Surface modification of metallic Co nanoparticles

Article abstract of DOI:10.1016/j.jmmm.2006.10.1203

Monodisperse Co nanoparticles were synthesized by thermal decomposition in the presence of aluminium alkyls yielding air-stable Co nanoparticles after surface passivation. Several procedures for surface modification of these pre-stabilized, metallic Co nanoparticles are presented, including direct anchoring of surface-active functional groups and biocompatible dextran layers as well as silica and polymer coatings. As a result, individually coated nanoparticles as well as microspheres can be obtained.

Full text of DOI:10.1016/j.jmmm.2006.10.1203

ARTICLE IN PRESS
Journal of Magnetism and Magnetic Materials 311 (2007) 92–96
www.elsevier.com/locate/jmmm
Surface modification of metallic Co nanoparticles
a,c
Nina Matoussevitch , Angelika Gorschinski , Wilhelm Habicht , Jens Bolle ,
a
a
b
a,c a,d
¨
Eckhard Dinjus , Helmut Bonnemann , Silke Behrens
a,
Ã
a
Institut f u¨ r Technische Chemie-CPV, Forschungszentrum Karlsruhe, Postfach 3640, 76021 Karlsruhe, Germany
Institut f u¨ r Technische Chemie-WGT, Forschungszentrum Karlsruhe, Postfach 3640, 76021 Karlsruhe, Germany
b
c
Universit a¨ t Heidelberg, Im Neuenheimer Feld 253, 69117 Heidelberg, Germany
Max-Planck-Institut f u¨ r Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 M u¨ lheim, Germany
d
Available online 3 January 2007
Abstract
Monodisperse Co nanoparticles were synthesized by thermal decomposition in the presence of aluminium alkyls yielding air-stable Co
nanoparticles after surface passivation. Several procedures for surface modification of these pre-stabilized, metallic Co nanoparticles are
presented, including direct anchoring of surface-active functional groups and biocompatible dextran layers as well as silica and polymer
coatings. As a result, individually coated nanoparticles as well as microspheres can be obtained.
r 2007 Elsevier B.V. All rights reserved.
Keywords: Nanoparticles; Microspheres; Co nanoparticles; Surface modification; Synthesis
Recently, the combination of the interesting optical,
magnetic, and electrical properties of nanoparticles to-
gether with the high specificity of biomolecular binding
tion, e.g., the saturation magnetization M of bulk cobalt is
s
2
162 A m /kg. Moreover, based on the higher magnetiza-
tion, Co nanoparticles exhibit a higher specific loss power
than the iron oxide ones, which makes them interesting as
potential candidates for applications in hyperthermia
(770 W/g for a Co magnetic fluid in kerosene at 400 kHz,
10 kA/m) [6]. However, stability is an important issue;
usually, metal nanoparticles get easily oxidized when
exposed to air and, thus, loose their magnetic properties.
Co, for example, easily oxidizes in air (free heat of
formation CoO À237.9 KJ/mol, Co O À891.0 KJ/mol).
[
1–3] has shown promising results in various areas of
biomedical in vitro and in vivo application. Magnetic
particles (including microspheres and nanoparticles) have
been widely studied for various fields of application in
biology and medicine, e.g., magnetic drug targeting,
magnetic resonance imaging, hypothermia, immunoassays,
separation/purification of nucleic acids, proteins, and
cells [4,5]. The combination of magnetic particles and
biomolecules has been mainly focused on iron oxide
nanoparticles; reports on biofunctionalized magnetic metal
nanoparticles are rather rare.
Compared to magnetic iron oxide particles, the sig-
nificance of using magnetic metal nanoparticles, e.g., Co,
Fe, FeCo, or FePt is manifold: firstly, they can be prepared
with a narrow size distribution whereas the direct synthesis
of iron oxides often results in rather broad particles size
distributions. For numerous biomedical applications
monodisperse particles are required. Secondly, magnetic
metal nanoparticles exhibit a high saturation magnetiza-
3
4
Co films with a thickness greater than 5 nm were shown to
instantaneously oxidize in air forming a 2.5 nm cobalt
oxide layer [7]. In addition, the synthesis of magnetic metal
nanoparticles is typically performed in organic media and,
thus, phase transfer of the as-prepared particles into the
aqueous phase is important for biomedical application.
Besides, the potential toxicity for in vivo applications of Co
nanoparticles has to be explored. On the other hand, using
magnetic metal nanoparticles, only small amounts of
particles would be required due to the excellent magnetic
characteristics, minimizing potential side effects. More-
over, appropriate surface engineering of the particles not
only allows binding of bioactive molecules to the particle
surface but also generates biocompatible particle surfaces.
Ã
Corresponding author. Tel.: +49 7247826512; fax: +497247822244.
E-mail address: silke.behrens@itc-cpv.fzk.de (S. Behrens).
0304-8853/$ - see front matter r 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmmm.2006.10.1203

ARTICLE IN PRESS
N. Matoussevitch et al. / Journal of Magnetism and Magnetic Materials 311 (2007) 92–96
93
We prepare metallic Co nanoparticles by thermolysis
of Co (CO) in the presence of an aluminium alkyl
catalyst yielding metallic Co nanoparticles. Details of the
synthesis are reported in Ref. [8]. The metal particles were
protected against oxidation by careful surface passivation
A monolayer of molecules that bear a functional thiol
group can be attached to the surface of the pre-stabilized
Co particles, allowing further functionalization with
bioactive molecules in a second step. We have peptized
pre-stabilized Co nanoparticles by L-cysteine ethyl ester
after extensive washing with ethanol and water, forming
stable colloidal solutions in water or ethanol. As shown by
IR spectra and X-ray absorption spectroscopy, the binding
of L-cysteine ethyl ester occurred via the thiol group,
with the amino group remaining free for further functio-
nalization (see Fig. 1) [9]. Other groups have reported
nitrilotriacetic acid-modified FePt nanoparticles obtained
via a functional thiol group which acted as anchoring
group to the particle surface [10]. The L-cysteine ethyl ester-
modified Co nanoparticles provide a basis for further
surface modification and functionalization with biocompa-
tible layers or bioactive molecules. Dextran coating of iron
oxide nanoparticles, for example, has been widely studied
due to its biocompatibility and its surface properties such
as its hydrophilic character. When colloidal solutions of
L-cysteine ethyl ester-modified Co nanoparticles in ethanol
were reacted with dextran (molecular weight 10,000) or
carbomethoxydextran nanoparticle aggregates covered by
the polysaccharide were formed [11]. The TEM images
display the formation of aggregates of 9-nm-sized Co
particles (Fig. 2). The corresponding EFTEM images show
the Co core of the particles covered by an oxygen-rich
2
8
(
pre-stabilization), exposing the Co nanoparticles to a
controlled, small amount of oxygen. The stabilization of
the metallic particles occurs via formation of a CoO layer
and via remaining surface carbonyls as well as surface
carbonates as shown by XPS and XAS analysis [8]. The
mean diameter of the applied pre-stabilized Co nanopar-
ticles was 9 nm. The magnetic nanoparticles and their
surface modification were characterized by a wide range of
techniques including transmission electron microscopy
(
TEM), energy-filtered TEM (EFTEM), infrared (IR)
spectroscopy, atomic emission spectroscopy with induc-
tively coupled plasma (AES–ICP), dynamic light scattering
(DLS), and magnetic measurements by an alternating
gradient magnetometer (AGM).
Surface engineering as well as the conjugation method
are crucial because they may affect not only the particle
stability and biocompatibility but also the biological
functionality of the prepared materials. There are several
ways to modify the particle surface and connect biofunc-
tional molecules to the surface of a nanoparticle. First,
the molecule contains a surface-active group, e.g., a thiol
and is chemisorbed directly on the nanoparticle surface.
Fig. 1. Surface-modified Co nanoparticles: surface modification via a surface-active thiol group using L-cysteine ethyl ester and via a silica network
exhibiting functional silanol groups.
Fig. 2. Co nanoparticles coated by carbomethoxydextran and corresponding EFTEM images (Co map (Co L2,3 edge) (b), and O map (O K edge) (c))
(scale bar 20 nm).

ARTICLE IN PRESS
N. Matoussevitch et al. / Journal of Magnetism and Magnetic Materials 311 (2007) 92–96
94
carbomethoxydextran layer. Dextran coating could be also
performed on unmodified Co nanoparticles in alkaline
solution.
for Si–O–Si bonding, IR spectra display the bands of alkyl
À1
chains at 2854–2966 cm suggesting unhydrolyzed ethoxy
groups at the particle surface. After separation and
extensive washing with ethanol, the silica coated particles
form colloidal solutions in ethanol. As determined by DLS
experiments, the particles were dispersed as individual
particles exhibiting a hydrodynamic radius of around
10 nm and only few aggregates were observed (Fig. 3).
Aside from using peptized nanoparticles together with the
base-catalyzed hydrolysis of TEOS, the Co particles could
By covalent attachment of functional biomolecules to
nanoparticle surfaces, problems of ligand exchange reac-
tions in solution may be overcome. Therefore, the
nanoparticles must be derivatized with a cross-linked
surface shell, which exhibit potential binding sites for
bioactive molecules. This cross-linked surface shell could
consist of polymer or inorganic networks. Such biomole-
cule-particle conjugates are stable, as long as covalent
bonds have to be broken to disintegrate them.
Coating with silica enables covalent functionalization via
its numerous surface silanol groups, easily reacting with
silane coupling agents, e.g., 3-aminoproyl silane as covalent
anchorage for bioactive molecules (Fig. 1). As mentioned
above, Co is easily oxidized, and thus stability is an
important issue during the procedure of silica coating, even
when pre-stabilized Co nanoparticles were applied. Co
nanoparticles were coated by hydrolysis of tetraethoxysi-
be directly covered by TEOS with a very thin SiO layer
2
even without peptizing oleic acid and in the absence of the
ammonium hydroxide catalyst, avoiding complexation and
oxidation reactions with the Co core which might destroy
the protective, pre-stabilizing cobalt oxide shell. TEM and
energy-filtered TEM images show the metallic Co core with
its stabilizing cobalt oxide layer covered by a thin silica
layer (Fig. 4). The saturation magnetization of the sample
2
was 80 A m /kg (Fig. 6c), which is comparable to the
saturation magnetization of the pre-stabilized Co particles
(Fig. 6a) representing about 50% of the Co bulk value. It
has to be stressed that these are raw data, including the
diamagnetic silica shell and the passivating, antiferromag-
netic CoO layer (Co content of the sample as determined
by AES–ICP was 70 wt%). These particles were stable in
water, although they did not form a colloidal solu-
tion. Preliminary experiments with 3-aminopropylsilane
reveal anchored amino groups (1.3% N as determined by
elemental analysis).
¨
lane (TEOS), a method related to the Stober process [12].
In brief, Co particles peptized with oleic acid in toluene
were coated with silica via the base-catalyzed hydrolysis of
TEOS under argon atmosphere [11]. The surface of the Co
particles exhibited an affinity toward silica, and, in contrast
to Au and Ag particles, no primer was applied to promote
the deposition and adhesion of silica on the Co particles.
TEM images show that an amorphous silica layer of about
3
Co particle (Fig. 3, magnetization curve see Fig. 6b).
.5 nm was formed homogeneously around each individual
Depending on the area of application, magnetic micro-
spheres, e.g., for cell separation or individually coated
À1
Besides a broad band at 1052 cm , which is characteristic
1
1
0
0
0
0
0
.2
.0
.8
.6
.4
.2
.0
0
1
10
100
1000 10000 100000
Radius [nm]
2
Fig. 3. TEM image of SiO coated Co particles (scale bar 20 nm) and dynamic light scattering (DLS).
Fig. 4. TEM image of SiO
scale bar 10 nm).
2
coated Co particles (a) and EFTEM images (Co map (Co L2,3 edge) (b), O map (O K edge) (c), and Si map (Si L2,3 edge) (d))
(

ARTICLE IN PRESS
N. Matoussevitch et al. / Journal of Magnetism and Magnetic Materials 311 (2007) 92–96
95
8.0%
7.2%
6.4%
5.6%
4.8%
4.0%
3.2%
2.4%
1.6%
0.8%
0.0%
0.5
1
2
5
10 20
60
Fig. 5. SEM image of Co nanoparticles embedded in poly(MMA–DVB) microspheres (size around 15 mm) and size distribution of poly(VA–DVB)
microspheres (size around 3 mm).
nanoparticles, which are able to penetrate small capillaries
a
1
00
within the body tissue or cell membranes, are preferable.
Due to the high saturation magnetization and the high
magnetophoretic mobility, microspheres equipped with
magnetic metal nanoparticles can be made much smaller
than commonly used ones containing iron oxide.
a
c
8
6
4
2
0
0
0
0
0
b
Silica microspheres containing Co particles could be
prepared by applying a two-step procedure [11]: firstly, the
pre-stabilized Co particles were treated with 3-aminopropyl
silane and active silica, followed by the ammonia-catalysed
hydrolysis of TEOS in a second step. The resulting uniform
silica spheres were about 520 nm in size exhibiting a
-
1.E+06 -8.E+05 -4.E+05 0.E+00
4.E+05
8.E+05
1.E+06
-20
-40
-60
-80
-100
2
saturation magnetization of 21 A m /kg (Fig. 6d; Co
H (A/m)
content as determined by AES–ICP was 24 wt%). The
diameter of the microspheres was dependent on the
reaction time. Further increase of the SiO microsphere
b
4
0
2
e
diameter could be achieved by applying a third step of
ammonia-catalysed hydrolysis of TEOS after washing the
above microspheres.
20
d
Besides silica beads, magnetic polymer microspheres are
very interesting for various applications in biology and
medicine. The surface of such polymer spheres can be
further functionalized with bioactive groups. Pre-stabilized
Co nanoparticles were encapsulated into polymer micro-
spheres by applying a suspension polymerization process
0
-
1.E+12 -8.E+11 -4.E+11 0.E+00
4.E+11
8.E+11
1.E+12
-20
-
40
[
11]. In brief, a magnetic fluid of pre-stabilized Co
H (A/m)
nanoparticles was prepared with KorantinSH (N-oleyl
sarcosine) in dichloromethane/hexane (1:1) and mixed with
the monomers (methylmetacrylate, MMA, or vinylacetate,
VA, and divinylbenzene, DVB) and the initiator (benzoyl
peroxide, BPO). Polyvinylalcohol as stabilizer and sodium
chloride were dissolved in the aqueous phase at 60 1C and a
few microlitre methylene blue were added. The two phases
were vigorously mixed under stirring and the formed
emulsion was heated for 4 h. Depending on the polymer-
ization conditions such as speed of mixing, concentration
of magnetic fluid, and the kind of monomer, different
microspheres were obtained, e.g., the use of MMA instead
of VA yielded 15 mm-sized polymer spheres instead of
Fig. 6. Magnetization curves of pre-stabilized Co nanoparticles (a),
individually SiO coated Co nanoparticles (SiO coating of Co nanopar-
ticles stabilized by oleic acid, 50 wt% Co (b), and SiO coating of Co
nanoparticles in the absence of NH OH, 73 wt% Co (c)) and Co
nanoparticles embedded in SiO microspheres (24 wt% Co (d)) and
2
2
2
4
2
poly(MMA–DVB) microspheres (35 wt% Co (e)).
tion of poly(VA–DVB) microspheres in the range of 3 mm.
The saturation magnetization of poly(MMA–DVB) micro-
2
spheres was 33 A m /kg (Fig. 6e) which is higher than those
reported for Fe O (Co content as determined by AES–ICP
3
4
was 35 wt%) [13].
3
mm-sized spheres. Fig. 5 displays the products after size
In summary, we report the synthesis of air-stable Co
nanoparticles and several procedures for their surface
modification. These methods for surface engineering will
allow future biofunctionalization of the particles with
fractioning by sedimentation. The SEM image in Fig. 5
shows poly(MMA–DVB) microspheres containing Co
nanoparticles in the range of 15 mm and the size distribu-

ARTICLE IN PRESS
N. Matoussevitch et al. / Journal of Magnetism and Magnetic Materials 311 (2007) 92–96
96
proteins, nucleic acid or other biomolecules, opening up
possibilities for biomedical applications. The toxicity of
such Co nanoparticles will be the issue of further studies.
[6] M. Zeisberger, S. Dutz, R. Mu
H. Bonnemann, J. Magn. Magn. Mater., submitted for publication.
7] L. Smardz, U. Kobler, W. Zinn, J. Appl. Phys. 71 (1992) 5199.
8] H. Bonnemann, W. Brijoux, R. Brinkmann, N. Matoussevitch,
N. Waldofner, N. Palina, H. Modrow, Inorg. Chim. Acta 350 (2003)
17.
9] S. Behrens, H. Bo
¨
ller, R. Hergt, N. Matoussevitch,
¨
[
[
¨
¨
¨
6
This work was supported by Deutsche Forschungsge-
meinschaft (DFG), priority program SPP 1104 ‘‘Kolloidale
¨
magnetische Flussigkeiten’’.
[
¨
nnemann, N. Matoussevitch, E. Dinjus,
H. Modrow, N. Palina, M. Frerichs, V. Kempter, W. Maus-
Friedrichs, A. Heinemann, M. Kammel, A. Wiedenmann, L. Pop,
S. Odenbach, E. Uhlmann, N. Bayat, J. Hesselbach, J.M. Guldbakke,
Z. Phys. Chem. 220 (2006) 3.
References
[10] C. Xu, K. Xu, H. Gu, X. Zhong, Z. Guo, R. Zheng, X. Zhang, B. Xu,
J. Am. Chem. Soc. 126 (2004) 3392.
¨
[11] S. Behrens, H. Bonnemann, N. Matoussevitch, A. Gorschinski,
[
1] S. Behrens, E. Dinjus, E. Unger, in: H.S. Nalwa (Ed.), Encyclopedia
of Nanoscience and Nanotechnology, vol. 10, American Scientific
Publ., Stevenson Ranch, CA, 2004, p. 381.
E. Dinjus, W. Habicht, J. Bolle, S. Zinoneva, N. Palina, J. Hormes,
H. Modrow, S. Bahr, V. Kempter, J. Phys.: Condens. Matter 18
(2006) S2543.
[
[
[
[
2] C.M. Niemeyer, Angew. Chem. Int. Ed. 40 (2001) 4129.
3] E. Katz, I. Willner, Angew. Chem. Int. Ed. 43 (2004) 6042.
4] A.K. Gupta, M. Gupta, Biomaterials 26 (2005) 3995.
5] H. Gu, K. Xu, C. Xu, B. Xu, Chem. Commun. (2006) 941.
¨
[12] W. Stober, A. Fink, E. Bohn, J. Colloid Interface Sci. 26 (1968) 62.
[13] C. Yang, H. Liu, Y. Guan, J. Xing, J. Liu, G. Shan, J. Magn. Magn.
Mater. 293 (2005) 187.