene), using dimethyl(cyclooctadiene)Pt(II) as
metal precursor. There are multiple advantages associated with this approach:
rst, the high permeation of CO2 in virtually all polymers and the wide range of
organic and organometallic reagents which are soluble in CO2 render this tech-
nique a generally useful scheme for the synthesis of polymer composites. Neither
the polymer substrate nor the reaction product needs to be soluble in CO2 . Second,
the sorption of CO2 is a very fast process which signicantly enhances the kinet-
ics of the penetrant absorption. The degree of polymer swelling, diffusion rates
within the substrate, and the partitioning of penetrates between the SCF and
the swollen polymer can be controlled by density mediated adjustments of solvent
strength via changes in temperature and pressure. Coupled with manipulation of
reaction rates, SCFs offer unprecedented control over composite composition and
morphology. In addition, SCFs such as CO2 are gases at ambient conditions and
the solvent dissipates rapidly upon the release of pressure. In the case of metala-
tion using dimethyl(cyclooctadiene)Pt(II), the process efuent consists of CO2 and
light hydrocarbons (methane and cyclooctane) derived from the precursor ligands.
Bronstein and co-workers (50) reported an in situ method for the preparation
of polymer-cobalt nanocomposites by mixing CO2 (CO)8 with a polyacrylonitrile
copolymer or an aromatic polyamide in dimethylformamide (DMF). The cobalt
carbonyl interacts with DMF giving the complex [Co(DMF)6 ]2+ [Co(CO)4 ]2 , which
is then converted to nanodispersed Co particles by thermolysis.
Analogously, silver/polyimide nanocomposites have been prepared by Fra-
gala` and co-workers (51) by thermolysis of a mixture of a metallorganic silver com-
plex and polyamidic acid (PPA). The PAA precursor transforms in polyimide (PI)
at a curing temperature compatible with the metallorganic precursor reduction
temperature, so that at the same time the precursor decomposes, giving metallic
silver particles, and the PAA transforms itself in PI, allowing the silver particles
to remain trapped in the polymeric cage.
Copper/polymer nanocomposites have been prepared in the solid state by
Lyons and co-workers (52). A soluble precursor was synthesized by complexing
poly(2-vinylpyridine) with copper formate (ie, Cu(HCO2 )2 ) in methanol. The ther-
mal decomposition of the complex results in a redox reaction whereby Cu(II) is
reduced to copper metal and the formate anion is oxidized to CO2 and H2 . By incor-
porating a reducing agent into the complex, the thermal decomposition reaction
is not diffusion-limited, and nanocrystalline copper particles can be prepared in
the solid state.
Chen and co-workers (53) describe the synthesis of iron nanoparticles dis-
persed in poly(4-vinylpyridine) homopolymer and vinylstyrene-4-vinylpyridine
copolymers by chemical reduction of thin lms of iron chloride/polymer

nanocomposite precursors swollen by DMF. Idrazine was used as reducing agent.

Such approach can be used because the metal ions have a high tendency to form
ionic clusters in a polymeric matrix. The produced microscopic phases, which are
separated by the hydrophobic portions of the polymer, have a diameter ranging
from 2 to 10 nm. The chemical reduction of the dispersed ionic metal domains
produces ultrane metal particles.
Huang and co-workers (54) prepared copper/poly(itaconic acid-co-acrylic
acid) nanocomposites via in situ chemical reduction of the Cu2+ polymer com-
plex by hydrazine hydrate aqueous solution.
Metal/polymer nanocomposites were prepared by Chen and co-workers (55)
using dispersion of metal chlorides in polyurethane. Both polyurethane and metal
salts were dissolved in N,N  -dimethylacetamide, followed by lm casting and re-
duction of the metal salts by sodium borohydrate. The metal particle size depended
on the type of metal salt used and on its concentration.
Zhu and co-workers (56) developed a method for the preparation of
metal/polymer nanocomposites based on -irradiation. In this method, the metal
salt was dissolved in the organic monomer, and the formation of nanocrystalline
metal particles and monomer polymerization was obtained simultaneously in solu-
tion by irradiation with a 60 Co -ray source, leading to a homogeneous dispersion
of nanocrystalline metal particles in the polymer matrix. Also UV-irradiation has
been used for the reduction of metal ions dispersed in polymer matrices (57).
Wizel and co-workers (58) have described the use of ultrasounds in the
in situ synthesis of composite materials made of polystyrene and iron. The prop-
agation of ultrasound waves through a uid causes the formation of cavita-
tion bubbles. The collapse of these bubbles, described as an implosion in the
hot-spot theory, is the origin of extreme local conditions: high temperatures
(500025,000 K) and high pressures (1000 atm). The cooling rates obtained during
the bubble collapse are greater than 107 K s 1 . These high cooling rates have
been utilized in the sonication of Fe(CO)5 as a neat liquid or in solution, to produce
amorphous iron nanoparticles. In addition, ultrasound radiation has been widely
used for the preparation of polymers without the use of initiators (59). These
two chemical processes have been combined for the fabrication of metal/polymer
nanocomposites by sonication of styrene solution of iron carbonyl.
Ex Situ Methods. Because of the high optical purity that can be achieved
in the nal product, the ex situ synthesis of metal/polymer nanocomposites is a
very attractive technique, especially in the preparation of materials for optical
applications. The particulate material of the required size, as obtained by a so-
lution chemistry route, is stabilized by legand chemisorption (eg, thiols) in order
to reduce their surface reactivity and tendency to agglomeration, and then it is
incorporated into a castable polymeric matrix. Usually, the passivated nanoparti-
cles are dispersed into a liquid monomerinitiator mixture (eg, styrene or methyl
methacrylate activated by benzoyl peroxide), which is then thermally polymer-
ized. Because a little amount of ller is required, the polymerization behavior of
the monomer is not signicantly inuenced by the presence of passivated nanopar-
ticles (the reaction rate is only slightly decreased (60)). Gonsalves and co-workers
(60,61) have prepared surface-derivatized gold particles by phase-transfer reac-
tion of gold ions with dodecanthiol and used these particles for the synthesis of
poly(methyl methacrylate) based nanocomposites. Carotenuto (62) has developed

a new technique for the preparation of bulk quantities of atomic clusters of gold,
derivatized by thiols by a modication of the well-known Polyol Process technique
(63). The ex situ approach has been used also in the synthesis of silver/polyethylene
nanocomposites (64,65). A variety of high nuclearity metal cluster compounds
which can be used for the polymeric nanocomposite preparation have been re-
cently synthesized and studied (6676).
Clusters of noble and easily reducible metals (eg, Ag, Au, Pd) can be ob-
tained by alcoholic reduction of metal ions in presence of polymeric stabilizers.
Poly(vinylpyrrolidinone), poly(vinyl alcohol), poly(methylvinyl ether), etc are usu-
ally utilized. High quality polymer-protected metal clusters can be obtained at
the end of reaction simply by addition of a nonsolvent liquid which causes the
nanocomposite separation from the reactive mixture (coprecipitation) (7779).
However, the use of such nanocomposites is limited because of their excessive
moisture sensitivity.

Characterization Techniques

For the comprehension of mechanisms involved in the appearance of novel prop-

erties in polymer-embedded metal nanostructures, their characterization repre-
sents the fundamental starting point. The microstructural characterization of
nanollers and nanocomposite materials is performed mainly by transmission
electron microscopy (TEM), large-angle X-ray diffraction (XRD), and optical spec-
troscopy (UVvis). These three techniques are very effective to determine particle
morphology, crystal structure, composition, and grain size (48).
Of the many techniques which have been used to study the structure of
metal/polymer nanocomposites, TEM has undoubtedly been the most useful. This
technique is currently used to probe the internal morphology of nanocomposites.
As visible in Figure 4, high quality images can be obtained because of the presence,
in the sample, of regions that do not allow the high voltage electron beam passage
(ie, the metallic domains) and region perfectly transparent to the electron beam
(ie, the polymeric matrix). High resolution TEM (HRTEM) allows morphological
investigations with a resolution of 0.1nm, and thus this technique makes possible
to accurately image nanoparticle sizes, shapes, and in some cases even inner atoms
(80) (see Fig. 5).
Large-angle X-ray powder diffraction (XRD) has been one of the most ver-
satile techniques utilized for the structural characterization of nanocrystalline
metal powders. The modern improvements in electronics, computers, and X-ray
sources have allowed it to become an indispensable tool for identifying nanocrys-
talline phases as well as crystal size and crystal strain (see Fig. 6). The comparison
of the crystallite size obtained by the XRD diffractogram using the Scherrer for-
mula with the grain size obtained from the TEM image allows to establish if the
nanoparticles have a mono- or polycrystalline nature (11).
Metal clusters are characterized by the surface plasmon resonance, which is
an oscillation of the surface plasma electrons induced by the electromagnetic eld,
and consequently their microstructure can be indirectly investigated by optical
spectroscopy (UVvis spectroscopy). The characteristics of this absorption (shape,
intensity, position, etc) are strictly related to the nature, structure, topology, etc

Fig. 4. TEM micrograph of metal gold clusters embedded in poly(vinylpyrrolidinone).

of the cluster system. In fact, the absorption frequency is a ngerprint of the par-
ticular metal, the eventual peak splitting reects aggregation phenomena, the
intensity of the peak is related to the particle size, the absorption wavelength
is related to the particle shape, the shift of the absorption with increasing of
temperature is indicative of a cluster melting, and so on. For bimetallic particles
information about inner structure (intermetallic or core/shell) and composition
can be obtained from the maximum absorption frequency (81). Differently from
out-line techniques (eg, TEM, XRD) this method allows on-line and in situ cluster
sizing and monitoring of morphological evolution of the system (see Fig. 7). This
method has been used also in the study of cluster nucleation and growth mech-
anisms (82,83). In addition to multielectron transition (ie, plasmon absorption)
also single-electron transition (interband transition) can be detected and studied
by optical spectroscopy in order to obtain important microstructural information.
This outline of the principal characterization techniques for nanocomposite
materials and nanosized metal llers is far from being complete. Advances in
Raman spectroscopy, energy dispersive spectroscopy, infrared spectroscopy, and
many other techniques are of considerable importance as well. In fact, the success
that nanostructured materials are having in the last few years is strictly related
to the advanced characterization techniques which are available today.

Fig. 5. HRTEM image of a very small metal cluster.

Fig. 6. XRD pattern of nanosized gold clusters.


Fig. 7. UVvis spectra of a colloidal gold suspension during the cluster growth. Temper-
ature = 60 C; 5 min, 10 min, - - - - 15 min.


Applications of metal/polymer nanocomposites have already been made in differ-

ent technological elds. However the use of a much larger number of devices based
on these materials is predicted for the near future.
Because of the plasmon surface absorption band, atomic clusters of metals
can be used as pigments for optical plastics. The color of the resulting nanocom-
posites is light-fast and intensive, and these materials are perfectly transparent,
since the cluster size is much lower than light wavelength. Gold, silver, and copper
can be used for color lter application. Also UV absorbers can be made for example
by using Pd clusters. The plasmon surface absorption frequency is modulated by
making intermetallic particles (eg, Pd/Ag, Au/Ag) of adequate composition.
As shown in Figure 8, polymeric lms containing uniaxially oriented pearl-
necklace type of arrays of nanoparticles exhibit a polarization-dependent and tun-
able color (64,65,8486). The color of these systems is very bright and can change
strongly, modifying the light polarization direction. These materials are obtained
by dispersing metal nanoparticles in polymeric thin lms and subsequently reor-
ganizing the dispersed phase into pearl-necklace arrays by solid-state drawing at
temperature below the polymer melting point. The formation of these arrays in
the lms is the cause of a strong polarization-direction-dependent color which can

Fig. 8. (a) UVvis spectra of drawn polyethylene/silver nanocomposites containing 4 wt%

dodecanethiol coated silver nanoparticles in linearly polarized light. (b) Absorption spectra
of drawn polyethylene/silver nanocomposite lm annealed for 15 h at 180 C. The angle
between the polarization direction of the light and the drawing direction of the lm is
given (64).

be used in the fabrication of liquid-crystal color display (see Fig. 9) and special
electrooptical devices (see Fig. 10).
Surface plasmon resonance has been used to produce a wide variety of optical
sensors, eg, systems which are able to change their color in presence of specic
analytes. These devices can be used as sensors for immunoassay, gas, and liquid
(see Fig. 11) (48).
Metals are characterized by ultrahigh/low refractive indices and therefore
they can be used to modify the refractive index of optical plastics (8789). Ultra-
high/low refractive index optical plastics can be used in the waveguide technology
(eg, planar waveguides and optical bers).
Plastics doped by atomic clusters of ferromagnetic metals show magneto-
optical properties (ie, when subject to a strong magnetic eld, they can rotate
the vibration plane of a plane-polarized light) and therefore they can be used

Fig. 9. Schematic representation of (a) a conventional color twisted nematic liquid-crystal

display; and (b) display set-up containing a color polarizing lter (64).

Fig. 10. Twisted nematic liquid-crystal cell equipped with a polyethylene/silver nanocom-
posite lter (65).

as Faraday rotators (90,91). These devices have a number of important optical

applications (eg, magneto-optic modulators, optical isolators, optical shutters).
Nanosized metals (eg, gold, silver) have attracted much interest because of
the nonlinear optical polarizability, which is caused by the quantum connement
of the metals electron cloud (92) (see NONLINEAR OPTICAL PROPERTIES). When ir-
radiated with light above a certain threshold power, the optical polarizability
deviates from the usual linear dependence on that power. By incorporating these
particles into a clear polymeric matrix, nonlinear optical devices can be made in
a readily processable form (61). These materials are used to prepare a number of
devices for photonics and electrooptics (see ELECTROOPTICAL APPLICATIONS).
Contact-free dispersion of noble metal clusters in polymer can be also used as
nonporous catalytic membranes (9395). Traditionally, nonporous catalytic sep-
aration layers are composed of palladium or palladium alloy foils. Reduction in

Fig. 11. Nanosized gold-based optical sensor for biological assay.


thickness by metal lm deposition on special supports increases the hydrogen

permeability and reduces the costs for the precious metal. Polymeric matrix lled
by a catalytic nanosized metal offers in addition to the above advantages, the
possibility of using the catalytic properties of nanoscale metal catalysts.
One of the most important characteristics of polymers lled by metal mi-
croclusters is their ability to absorb microwave radiation, producing heat (96).
These plastic materials can be processed by using new technologies based on
the microwave heating and can be used for food packaging, microwave shielding,
RADAR camouage, plastic welding by microwaves, etc.
Metal/polymer nanocomposites can have many other important applications.
For example, nanoparticles embedded into poly(vinylpyrrolidinone) can be used
for the electroless plating of polymeric, ceramic, and semiconductor substrates
(9398). These materials have also been used for the preparation of smart sys-
tems that experience a reversible alteration of their properties upon exposure to
light. They are used as infrared barriers against exposures to intense solar light
or res (99).


Metal clusterpolymer systems are receiving increased attention because of the

interesting properties and large potentialities for technological applications. In
particular, these nanocomposite systems can provide simple, low cost options for
obtaining tailored materials with high promise for various catalytic, optical, mag-
netic, and electronic applications. Usually, the control of the nanoparticle mor-
phology, ie, particle size, size distribution, shape, and composition, is of main in-
terest. Nanoparticle features can be varied by selection of preparation method and
variation of experimental conditions. In addition to the versatility of the nanopar-
ticle features and the polymer morphologies, options can be provided to tailor the
topology of the metalpolymer systems. This includes for example the uniaxially
oriented pearl-necklace type distribution of metal nanoparticles within the poly-
mer matrix. For a variety of applications special topologies are important and
determining factors for tuning the composite material performance.
In conclusion, further research activity in this eld is extremely important
for the development of advanced devices for functional applications based on these


1. A. B. R. Mayer, Mater. Sci. Eng., C 6, 155 (1998).

2. W. Caseri, Macromol. Rapid Commun. 21, 705 (2000).
3. W. R. Moser, Advanced Catalysts and Nanostructured Materials: Modern Synthetic
Methods, Chapman & Hall, New York, 1996.
4. R. C. Ashoori, Nature 379, 413 (1996).
5. G. Carotenuto, Polym. News 25, 191 (2000).
6. G. Carotenuto, Polym. News 25, 260 (2000).
7. G. Carotenuto, Polym. News 25, 365 (2000).

8. G. Carotenuto, Polym. News 26, 48 (2001).

9. G. Carotenuto, Polym. News 26, 119 (2001).
10. G. Carotenuto, Polym. News 26, 234 (2001).
11. G. Carotenuto, Polym. News 26, 278 (2001).
12. G. Carotenuto, Polym. News 26, 342 (2001).
13. C. U. Shepard, Sill. J. (Am. J. Sci. Arts) 28(1), 145 (1835).
14. H. Ambronn, Kgl. Sachs.
Ges. Wiss. 8, 613 (1896).
15. H. Ambronn and R. Zsigmondy, Ber. Sachs. Ges. Wiss. 51, 13 (1899).
16. H. Ambronn, Kgl. Sachs.
Ges. Wiss. 8, 613 (1896).
17. H. Ambronn, Kgl. Sachs.
Ges. Wiss. 8, 613 (1896).
18. H. Ambronn and R. Zsigmondy, Ber. Sachs. Ges. Wiss. 51, 13 (1899).
19. E. Kolbe, Ph.D. dissertation, Jena, Germany, 1913 (cited by A. Frey-Wyssling, Proto-
plasma 27, 372 (1937)).
20. F. Kirchner and R. Zsigmondy, Ann. Phys., vierte Folge (Drudes Ann.) 15, 573
21. J. C. Maxwell Garnett, Philos. Trans. R. Soc. London, A 203, 385 (1904).
22. J. C. Maxwell Garnett, Philos. Trans. R. Soc. London, A 205, 237 (1906).
23. F. Braun, Ann. Phys., vierte Folge (Drudes Ann.) 16, 238 (1905).
24. A. Frey, Jahrbuch Wiss. Bot. 67, 597 (1927).
25. A. Frey, Z. Wiss. Mikrosk 42, 421 (1925).
26. A. Frey, Naturwiss 13, 403 (1925).
27. F. Bion, Helv. Phys. Acta 1, 165 (1928).
28. E. H. Land and C. D. West, in J. Alexander, ed., Colloid Chemistry, Reinhold, New
York, 1946, p. 160.
29. H. Ambronn, Z. Wiss. Mikrosk. 22, 349 (1906).
30. A. Frey-Wyssling, Protoplasma 27, 563 (1937).
31. E. H. Land, J. Opt. Soc. Am. 41, 957 (1952).
32. S. Berkman, J. Bohm, and H. Zocher, Z. Phys. Chem. 124, 83 (1926).
33. A. Frey-Wyssling, Protoplasma 27, 372 (1937).
34. U. Kreibig and M. Vollmer, in J. P. Toennies, ed., Optical Properties of Metal Cluster,
Springer-Verlag, Berlin, 1993.
35. R. E. Newnham, D. P. Skinner, and L. E. Cross, Mater. Res. Bull. 13, 525 (1978).
36. A. Rosen, Adv. Quantum Chem. 30, 235 (1998).
37. K. J. Klabunde, ed., Nanoscale Materials in Chemistry, Wiley-Interscience, New York,
38. M. Wautelet, Solid State Commun. 74, 1237 (1990).
39. Ph. Buffat and J. P. Borel, Phys. Rev. A 13, 2287 (1997).
40. Q. Jiang, Z. Zhang, D. T. Hsu, H. Y. Tong, and M. Iskandar, J. Mater. Sci. 34, 5919
41. G. L. Allen, R. A. Bayles, W. W. Gile, and W. A. Jesser, Thin Solid Films 144, 297 (1986).
42. C. R. M. Wronski, Br. J. Appl. Phys. 18, 1731 (1967).
43. Z. L. Wang, J. M. Petroski, T. C. Green, and M. A. El-Sayed, J. Phys. Chem. B 102, 6145
44. Y. Zhang, S. Liao, Y. Fan, J. Xu, and F. Wang, J. Nanoparticle Res. 3, 23 (2001).
45. R. D. Rieke and S. E. Bales, J. Am. Chem. Soc. 96, 1775 (1974).
46. J. Zhang, Z. Wang, J. Liu, S. Chen, and G. Liu, Self-Assembled Nanostructures, Kluwer
Academic, New York, 2003.
47. J. S. Miller and M. Drillon, eds., Magnetism: Molecules to Materials III, Nanosized
Magnetic Materials, Wiley-VCH, Weinheim, Germany, 2002.
48. H. S. Nalwa, ed., Nanostructured Materials and Nanotechnology, Academic Press, San
Diego, Calif., 2001.
49. J. J. Watkins and T. J. McCarthy, Chem. Mater. 7, 1991 (1995).

50. M. Bronstein, E. Sh. Mirzoeva, M. V. Seregina, P. M. Valetsky, S. P. Solodovnikov and

R. A. Register, ACS Symp. Ser. 622, 102 (1996).
51. M. E. Fragala,` G. Compagnini, G. Malandrino, C. Spinella, and O. Puglisi, Eur. Phys.
J. D 9, 631 (1999).
52. A. M. Lyons, S. Nakahara, M. A. Marcus, E. M. Pearce, and J. V. Waszczak, J. Phys.
Chem. 95, 1098 (1991).
53. L. Chen, W. J. Yang, and C. Z. Yang, J. Mater. Sci. 32, 3571 (1997).
54. C. Huang and C. Z. Yang, Appl. Phys. Lett. 74, 1692 (1999).
55. G. M. Nasr and M. M. Badawy, Polym. Int. 38, 249 (1995).
56. Y. Zhu, Y. Qian, X. Li and M. Zhang, Chem. Commun. 1081 (1997).
57. M. Zheng, M. Gu, Y. Jin and G. Jin, Mater. Res. Bull. 36, 853 (2001).
58. S. Wizel, S. Margel and A. Gedanken, Polym. Int. 49, 445 (2000).
59. P. Kruus, M. ONeill and D. Robertson, Ultrasonics 28, 304 (1990).
60. K. E. Gonsalves, G. Carlson, X. Chen, S. K. Gayen, R. Perez and M. Jose-Yacaman,
Nanostruct. Mater. 7, 293 (1996).
61. K. E. Gonsalves, G. Carlson, X. Chen, J. Kumar, F. Aranda, R. Perez and
M. Jose-Yacaman, J. Mater. Sci. Lett. 15, 948 (1996).
62. G. Carotenuto and L. Nicolais, J. Mater. Chem. 13, 1038 (2003).
63. T. Sugimoto, ed., Fine Particles Synthesis, Characterization, and Mechanism of Growth,
Marcel Dekker, Inc., New York, 2000, p. 460496.
64. Y. Dirix, C. Bastiaansen, W. Caseri and P. Smith, J. Mater. Sci. 34, 3859 (1999).
65. Y. Dirix, C. Bastiaansen, W. Caseri and P. Smith, Adv. Mater. 11, 223 (1999).
66. R. L. Whetten, J. T. Khoury, M. M. Alvarez, S. Murty, I. Vezmar, Z. L. Wang, P. W.
Stephens, C. L. Cleveland, W. D. Luedtke and U. Landman, Adv. Mater. 8, 428 (1996).
67. R. L. Whetten, M. N. Shagullin, J. T. Koury, T. G. Scaaff, I. Vezmar, M. M. Alvarez
and A. Wilkinson, Acc. Chem. Res. 32, 397 (1999).
68. W. D. Luedtke and U. Landman, J. Phys. Chem. 100, 13324 (1996).
69. P. C. Ohara, D. V. Leff, J. R. Heath and W. M. Gelbart, Phys. Rev. Lett. 75, 3466 (1995).
70. R. L. Whetten, M. N. Shagullin, J. T. Khoury, T. G. Schaaff, I. Vezmar, M. M. Alvarez
and A. Wilkinson, Acc. Chem. Res. 32, 397 (1999).
71. Z. L. Wang, S. A. Harfenist, R. L. Whetten, J. Bentley and N. D. Evans, J. Phys. Chem.
B 102, 3068 (1998).
72. M. M. Alvarez, J. T. Khoury, T. G. Scheeff, M. Shagullin, I. Vezmar and R. L. Whetten,
Chem. Phys. Lett. 266, 91 (1997).
73. A. Badia, W. Gao, S. Singh, L. Demers, L. Cuccia and L. Reven, Langmuir 12, 1262
74. C. B. Murray, S. Sun, W. Gaschler, H. Doyle, T. A. Betley and C. R. Kagan, IBM J. Res.
Dev. 45, 47 (2001).
75. G. Katabi, Y. Koltypin, X. Cao and A. Gedanken, J. Cryst. Growth 166, 760 (1996).
76. Z. L. Wang, S. A. Harfenist, I. Vezmar, R. L. Whetten, J. Bentley, N. D. Evans and
K. B. Alexander, Adv. Mater. 10, 808 (1998).
77. G. Carotenuto, G. P. Pepe, and L. Nicolais, Eur. Phys. J. B 16, 11 (2000).
78. G. Carotenuto, G. P. Pepe, L. Parlato, and L. Nicolais, Mater. Eng. 11, 261 (2000).
79. G. Carotenuto, Appl. Organomet. Chem. 15, 1 (2001).
80. M. J. Y. Yacaman, R. Perez, P. Santiago, M. Banaissa, K. Gonsalves, and G. Carlson,
Appl. Phys. Lett. 69, 913 (1996).
81. H. Remita, J. Khatouri, M. Treguer, J. Amblard, and J. Belloni, Z. Phys. D 40, 127
82. G. Carotenuto, S. DeNicola, G. P. Pepe, and L. Nicolais, Eur. Phys. J. B 24, 437 (2001).
83. G. Carotenuto, S. DeNicola, and L. Nicolais, J. of Nanoparticle Res. 3, 469 (2001).
84. N. A. F. Al-Rawashdeh, M. L. Sanfrock, C. J. Seugling, and C. A. Foss, J. Phys. Chem.
102, 361 (1998).

85. J. C. Wittmann and P. Smith, Nature 352, 414 (1991).

86. A. H. Lu, G. H. Lu, A. M. Kessinger, and C. A. Foss, J. Phys. Chem. B 101, 9139 (1997).
87. L. Zimmerman, M. Weibel, W. Caseri, and U. W. Suter, J. Mater. Res. 8, 1742 (1993).
88. M. Weibel, W. Caseri, U. W. Suter, H. Kiess, and E. Wehrli, Polym. Adv. Technol. 2, 75
89. L. Zimmerman, M. Weibel, W. Caseri, U. W. Suter, and P. Walther, Polym. Adv. Technol.
4, 1 (1992).
90. A. H. Lu, G. H. Lu, A. M. Kessinger, and C. A. Foss, J. Phys. Chem. B 101, 9139 (1997).
91. K. Baba, F. Takase, and M. Miyagi, Opt. Commun. 139, 35 (1997).
92. K. Puech and W. J. Blau, J. Nanoparticle Res. 3, 13 (2001).
93. L. Troger, H. Hunnefeld, S. Nunes, M. Oehring, and D. Fritsch, Z. Phys. D 40, 81 (1997).
94. D. Fritsch and K. V. Peinemann, Catal. Today 25, 277 (1995).
95. M. Ohtaki, M. Komiyama, H. Hirai and N. Toshima, Macromolecules 24, 5567 (1991).
96. D. Acierno, C. Leonelli, G. C. Pellacani, eds., Application of the Microwave Technology,
Mucchi Editore, Modena, Italy, 2000.
97. E. P. Boonekamp, J. J. Kelly, and L. G. J. Fokkink, Langmuir 10, 4089 (1994).
98. A. M. T. Van der Putten, J. W. G. De Bakker, and L. G. J. Fokkink, J. Electrochem. Soc.
139, 3475 (1992).
99. G. A. Gaddy, J. Mclain, S.-W. Huang, B. L. Slaten, and G. Mills, in Proceedings 67th
Annual Meeting of the APS Southeastern Section, Mississippi State, 2000.

Institute for the Composite Material Technology,
National Research Council