View Article Online / Journal Homepage / Table of Contents for this issue

Chemical synthesis of magnetic nanoparticles

Taeghwan Hyeon
National Creative Research Initiative Center for Oxide Nanocrystalline Materials and School of Chemical
Engineering, Seoul National University, Seoul 151-744, Korea. E-mail: [email protected]
Received (in Cambridge, UK) 9th September 2002, Accepted 12th November 2002
First published as an Advance Article on the web 3rd December 2002

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

Recent advances in the synthesis of various magnetic nanoparticles using colloidal chemical approaches are reviewed.
Typically, these approaches involve either rapid injection of
reagents into hot surfactant solution followed by aging at high
temperature, or the mixing of reagents at a low temperature and
slow heating under controlled conditions. Spherical cobalt
nanoparticles with various crystal structures have been synthesized by thermally decomposing dicobalt octacarbonyl or by
reducing cobalt salts. Nanoparticles of FePt and other related
iron or cobalt containing alloys have been made by simultaneously reacting their constituent precursors. Many different
ferrite nanoparticles have been synthesized by the thermal
decomposition of organometallic precursors followed by oxidation or by low-temperature reactions inside reverse micelles.
Rod-shaped iron nanoparticles have been synthesized from the
oriented growth of spherical nanoparticles, and cobalt nanodisks were synthesized from the thermal decomposition of
dicobalt octacarbonyl in the presence of a mixture of two
surfactants.

Introduction

DOI: 10.1039/b207789b

The development of uniform nanometer sized particles has been

intensively pursued because of their technological and fundamental scientific importance.1 These nanoparticulate materials
often exhibit very interesting electrical, optical, magnetic, and
chemical properties, which cannot be achieved by their bulk
counterparts.2 The synthesis of discrete magnetic nanoparticles
with sizes ranging from 2 to 20 nm, is of significant importance,
because of their applications in multi-terabit in22 magnetic
storage devices.3 Such magnetic nanoparticles could also find
Taeghwan Hyeon received his B. S. (1987) and M. S. (1989) in
Chemistry from Seoul National University, Seoul, Korea. He
obtained his Ph.D. from the University of Illinois at UrbanaChampaign (1996) under the supervision of Professor Kenneth
S. Suslick. After conducting postdoctoral research with Professor Wolfgang M. H. Sachtler at Northwestern University, he
joined the faculty of the School of Chemical Engineering of
Seoul National University in September 1997. He has received
several awards including the T. S. Piper Award (University of
Illinois, 1996), Korean Young Scientist Award (Korean Government, 2002), KCSWiley Young Chemist Award (Korean
Chemical Society, 2001), and The Scientist of the Month Award
(Korean Science and Engineering Foundation, 2002). In July
2002, he became the director of the National Creative Research
Initiative Center for Oxide Nanocrystalline Materials supported by the Korean Ministry of Science and Technology. His
current research interests include synthesis of metallic and
oxide nanoparticles, synthesis of nanoporous carbon materials,
and catalytic applications of nanostructured materials.

applications in ferrofluids, magnetic refrigeration systems,

contrast enhancement in magnetic resonance imaging, magnetic
carriers for drug targeting and catalysis.4 In the current article,
I will discuss recent advances made in the synthesis of various
magnetic nanoparticles using colloidal chemical synthetic
procedures. A general review on the synthesis and understanding of nanostructured magnetic materials up to 1996 is
given in the article by Leslie-Pelecky and Rieke.5 The materials
covered in the current article are colloidal systems composed of
isolated particles with nanometer-sized dimensions that are
stabilized by surfactant molecules and dispersed in solvent
media. In the ideal case, these non-interacting systems derive
their unique magnetic properties mostly from the reduced size
of the isolated nanoparticles, and contributions from interparticle interactions are negligible. The surfactant coating on
magnetic nanoparticles prevents clustering due to steric repulsion. Dynamic adsorption and desorption of surfactant molecules onto particle surfaces during synthesis enables reactive
species to be added onto the growing particles. These
nanoparticles can be dispersed in many organic solvents and can
be retrieved as powder forms by removing the solvent.
The followings are several key issues of nanoparticle
synthesis.
4 Particle size distribution (uniformity): Can we synthesize
monodisperse nanocrystals?
4 Particle size control: Can we control the particle size of
nanocrystals in a reproducible manner?
4 Crystallinity and crystal structure: Can we obtain materials
with satisfactory high crystallinity and the desired crystal
structure?
4 Shape-control: Can we synthesize non-spherical and anisotropic nanoparticles?
4 Alignment: For device applications, we should also consider
the alignment of nanoparticles on substrates.
In the present article, I am especially interested in the
synthesis of monodisperse magnetic nanoparticles. Monodisperse nanoparticles are generally considered to be samples
with standard deviations s @ 5% diameter for spherical
particles. Often nanocrystals are referred to as well-defined
crystalline materials, whereas nanoparticles is a term used more
generally for particles with diameters of 250 nm with variable
crystallinity.
Chemical methods have been widely used to produce
nanostructured materials due to their straightforward nature and
their potential to produce large quantities of the final product.
Realizable particle sizes range in size from nanometers to
micrometers, by controlling particle size during synthesis by
using competition between nucleation and growth. There are
several synthetic procedures for synthesizing monodisperse
nanoparticles. It is well known that a short burst of nucleation
followed by slow controlled growth is critical to produce
monodisperse particles. The following describe two representa-

This journal is The Royal Society of Chemistry 2003

CHEM. COMMUN., 2003, 927934

927

View Article Online

tive synthetic procedures. In the first synthetic procedure (Fig.

1), the rapid injection of the reagents, often organometallic

Short-chain alkylphosphines allowed faster growth, and resulted in the bigger particles, while bulkier surfactants reduced
particle growth and favored production of smaller nanoparticles. For example, when bulky trioctylphosphine was
applied in the synthesis, 26 nm sized particles were generated.
In contrast, when tributylphosphine was used as a stabilizer
711 nm sized nanoparticles were produced. A further narrowing of particle size distribution was proceeded using a size
selective precipitation by the gradual addition of alcohol (e.g.
ethanol) into a hydrocarbon (e.g. hexane) dispersion containing
nanoparticles with wide size distribution. Monodisperse cobalt
nanoparticles organize into two- and three-dimensional superlattices. Fig. 2 shows a 2D assembly of 9 nm cobalt

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

Fig. 1 Generalized synthesis of monodisperse nanoparticles by the injection

of reagents into hot surfactant solution followed by aging and size-selective
process.

compounds, into hot surfactant solution induces the simultaneous formation of many nuclei.6 Alternatively, reducing agents
are added to metal salt solutions at high temperature. In the
second procedure, reagents are mixed at low temperature and
the resulting reaction mixtures are slowly heated in a controlled
manner to generate nuclei. Subsequently, particle growth occurs
by the addition of reactive species. Particle size can be also
increased by aging at high temperature by Oswalt ripening, in
which smaller nanoparticles dissolve and deposit on the bigger
nanoparticles. The growth of nanoparticles can be stopped by
rapidly decreasing the reaction temperature. Nanoparticles
produced using these synthetic procedures often have particle
size distributions with s ~ 10%. Further size-selection processes can narrow the particle size distribution below s = 5%.
The most frequently applied size-selection involves the addition
of a poor solvent to precipitate the larger particles. When poor
solvent is added to a mixture containing nanoparticles of
various sizes, the biggest particles flocculate first because of
their greatest van der Waals attraction. This precipitate can be
retrieved by centrifugation. The precipitate can be re-dissolved
in solvent, and further size-selection process can be performed
to yield particles with narrower particle size distribution. Very
recently, Klabunde and coworkers reported the fabrication of
extremely monodisperse gold nanoparticles through digestive
ripening.7
The particle sizes of nanoparticles can be controlled by
systematically adjusting the reaction parameters, such as time,
temperature, and the concentrations of reagents and stabilizing
surfactants. In general, particle size increases with increasing
reaction time, because more monomeric species are generated,
and with increasing reaction temperature because the rate of
reaction is increased.

Nanoparticles of iron, cobalt and nickel

The Murray group at the IBM Watson research center has
studied the synthesis of monodisperse metallic magnetic
nanoparticles intensively.8,9 They have synthesized cobalt
nanoparticles with several different crystal structures using
different synthetic procedures. High temperature reduction of
cobalt chloride was employed to synthesize e-phase cobalt
nanoparticles.9a The injection of superhydride (LiBEt3H)
solution in dioctyl ether into a hot cobalt chloride solution in
dioctyl ether (200 C) in the presence of oleic acid and
trialkylphosphine induced the simultaneous formation of many
small metal clusters, which acted as nuclei for nanoparticle
formation. Continued heating at 200 C induced the growth of
these clusters to the nanoparticle level. Particle size was
controlled by the steric bulkiness of the stabilizing surfactants.
928

CHEM. COMMUN., 2003, 927934

Fig. 2 TEM images of 9 nm e-Co nanoparticles (inset, high-resolution TEM

image).9a Reprinted with permission from reference 9(a). Copyright 1999
American Institute of Physics.

nanoparticles. A high-resolution transmission electron micrograph of a single nanoparticle, shown in the inset of Fig. 2,
reveals its highly crystalline nature. The X-ray diffraction
pattern of these nanoparticles reveals a complex cubic structure
related to the beta phase of elemental manganese (e-Co).
Heating the e-Co nanoparticles at 300 C generated hcp Co
nanoparticles. The Bawendi group have reported on the
synthesis of e-Co nanoparticle powders from the thermal
decomposition of Co2(CO)8 in TOPO.10 The Alivisatos group
have reported on the synthesis of monodisperse e-Co nanoparticles by the rapid pyrolysis of dicobalt octacarbonyl in the
presence of a surfactant mixture composed of oleic acid, lauric
acid and trioctylphosphine.11 They could synthesize 317 nm
sized cobalt nanoparticles by controlling the precursor/surfactant ratio, the reaction temperature, and the injection time. Black
et al. later demonstrated spin-dependent electron transport using
self-assembled 10 nm e-Co nanoparticles over 100 nm wide
electrodes.12 Single electron tunneling was clearly observed for
the two-dimensional hexagonal arrays of cobalt nanoparticles.
Similar single electron tunneling in cobalt nanoparticles was
also reported by Pileni and workers.13
Hcp cobalt nanoparticles were synthesized by using the socalled polyol process, in which high boiling alcohol is applied
as both a reductant and a solvent.8 In a typical synthesis,
1,2-dodecanediol is added into hydrated cobalt acetate solution
dissolved in diphenyl ether containing oleic acid and trioctylphosphine at 250 C. Nanoparticles were isolated by sizeselective precipitation, and particle size was controlled by
changing the relative concentration of precursor and stabilizer.
For example, when a 1+1 molar ratio of cobalt acetate and oleic
acid was used in the synthesis, 68 nm sized cobalt nanoparticles were produced, while increasing the concentration of
stabilizing surfactants by a factor of two yielded smaller 36 nm
nanoparticles. Particle size could be also varied by controlling
the steric bulkiness of the phosphine stabilizers, for example,

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

View Article Online

using tributylphosphine, 1013 nm Co nanoparticles were

generated. Using hydrated nickel acetate as a metal precursor,
nickel nanoparticles with particle sizes in the range 813 nm
were obtained. Co/Ni alloy nanoparticles were also produced
using a mixture of cobalt acetate and nickel acetate through a
similar synthetic procedure.8b Diehl et al. used ferromagnetic
resonance (FMR) techniques to investigate the detailed magnetic characteristics of cobalt nanoparticles with different
crystalline structures.9b The structural characterization of an
ordered assembly of cobalt nanoparticles using high-resolution
transmission electron microscopy was reported by Wang et al.9c
They synthesized multiply twinned fcc cobalt nanocrystals from
the thermal decomposition of dicobalt octacarbonyl in the
presence of a surfactant mixture composed of oleic acid and
tributylphosphine. An HRTEM image of the nanoparticles
revealed complicated interference patterns.8
Reverse micelles, which are water-in-oil droplets stabilized
by a monolayer of surfactant, have been applied as nanoscale
reactors for the synthesis of various nanoparticles.14 Chen et al.
and later the Pileni group reported on the synthesis of relatively
uniform cobalt nanoparticles by the reduction of cobalt salt
inside reverse micelles.15 Chen et al. synthesized cobalt
nanoparticles with particle size in the range 1.84.4 nm by
reducing CoCl2 with NaBH4 in reverse micelles formed using
didodecyldimethylammonium bromide (DDAB). The Pileni
group reported the synthesis of cobalt nanoparticles from the
reaction between a micellar solution containing NaAOT
(sodium bis(2-ethylhexyl)sulfosuccinate) and Co(AOT)2, and
NaBH4 dissolved in NaAOT solution.16 Collisions followed by
exchange between micellar droplets induce chemical reaction of
cobalt ions and sodium borohydride to generate cobalt nanoparticles. The as-synthesized nanoparticles were poorly crystalline while XRD, after heating at 500 C, revealed the
characteristic pattern of fcc cobalt. The average particle size
measured using TEM data was 6 nm with a particle size
distribution s = 9%. A size selection process involving the
extraction of nanoparticles from reverse micelles significantly
reduced the polydispersity of the particle size distribution and
these uniform particles self-assembled to generate two-dimensionally hexagonally ordered arrays.17
Relatively very little work has been done to synthesize
uniform iron nanoparticles. Suslick et al. reported the synthesis
of iron nanoparticles from the sonochemical decomposition of
iron pentacarbonyl in the presence of polyvinylpyrrolidone
(PVP) or oleic acid.18 The sonochemical decomposition of
volatile organometallic compounds has been applied to synthesize various nanostructured materials.19 Transmission electron
micrographs showed that the iron particles ranged in size from
3 to 8 nm. Electron diffraction revealed that the particles asformed are amorphous, and that after in situ electron beam
heating they crystallized to bcc iron. These iron nanoparticles
were readily oxidized to FeO when exposed to air. Gedanken
and coworkers reported the synthesis of amorphous cobalt
nanoparticles by sonicating Co(CO)3(NO) in decane solution in
the presence of oleic acid.20 Spherical 510 nm sized amorphous nanoparticles were produced. One drawback of the
sonochemical process in the synthesis of nanoparticles is its
inability to control particle size.
Very recently, our research group synthesized monodisperse
iron nanoparticles from the high temperature (300 C) aging of
an ironoleic acid metal complex, which was prepared by the
thermal decomposition of iron pentacarbonyl in the presence of
oleic acid at 100 C.21 We were able to synthesize monodisperse
nanoparticles with sizes ranging from 4 to 20 nm without using
any size selection process. Initially, the iron oleate complex was
prepared by reacting Fe(CO)5 and oleic acid at 100 C. Iron
nanoparticles were then generated by aging the iron complex at
300 C. A TEM image of the iron nanoparticles revealed that the

nanoparticles were monodisperse and the electron diffraction

pattern showed that the nanoparticles were nearly amorphous.
The XRD pattern of the sample after being heat-treated in an
argon atmosphere at 500 C revealed a bcc a-iron structure. The
particle size could be controlled from 4 to 11 nm by using
different molar ratios of iron pentacarbonyl to oleic acid. Fig.
3(a) and (b) show TEM images of iron nanoparticles with

Fig. 3 TEM images of iron nanoparticles: (a) three-dimensional array of 7

nm Fe nanoparticles and (b) 11 nm Fe nanoparticles.

particle sizes of 7 and 11 nm, respectively, which were prepared

using 1+2 and 1+3 molar ratios of Fe(CO)5+oleic acid. In order
to produce nanoparticles larger than 11 nm, a reaction mixture
with a 1+4 molar ratio of Fe(CO)5 and oleic acid was used
during the synthesis. However, the particle size of the resulting
nanoparticles was still around 11 nm. We could produce iron
nanoparticles with particle sizes bigger than 11 nm by adding
more iron oleate complex to previously prepared 11 nm iron
nanoparticles, and then aging at 300 C. Using this synthetic
procedure, we were able to tune the particle sizes of the
nanoparticles from 11 to 20 nm.
Using a similar synthetic procedure, we prepared moderately
monodisperse cobalt nanoparticles and successfully applied
them to recyclable catalysts for PausonKhand reactions, which
involve the cycloaddition of alkynes, alkenes and carbon
monoxide to synthesize cyclopentenones.22 We also used
aqueous colloidal cobalt nanoparticles stabilized by sodium
dodecyl sulfate, which were synthesized by the reverse micelle
method described above, for use as recyclable aqueous-phase
PausonKhand catalysts.23

Nanoparticles of alloys of cobalt and iron

Sun et al. synthesized monodisperse ironplatinum nanoparticles by the simultaneous reduction of platinum acetylacetonate (Pt(acac)2) and the thermal decomposition of iron
pentacarbonyl (Fe(CO)5) in the presence of oleic acid and oleyl
amine.24a The composition was adjusted by changing the molar
ratio of the two precursors. Using a 3+2 molar ratio of Fe(CO)5
to Pt(acac)2, Fe48Pt52 nanoparticles were produced, while using
4+1 mixtures Fe70Pt30 nanoparticles were formed. The particle
size could be varied from 3 to 10 nm by adding more precursors
to previously synthesized 3 nm particles, which acted as nuclei.
When the solvent was evaporated slowly, three-dimensional
superlattices are generated, and by controlling the chain lengths
of the stabilizing surfactants, inter-particle spacing could be
tuned. For example, 6 nm Fe48Pt52 nanoparticles were arranged
CHEM. COMMUN., 2003, 927934

929

View Article Online

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

with an interparticle spacing of 4 nm using oleic acid and oleyl

amine as stabilizers, while the same sized particles stabilized
with hexanoic acid and hexyl amine exhibited ~ 1 nm spacing.
Fig. 4 shows the TEM images of 6 nm sized Fe50Pt50

Moreover, these nanoparticles self-assemble to generate superlattices, and when annealed the nanoparticles were transformed
to the tetragonal phase.
Park and Cheon synthesized nanoparticles of CoPt alloys
and CocorePtshell via transmetalation reactions.28 The injection
of 0.25 mmol of Co2(CO)8 into hot toluene solution containing
0.5 mmol of Pt(hfac)2 and oleic acid produced CoPt3 alloy
nanoparticles with a particle size of 1.8 nm (s = 0.1 nm).
Reaction of Pt(hfac)2 with previously prepared 6.33 nm sized
Co nanoparticles produced moderately monodisperse CocorePtshell nanoparticles with a particle size of 6.27 nm (s = 0.58 nm).
The Pt shell thickness and the Co core size were estimated to be
1.82 and 4.75 nm, respectively.
Chaudret and coworkers reported the synthesis of 12 nm
sized bimetallic CoPt nanoparticles from the reaction of
Co(h3-C8H13)(h4-C8H12) and Pt2(dba)3 (dba = dibenzylideneacetone) under dihydrogen.29 The composition of the nanoparticles could be varied by changing the initial molar ratio of
the two organometallic precursors. XRD and TEM analysis
revealed that the Pt rich particles adopted a fcc crystalline
structure while the cobalt rich particles adopted a polytetrahedral arrangement.

Nanoparticles of ferrites

Fig. 4 TEM images of 6 nm sized Fe50Pt50 nanoparticles stabilized with

oleic groups (a) and hexanoic groups (b).24a Reprinted with permission from
reference 24(a). Copyright 2000 American Association for the Advancement of Science.

nanoparticles stabilized with oleic groups (Fig. 4(a)) and

hexanoic groups (Fig. 4(b)), respectively. XRD analysis
revealed that the as-synthesized nanoparticles showed a disordered face centered cubic (fcc) crystal structure and that this
was transformed to an ordered face centered tetragonal structure
when heated at 500 C. Magnetic studies on 4 nm sized Fe52Pt48
nanoparticles showed that the as-synthesized nanoparticles
were superparamagnetic at room temperature and that they
became ferromagnetic after heating at 500 C. A write/read
experiment demonstrated that the annealed 120 nm thick
superlattice of 4 nm Fe48Pt52 nanoparticles supported magnetization reversal transitions at moderate linear densities. Dai et al.
reported the detailed structural characterization of FePt
nanoparticles during annealing using high-resolution transmission electron microscopy.24b Very recently, Rogach and
coworkers grew micrometer-sized colloidal crystals of FePt
nanoparticles using a three-layer technique based on the slow
diffusion of a non-solvent into a concentrated solution of
nanoparticles.25
Using a similar synthetic procedure, Chen and Nikles
reported the synthesis of FePd and CoPt nanoparticles.26 7 nm
sized Co48Pt52 nanoparticles were synthesized by the simultaneous reduction of platinum acetylacetonate and the thermal
decomposition of cobalt tricarbonylnitrosyl (Co(CO)3(NO)) in
the presence of oleic acid and oleyl amine. 11 nm Fe50Pd50
nanoparticles were produced using a similar synthetic method
employing palladium acetylacetonate and iron pentacarbonyl as
precursors. They also synthesized FexCoyPt1002x2y nanoparticles from the simultaneous reduction of cobalt acetylacetonate and platinum acetylacetonate and the thermal
decomposition of iron pentacarbonyl in the presence of oleic
acid and oleyl amine.27 The average particle diameter was 3.5
nm and the particle size distribution was very narrow.
930

CHEM. COMMUN., 2003, 927934

Transition metal oxides constitute one of the most fascinating

classes of inorganic solids, as they exhibit a very wide variety of
structures, properties, and phenomena.30 Oxide materials have
been employed in many important advanced technology areas
due to their many interesting physical properties, including
magnetic, ferroelectric, superconducting, ionic and electrical
conducting characteristics. Nanoparticles of magnetic oxides,
including most representative ferrites, have been studied for
many years for their applications as magnetic storage media and
as ferrofluids. However, the synthesis of monodisperse magnetic oxide nanoparticles was only reported very recently.
The Alivisatos group reported the synthesis of g-Fe2O3
nanoparticles from the thermal decomposition of iron Cupferron
complexes
(Cup:
N-nitrosophenylhydroxylamine,
C6H5N(NO)O2).31 Particle sizes were controlled by either
varying the reaction temperature or by using a varying amount
of the complex. Relatively uniform 410 nm sized g-Fe2O3
nanoparticles were synthesized. Using different metal Cupferron complexes, nanoparticles of Mn3O4 and Cu2O were also
generated.
Very recently, our group reported on the synthesis of
monodisperse and highly-crystalline maghemite nanoparticles,
which did not involve a size selection process.21 The overall
synthetic procedure is depicted in the Fig. 5. Monodisperse iron
nanoparticles were first synthesized, and then they were
transformed to monodisperse g-Fe2O3 nanocrystals by controlled oxidation using trimethylamine N-oxide ((CH3)3NO), as
a mild oxidant. In the current synthetic process, the size
uniformity was determined during the synthesis of metallic
nanoparticles. Transmission electron microscopic images of the
particles showed two- and three-dimensional assembly of
particles, demonstrating the uniformity of these nanoparticles.
Electron diffraction, X-ray diffraction, and high-resolution
transmission electron microscopy proved the highly crystalline
nature of the g-Fe2O3 structures. XPS and Raman spectroscopic
results confirmed the g-Fe2O3 structures. Particle size could be
varied from 4 to 20 nm by altering the experimental parameters.
The dominating size controlling factor was the molar ratio of
iron pentacarbonyl to oleic acid. Fig. 6(a) and (b) show the TEM
images of maghemite nanoparticles with sizes of 7 and 11 nm
that were synthesized by using reaction mixtures with [Fe(CO)5]+[oleic acid] molar ratios of 1+2 and 1+3, respectively.
As described in the synthesis of iron nanoparticles, nano-

View Article Online

Fig. 5 Synthetic procedure for monodisperse g-Fe2O3 nanoparticles.

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

by mild oxidation allowed better control of particle size and

reproducibility.
Using a similar procedure, based on the thermal decomposition of a metalsurfactant complex followed by mild oxidation,
we synthesized highly crystalline and monodisperse cobalt
ferrite (CoFe2O4) nanocrystals.32 In this synthesis, uniform
ironcobalt alloy nanoparticles were first generated from the
thermal decomposition of a metaloleate complex, and then
further oxidized to yield cobalt ferrite nanocrystals. A single
molecular precursor, (h5-C5H5)CoFe2(CO)9, was employed in
the synthesis of the mixed-metaloleate complex. The assynthesized poorly crystalline metallic nanoparticles were then
transformed into cobalt ferrite nanocrystals by oxidizing them
with the mild oxidant, trimethylamine N-oxide. Cobalt ferrite
nanoparticles with particle diameters of 6 nm were synthesized
using starting reaction mixtures with a metal precursor to oleic
acid molar ratio of 1+3. The uniformity of the nanocrystals was
further demonstrated by the formation of a three-dimensional
close-packed superlattice assembly. Larger 9 nm sized nanocrystals, as shown in Fig. 7, were produced when lauric acid was

Fig. 7 TEM images of 9 nm sized cobalt ferrite nanoparticles.32 Reprinted

with permission from reference 32. Copyright 2002 American Chemical
Society.

Fig. 6 TEM images of g-Fe2O3 nanoparticles: (a) 7 nm, (b) 11 nm, (c) 13
nm.21 Reprinted with permission from reference 21. Copyright 2001
American Chemical Society.

particles larger than 11 nm were synthesized by a seeded growth

process, employing addition of more ironoleate complex to the
11 nm sized particles followed by aging at 300 C. The resulting
larger iron nanoparticles were then transformed to iron oxide
nanocrystals by oxidizing with trimethylamine N-oxide. We
could also synthesize monodisperse g-Fe2O3 nanocrystals by
directly injecting Fe(CO)5 into a solution containing both the
surfactant and trimethylamine N-oxide. Fig. 6(c) shows a TEM
image of 13 nm sized g-Fe2O3 nanocrystals synthesized through
the direct injection method. However, the first method employing the synthesis of monodisperse iron nanoparticles followed

used in combination with oleic acid at a molar ratio of [metal

precursor]+[oleic acid]+[lauric acid] = 1+1+2. Monodisperse
nanocrystals were obtained without a further size selection
process. Nanoscale elemental analysis of ten 9 nm cobalt ferrite
nanocrystals by energy-dispersive X-ray spectroscopy (EDS)
showed that an average molar atomic ratio of Fe to Co was 2.3
with a standard deviation of 20%. The EDS data well matched
the bulk elemental analysis results obtained by inductively
coupled plasma atomic emission spectrometry (ICP-AES).
Moumen and Pileni reported on a synthesis of cobalt ferrite
nanoparticles using a microemulsion method,33 which had been
introduced earlier by Klabunde and coworkers.34 Aqueous
methylamine (CH3NH3OH) was added to sodium dodecyl
sulfate solution containing cobalt(II) dodecyl sulfate (Co(DS)2)
and iron(II) dodecyl sulfate (Fe(DS)2) to generate a precipitate.
The size of the nanoparticles was controlled from 2 to 5 nm by
changing the sodium dodecyl sulfate concentration and by
CHEM. COMMUN., 2003, 927934

931

View Article Online

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

maintaining constant concentrations of Co(DS)2, Fe(DS)2, and

methylamine. The particles were not well-isolated and aggregated into networks, and the particle size was not uniform with
a standard deviation of about 30%. Later, detailed structural and
magnetic characterizations of these cobalt ferrite nanoparticles
were reported by Zhang and coworkers.35 From neutron
diffraction studies, they calculated that 78% of tetrahedral sites
(the A site in normal spinel structure with the general formula of
AB2O4) were occupied by Fe3+ cations. The 12 nm nanoparticles had a blocking temperature of 320 K. The same
research group also prepared magnesium ferrite nanoparticles
using a similar microemulsion method, and the magnetic
properties were compared with cobalt ferrite nanoparticles.36
The blocking temperature of cobalt ferrite was found to be at
least 150 K higher than that of the same sized magnesium ferrite
nanoparticles.
Markovich and coworkers synthesized magnetite (Fe3O4)
nanoparticles by reacting aqueous ammonia with an aqueous
solution containing FeCl3 and FeCl2 at a molar ratio of 2+1.37
The aqueous nanoparticles were dispersible in hexane after
being coated with oleic acid. After several cycles of sizeselective precipitation, uniform nanoparticles were obtained.
The LangmuirBlodgett (LB) technique was used to produce
two-dimensional (2D) arrays of these uniform nanoparticles.
Very recently, Sun and Zeng synthesized monodisperse
magnetite nanoparticles from a high temperature reaction of
iron(III) acetylacetonate (Fe(acac)3).38 For example, 4 nm sized
magnetite nanoparticles were synthesized by refluxing a
reaction mixture composed of 2 mmol of Fe(acac)3, 20 mL of
diphenyl ether, 10 mmol of 1,2-hexadecanediol, 6 mmol of oleic
acid, and 6 mmol of oleyl amine. A seed-mediated growth
method was used to synthesize larger nanoparticles. Smaller
sized nanoparticles ( < 4 nm) were mixed with more precursor
materials and the resulting mixture was refluxed to produce the
larger nanoparticles. By controlling the quantity of the nanoparticle seeds, various sized larger nanoparticles were generated. Fig. 8 shows the TEM image of 16 nm Fe3O4

Fig. 8 TEM images of three-dimensional array of 16 nm Fe3O4

nanoparticles.38 Reprinted with permission from reference 38. Copyright
2002 American Chemical Society.

nanoparticles. The as-synthesized magnetite nanoparticles were

transformed to either maghemite (g-Fe2O3) or iron (a-Fe) by
annealing in an oxygen or in an argon atmosphere, respectively.

Shape control of nanoparticles

Recently, the shape control of nanoparticles to generate
anisotropic nanoparticles was recognized as a very important
issue. The Alivisatos group reported the synthesis of rod-shaped
CdSe nanocrystals from the high temperature reaction between
dimethylcadmium and selenium in the presence of a mixed
surfactant solution composed of trioctylphosphine oxide and
hexylphosphonic acid.39 Later the same group reported on the
synthesis of rod-, arrow-, teardrop- and tetrapod-shaped CdSe
932

CHEM. COMMUN., 2003, 927934

nanocrystals by controlling experimental conditions such as the

ratio of surfactants, the injection volume, and the timedependent monomer concentration.40 They explained the formation of anisotropic rod-shaped nanoparticles based on the
relative differences between the growth rates of different crystal
faces at high monomer concentration. Cheon and coworkers
described the synthesis of CdS nanorods and of related
nanostructures in pure hexadecylamine using a single-source
molecular precursor, Cd(S2CNEt2)2.41 The Peng group conducted detailed mechanistic studies on the shape-controlled
synthesis of CdSe nanoparticles, and explained the growth of
anisotropic CdSe nanocrystals on the basis of diffusion
control.42 Wang and coworkers synthesized gold nanorods
using an electrochemical method.43 In this synthesis, the
electrochemical cell consisted of a gold anode and platinum
cathode and the electrolyte solution was a solution of hexadecyltrimethylammonium bromide and tetradecylammonium
bromide. The aspect ratios of the nanorods were controlled by
surfactant molar ratios.
Anisotropic magnetic nanoparticles are expected to offer
many advantages because shape anisotropy would exert a
tremendous influence on their magnetic properties. The tendency for the magnetization to align in a particular direction in
a specimen is denoted magnetic anisotropy. Magnetic anisotropy generally originates from crystal symmetry, shape, stress,
or directed atomic pair ordering.3d,5,44 Of these, crystalline
anisotropy and shape anisotropy are common forms of anisotropy in magnetic materials. Symmetrically shaped nanoparticles, such as spheres, do not have any net shape anisotropy.
However, rod-shaped nanoparticles have shape anisotropy in
addition to crystalline anisotropy, which will increase the
coercivity. Submicrometer-sized acicular (rod-shaped) magnetic particles have been used in commercial particulate
magnetic recording media.
Relatively few anisotropic magnetic nanoparticles have been
reported until 2000. Particle dimensions of the anisotropic
magnetic nanoparticles are too big to exhibit any nanosize effect
and these materials were often extensively agglomerated.45
Gibson and Putzer reported on the synthesis of disk-shaped
cobalt nanoparticles of width 100 nm and thickness 15 nm by
sonicating an aqueous solution of Co2+ and hydrazine.45 The
nanoparticles obtained were single-domain particles. The
particles were not uniform and agglomerated.
Our group reported the synthesis of nearly uniform rodshaped iron nanoparticles from the controlled growth of
monodisperse spherical nanoparticles.46 In the first step of the
synthesis, monodisperse 2 nm spherical iron nanoparticles were
prepared by the thermal decomposition of an organometallic
precursor (Fe(CO)5) in the presence of a stabilizing surfactant,
trioctylphosphine oxide (TOPO). An aliquot of Fe(CO)5 in
trioctylphosphine (TOP) was added to the resulting spherical
nanoparticles in TOPO, at 320 C, and the resulting black
solution was aged for 30 min at 320 C. The nanoparticles so
produced were retrieved by precipitation. The precipitate was
dissolved in 19 mL of pyridine containing 0.5 g of didodecyldimethylammonium bromide (DDAB), and the resulting solution
was refluxed for 12 h. The precipitate formed during the reflux
was removed by centrifugation and the supernatant was
vacuum-dried to yield a black powder. A TEM image of the
sample (Fig. 9) revealed nearly monodisperse rod-shaped
particles of dimensions 2 nm (width) 3 11 nm (length) with a
standard deviation of 5.7% in the length. The electron
diffraction pattern showed the bcc structure of a-Fe. When the
DDAB concentration in pyridine was increased, rod-shaped
particles with higher aspect ratios were obtained. The width of
these nanorods (2 nm) was almost unchanged, but the length
increased to 22 nm (standard deviation of 5.9%) and 27 nm
(standard deviation of 5.5%). The transformation of nano-

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

View Article Online

plained by the high surface tension that reduces the surface-tovolume ratio. The size of nanoparticles could be controlled by
the ratio of surfactant to precursor, which is consistent with
observations in the synthesis of CdSe and other nanoparticles.
The use of two different surfactants seemed to be important for
the synthesis of disk-shaped nanoparticles by modulating the
relative growth rates of different faces. At a fixed oleic acid
concentration, the length of the nanorods (nanodisks) was found
to be proportional to the concentration of TOPO. High
resolution transmission electron microscopic images showed
that the long direction of the nanorods (nanodisks) is parallel to
the (101) planes, demonstrating that TOPO selectively stabilizes the (101) face of Co and decreases its relative growth rate.
The size uniformity was demonstrated by the spontaneous selfassembly of the nanocrystals and by the formation of superstructures such as ribbons of nanorods (nanodisks). A TEM
image of self-assembled 4 nm by 25 nm hcp Co nanorods
(nanodisks) is shown in Fig. 10.
Fig. 9 TEM images of rod-shaped 2 nm (width) 3 11 nm (length) iron
nanoparticles.46 Reprinted with permission from reference 46. Copyright
2000 American Chemical Society.

spheres to nanorods seems to be due to the so-called oriented

attachment mechanism, which is initiated by the relatively
strong binding of DDAB surfactant on the central region of the
growing nanoparticles. After the fusion of two nanospheres, a
third nanosphere will bind on the edge (instead of the central
region where the DDAB is strongly bound), thus generating a
catenated structure. The continued attachment of nanospheres
onto the ends of the growing nanoparticles generates the
unidirectional nanorods. In some parts of TEM image, catenated
nanospheres, which seem to be of the intermediate structure,
were observed. Magnetic studies revealed that the blocking
temperatures of 2 nm 3 11 nm rod-shaped nanoparticles and 2
nm spherical nanoparticles were 110 and 12 K, respectively.
These results are consistent with classical micromagnetic
theory, which predicts that the anisotropy energy is proportional
to the volume of a single particle and the anisotropy constant.
The magnetic anisotropy constant (K) was deduced from the
blocking temperature using the equation K = 25kbTB/V, where
kb is the Boltzman constant and V is the volume of single
nanoparticle.47 The magnetic anisotropy constant of 2 nm
nanospheres was calculated to be 9.1 3 106 erg cm23, and that
of the rod-shaped particles as 1.6 3 107 erg cm23. By treating
the rod-shaped particles as prolate spheroids, the shape
anisotropy constant of 2 nm 3 11 nm rod-shaped nanoparticles
was calculated to be 7.9 3106 erg cm23. When this shape
anisotropy constant was added to the magnetocrystalline
anisotropy constant of the spheres, the overall value was found
to agree well with the experimentally determined anisotropy
constant of the rod-shaped particles.48 The longer iron nanorods
with lengths of 22 nm, 27 nm, and 37 nm were found to be
readily oxidized to FeO.
The Alivisatos group extended the synthetic procedure
developed for CdSe nanorods to synthesize cobalt nanorods,49a
which are now confirmed to be nanodisks as will be discussed
below.49b In this synthesis, dicobalt octacarbonyl, Co2(CO)8, in
o-dichlorobenzene was injected into hot (185 C) o-dichlorobenzene solution containing oleic acid and trioctylphosphine oxide (TOPO). Quenching of the reaction mixture shortly
after this injection then generated relatively uniform sized
cobalt nanorods (nanodisks) with a hcp crystal structure. Upon
further thermal aging of several minutes, the high-energy hcp
rod (disk)-shaped particles were transformed to monodisperse
spherical e-Co nanocrystals. The change of crystal structure at
the relatively low temperature of < 200 C can be explained by
the facile diffusion and phase transitions for nanometer sized
particles. The transformation to spherical particles was ex-

Fig. 10 TEM images of 4 nm (width) 3 25 nm (length) disk-shaped cobalt

nanoparticles.49a Reprinted with permission from reference 49(a). Copyright 2001 American Association for the Advancement of Science.

Very recently, the same research group reported the synthesis

of hcp-Co nanodisks, which were thought to be nanorods as
described above.49b The disk-shape of the nanocrystals was
verified by tilting experiments in TEM. The same workers
developed a much easier procedure to synthesize cobalt
nanodisks, employing a linear amine instead of TOPO. Thermal
decomposition of dicobalt octacarbonyl in the presence of oleic
acid and a linear amine followed by aging generated diskshaped nanoparticles along with spherical nanoparticles. A pure
fraction of hcp Co nanodisks could be obtained after magnetic
separation. The length and diameter can be controlled, primarily
by variation of the reaction time following nucleation as well as
by varying the precursor to surfactant ratios. The average size of
the smaller disks was about 2 3 4 nm and about 4 3 90 nm for
the largest.
Chaudret and coworkers reported on the synthesis of nickel
nanorods from the reduction of Ni(COD)2 (COD = cycloocta1,5-diene) in the presence of hexadecylamine.50 They reported
that increasing concentrations of amine induced the formation
of nanorods. Magnetic studies revealed that these nanorods
showed a saturation magnetization close to the bulk value,
demonstrating the absence of the influence of amine coordination on magnetic properties of the nanorods. Gedanken and
coworkers reported the fabrication of the magnetite nanorods by
the sonication of aqueous iron(II) acetate in the presence of bcyclodextrin.51

Conclusion and outlook

In the past few years, considerable progress has been made in
the synthesis of monodisperse and well-defined structured
magnetic nanoparticles with sizes ranging from 2 to 20 nm. The
outlook of such magnetic nanoparticles is very promising
because these materials will find many important industrial
applications including ultrahigh-density magnetic storage
CHEM. COMMUN., 2003, 927934

933

View Article Online

media and drug-delivery systems. From a synthetic point of

view, there are several interesting research areas worth
pursuing. First, we need a generalized synthetic procedure,
which can synthesize various monodisperse nanoparticles
directly without any further size-selection process. Second,
magnetic nanoparticles with various chemical compositions,
such as CrO2 and lanthanide-containing materials, should be
prepared. Third, more extensive shape-control of magnetic
nanoparticles should be investigated. Forth, a large-area
alignment of magnetic nanoparticles is critical in magnetic
storage media applications. To improve our understanding of
such systems and their applications, a collaborative multidisciplined effort is anticipated.

Published on 03 December 2002. Downloaded on 28/04/2015 17:30:11.

Acknowledgements
The work was supported by the National Creative Research
Initiative Program of the Korean Ministry of Science and
Technology. I thank my students, Dr Sang-Wook Kim,
Jongnam Park, Sang-Jae Park and Yunhee Chung for their
important contributions to the work discussed herein. I also
thank Dr Shouheng Sun and Dr Chris Murray at IBM T. J.
Watson Research Center for very helpful discussion on FePt
alloy nanoparticles.

Notes and references

1 (a) A. P. Alivisatos, Science, 1996, 271, 933; (b) H. Weller, Angew.
Chem., Int. Ed. Engl., 1993, 32, 41; (c) Clusters and Colloids, ed. G.
Schmid, VCH Press, New York, 1994; (d) Nanoscale Materials in
Chemistry, ed. K. J. Klabunde, Wiley-Interscience, New York, 2001; (e)
Nanoparticles and Nanostructured Films, ed. J. H. Fendler, WileyVCH, Weinheim, 1998.
2 (a) S. A. Majetich and Y. Jin, Science, 1999, 284, 470; (b) C. B. Murray,
C. R. Kagan and M. G. Bawendi, Science, 1995, 270, 1335; (c) A. J.
Zarur and J. Y. Ying, Nature, 2000, 403, 65.
3 (a) D. D. Awschalom and D. P. DiVicenzo, Phys. Today, 1995, 4, 43; (b)
K. Raj and R. Moskowitz, J. Magn. Magn. Mater., 1990, 85, 233; (c) I.
M. L. Billas, A. Chatelain and W. A. de Heer, Science, 1994, 265, 1682;
(d) R. C. OHandley, Modern Magnetic Materials, Wiley, New York,
1999.
4 (a) V. E. Fertman, Magnetic Fluids Guidebook: Properties and
Applications, Hemisphere Publishing Co., New York, 1990; (b) B. M.
Berkovsky, V. F. Medvedev and M. S. Krakov, Magnetic Fluids:
Engineering Applications, Oxford University Press, Oxford, 1993; (c)
R. F. Ziolo, E. P. Giannelis, B. A. Weinstein, M. P. OHoro, B. N.
Ganguly, V. Mehrotra, M. W. Russel and D. R. Huffman, Science, 1992,
257, 219.
5 D. L. Leslie-Pelecky and R. D. Rieke, Chem. Mater., 1996, 8, 1770.
6 C. B. Murray, D. J. Norris and M. G. Bawendi, J. Am. Chem. Soc., 1993,
115, 8706.
7 (a) X. M. Lin, H. M. Jaeger, C. M. Sorensen and K. J. Klabunde, J. Phys.
Chem. B, 2001, 105, 3353; (b) S. Stoeva, K. J. Klabunde, C. M.
Sorensen and I. Dragieva, J. Am. Chem. Soc., 2002, 124, 2305.
8 (a) C. B. Murray, S. Sun, W. Gaschler, H. Doyle, T. A. Betley and C. R.
Kagan, IBM J. Res. & Dev., 2001, 45, 47; (b) C. B. Murray, S. Sun, H.
Doyle and T. A. Betley, MRS Bull., 2001, 985.
9 (a) S. Sun and C. B. Murray, J. Appl. Phys., 1999, 85, 4325; (b) M. R.
Diehl, J.-Y. Yu, J. R. Heath, G. A. Held, H. Doyle, S. Sun and C. B.
Murray, J. Phys. Chem. B, 2001, 105, 7913; (c) Z. L. Wang, Z. Dai and
S. Sun, Adv. Mater., 2000, 12, 1944.
10 D. P. Dinega and M. G. Bawendi, Angew. Chem., Int. Ed., 1999, 38,
1788.
11 V. F. Puntes, K. M. Krishnan and A. P. Alivisatos, Appl. Phys. Lett.,
2001, 78, 2187.
12 C. T. Black, C. B. Murray, R. L. Sandstrom and S. Sun, Science, 2000,
290, 1131.

934

CHEM. COMMUN., 2003, 927934

13 C. Petit, T. Cren, D. Roditchev, W. Sacks, J. Klein and M. P. Pileni, Adv.

Mater., 1999, 11, 1198.
14 M. P. Pileni, Langmuir, 1997, 13, 3266.
15 J. P. Chen, C. M. Sorensen, K. J. Klabunde and G. C. Hadjipanayis, J.
Appl. Phys., 1994, 76, 6316.
16 C. Petit, A. Taleb and M. P. Pileni, J. Phys. Chem. B, 1999, 103,
1805.
17 C. Petit, A. Taleb and M. P. Pileni, Adv. Mater., 1998, 10, 259.
18 K. S. Suslick, M. Fang and T. Hyeon, J. Am. Chem. Soc., 1996, 118,
11960.
19 K. S. Suslick, T. Hyeon and M. Fang, Chem. Mater., 1996, 8, 2172.
20 K. V. P. M. Shafi, A. Gedanken and R. Prozorov, Adv. Mater., 1998, 10,
590.
21 T. Hyeon, S. S. Lee, J. Park, Y. Chung and H. B. Na, J. Am. Chem. Soc.,
2001, 123, 12798.
22 S.-W. Kim, S. U. Son, S. S. Lee, T. Hyeon and Y. K. Chung, Chem.
Commun., 2001, 2212.
23 S. U. Son, S. I. Lee, Y. K. Chung, S.-W. Kim and T. Hyeon, Org. Lett.,
2002, 4, 277.
24 (a) S. Sun, C. B. Murray, D. Weller, L. Folks and A. Moser, Science,
2000, 287, 1989; (b) Z. R. Dai., S. Sun and Z. L. Wang, Surf. Sci., 2002,
505, 325.
25 E. Shevchenko, D. Talapin, A. Kornowski, F. Wiehorst, J. Koetzler, M.
Haase, A. Rogach and H. Weller, Adv. Mater., 2002, 14, 287.
26 M. Chen and D. E. Nikles, J. Appl. Phys., 2002, 91, 8477.
27 M. Chen and D. E. Nikles, NanoLett., 2002, 2, 211.
28 J.-I. Park and J. Cheon, J. Am. Chem. Soc., 2001, 123, 5743.
29 T. O. Ely, C. Pan, C. Amiens, B. Chaudret, F. Dassenoy, P. Lecante, M.J. Casanove, A. Mosset, M. Respaud and J.-M. Broto, J. Phys. Chem. B,
2000, 104, 695.
30 (a) C. N. R. Rao and B. Raveau, Transition Metal Oxides, Wiley-VCH,
New York, 2nd edn., 1998; (b) R. M. Cornell and U. Schwertmann, The
iron oxides, VCH, Weinheim, 1996.
31 J. Rockenberger, E. C. Scher and A. P. Alivisatos, J. Am. Chem. Soc.,
1999, 121, 11595.
32 T. Hyeon, Y. Chung, J. Park, S. S. Lee, Y.-W. Kim and B. H. Park, J.
Phys. Chem. B, 2002, 106, 6831.
33 (a) N. Moumen and M. P. Pileni, Chem. Mater., 1996, 8, 1128; (b) N.
Moumen and M. P. Pileni, J. Phys. Chem., 1996, 100, 1867.
34 J. P. Chen, K. M. Lee, C. M. Sorensen, K. J. Klabunde and G. C.
Hadjipanayis, J. Appl. Phys., 1994, 75, 5876.
35 C. Liu, B. Zou, A. J. Rondinone and Z. J. Zhang, J. Am. Chem. Soc.,
2000, 122, 6263.
36 A. J. Rondinone, A. C. S. Samia and Z. J. Zhang, J. Phys. Chem., 1999,
103, 6876.
37 T. Fried, G. Shemer and G. Markovich, Adv. Mater., 2001, 13, 1158.
38 S. Sun and H. Zeng, J. Am. Chem. Soc., 2002, 124, 8204.
39 X. Peng, L. Manna, W. Yang, J. Wickham, E. Scher, A. Kadavanich and
A. P. Alivisatos, Nature, 2000, 404, 59.
40 L. Manna, E. Scher and A. P. Alivisatos, J. Am. Chem. Soc., 2000, 122,
12700.
41 Y.-W. Jun, S.-M. Lee, N.-J. Kang and J. Cheon, J. Am. Chem. Soc.,
2001, 123, 5150.
42 (a) Z. A. Peng and X. Peng, J. Am. Chem. Soc., 2001, 123, 1389; (b) Z.
A. Peng and X. Peng, J. Am. Chem. Soc., 2002, 124, 3343.
43 (a) Y. Y. Yu, S. Chang, C. J. Lee and C. R. C. Wang, J. Phys. Chem. B,
1997, 101, 6661; (b) S.-S. Chang, C.-W. Shih, C.-D. Chen, W.-C. Lai
and C. R. C. Wang, Langmuir, 1999, 15, 701; (c) S. Link, M. B.
Mohamed and M. A. El-Sayed, J. Phys. Chem. B, 1999, 103, 3073.
44 For the general introduction on the magnetic properties of anisotropic
magnetic nanoparticles, please refer to ch. 6 (Magnetism) of ref. 1c.
45 C. P. Gibson and K. J. Putzer, Science, 1995, 267, 1338.
46 S.-J. Park, S. Kim, S. Lee, Z. G. Khim, K. Char and T. Hyeon, J. Am.
Chem. Soc., 2000, 122, 8581.
47 Q. Chen and Z. J. Zhang, Appl. Phys. Lett., 1998, 73, 3156.
48 B. D. Cullity, Introduction to Magnetic Materials, Addison-Wesley,
Reading, MA, 1972.
49 (a) V. F. Puntes, K. M. Krishnan and A. P. Alivisatos, Science, 2001,
291, 2115; (b) V. F. Puntes, D. Zanchet, C. K. Erdonmez and A. P.
Alivisatos, J. Am. Chem. Soc., 2002, 124, 12874.
50 N. Cordente, M. Respaud, F. Senocq, J.-J. Casanove, C. Amiens and B.
Chaudret, NanoLett., 2001, 1, 565.
51 R. V. Kumar, Y. Koltypin, X. N. Xu, Y. Yeshurun, A. Gedanken and I.
Felner, J. Appl. Phys., 2001, 89, 6324.