The role of synthetic parameters in the magnetic behavior of relative large hcp Ni nanoparticles

The role of synthetic parameters in the magnetic behavior
of relative large hcp Ni nanoparticles
A. Kotoulas • M. Gjoka • K. Simeonidis •
I. Tsiaoussis • M. Angelakeris • O. Kalogirou •
C. Dendrinou-Samara
Received: 11 January 2010 / Accepted: 19 April 2010
Springer Science+Business Media B.V. 2010
Abstract The controllable synthesis of relatively
large nickel nanoparticles via thermal decomposition
of nickel acetate tetrahydrate in oleylamine in the
presence of 1-adamantane carboxylic acid (ACA) and
trioctylphosphine oxide (TOPO) is reported. High
crystalline hcp nanoparticles of different sizes have
been prepared at 290 C, whereas at relative lower
temperatures fcc are favored. The particle size was
varying between 50 and 150 nm by properly adjusting
the proportion of the capping ligands. TOPO-to-
ACA ratio was also found to have an influence on the
magnetic properties through the potential formation
of a NiO shell. Pure hcp Ni nanoparticles over 50 nm
in size served as models to illuminate the magnetic
behavior of this metastable hexagonal Ni phase.
Contrary to the net ferromagnetic characteristics of
fcc Ni nanoparticles in the same size range, hexagonal
structured particles exhibit superparamagnetic
behavior at room temperature and a weak ferromagnetic
contribution below 15 K.
Keywords Nickel nanoparticles hcp Nickel
Magnetism Thermal decomposition Surfactants
Introduction
Nickel is a ferromagnetic transition metal that
naturally crystallizes in the face-centered cubic
(fcc) crystal structure. The preparation of nanoparticulate
nickel materials is of great interest due to
their applications in magnetic sensors, electronics,
catalysts (Chen et al. 2007b; Singla et al. 2007) and
recently in biomedicine (Guo et al. 2009). The
synthetic methods of Ni nanoparticles include thermal
decomposition of several precursors, electrochemical
or chemical reduction, and laser pyrolysis
(Hou and Gao 2004). Among them, the most widely
applied is the thermal decomposition of a variety of
precursors such as NiCl2 (Vergara and Mandurga
2002), Ni(acac)2 (acac = acetylacetonate) (Hou and
Gao 2003; Hou et al. 2005; Winnischofer et al. 2008;
Chen et al. 2009) and Ni(COD)2 (COD = cycloocta-
1,5-diene) (Cordente et al. 2003) in the presence of
different kind of surfactants.
Although the cubic structure is the equilibrium
bulk state of Ni, another metastable phase, the
hexagonal close-packed (hcp) structure, is often
A. Kotoulas K. Simeonidis I. Tsiaoussis
M. Angelakeris O. Kalogirou
Department of Physics, Aristotle University
of Thessaloniki, 54124 Thessaloniki, Greece
M. Gjoka
Institute of Materials Science, N.C.S.R. ‘‘Demokritos’’,
Agia Paraskevi, 15310 Athens, Greece
C. Dendrinou-Samara (&)
Department of Chemistry, Aristotle University
of Thessaloniki, 54124 Thessaloniki, Greece
e-mail: samkat@chem.auth.gr
123
J Nanopart Res
DOI 10.1007/s11051-010-9941-2
observed in Ni nanoparticles and thin films (Tzitzios
et al. 2006; Han et al. 2007; Wang et al. 2008;
Mourdikoudis et al. 2009; Luo et al. 2009; Mi et al.
2005). Due to its instability, not much information
and comparable conclusions concerning the mechanisms
of formation and the magnetism of hcp nickel
are available. However, once hexagonal Ni is formed
it does not spontaneously transform into the cubic
phase unless being heated at 400 C, allowing the
study of its properties at room temperature. An
overview of the literature reports indicates that even
small variations in the preparation pathway of Ni
nanoparticles are adequate to change the obtained
atomic structure. Generally, large Ni concentrations
during the nucleation lead to hcp Ni nanoparticles.
This is achieved by relative high heating rates and
temperatures (*260 C) or by the hot-injection of
the nickel precursor (Mourdikoudis et al. 2009).
Intermediate heating conditions may favor the formation
of fcc Ni while low temperatures combined
with a source of hydrogen are sufficient for the
production of Ni hydrides or the stabilization of
vacancies in the crystal structure (Jeon et al. 2005).
The bond distances in the hcp lattice probably
affect the magnetism of the product. Apart from the
heating protocol, the use of different nickel precursors,
solvents, surfactants, and reducing agents also
affects the atomic constants. Depending on the
synthetic procedure and the geometrical characteristics
(size and shape) of hcp Ni nanoparticles several
different results on magnetism have been revealed,
still most of them show a very weak ferromagnetic
character (Han et al. 2007; Mourdikoudis et al. 2009;
Jeon et al. 2006). For instance, hexagonal nickel
nanoparticles smaller than 25 nm, synthesized
through the reduction of nickel chloride by KBH4
in ethylenediamine at 300 C, have been reported to
be ferromagnetic with a saturation magnetization and
coercivity of 7.4 emu/g (or 0.07 lB per atom) and
94 Oe, respectively, at room temperature (Tzitzios
et al. 2006). This is consistent to theoretical
approaches in which the Stoner theory of ferromagnetism
has been applied to 3d-transition Ni metal in
the hexagonal-close-packed (hcp) phase yielding a
stable magnetic moment of 0.59–0.76 lB per atom
(Papaconstantopoulos et al. 1989; Maksimovic and
Vukajlovic 1992). In contrast, larger hcp Ni nanoparticles
(50 nm) synthesized in tetraethylene glycol
using the modified polyol process show non-magnetic
behavior (Chinnasamy et al. 2005) similarly to the
case of hcp Ni thin films (Chen et al. 2007a), while
nanoparticles produced by the hot-injection of
Ni(acac)2 in a reducing solution of hydrazine have
been found to be antiferromagnetic (Jeon et al. 2006).
Herein, we report on the synthesis and the magnetic
behavior of pure hcp Ni nanoparticles stabilized at
diameters higher than 50 nm. For their preparation the
thermal decomposition of nickel acetate tetrahydrate
(Ni(CH3COO)2 4H2O) in the presence of oleylamine
was applied. In order to extent the growth stage and
increase the particle diameter trioctylphosphine oxide
(TOPO) and 1-adamantane carboxylic acid (ACA)
were employed as capping ligands. In a previous study
(Mourdikoudis et al. 2009), we studied the conditions
under which hcp and/or fcc nickel nanoparticles are
formed in primary and tertiary amine solutions, but the
exact contribution of the surfactants in the final size has
not been clarified. Hence, our main concept in that
work was to establish a controllable synthetic procedure
for large hcp Ni nanoparticles by investigating the
influence of the proportions of TOPO and ACA, under
constant reaction conditions (290 C). Furthermore, in
the present study, the performed study of the magnetic
properties of hcp Ni particles between 50 and 100 nm
and the comparison with similar fcc nanostructures
should be considered as more representative due to the
minimization of the contribution of size effects within
this dimensional range. Additionally, we suggest the
presence of a surface oxidation mechanism which
might be responsible for the contradictory literature
reports concerning the magnetic characteristics of hcp
Ni nanoparticles.
Experimental
Synthesis
All the synthetic routes followed the same procedure
of thermal decomposition of Ni(ac)2 4H2O selected
as a relatively cheap Ni precursor with the simultaneous
presence of the surfactants in a solvent. Eight
different samples consisting from hexagonal phase
nanoparticles were examined (denoted Ni0–Ni7).
Their preparation took place in relatively high
temperatures as according to our previous results
(Mourdikoudis et al. 2009), such conditions favor the
hcp Ni structure. In a typical procedure (e.g., sample
J Nanopart Res
123
Ni2) 0.558 g Ni(ac)2 4H2O (2.2 mmol), 0.309 g
TOPO (0.8 mmol), and 0.180 g ACA (1.0 mmol)
were inserted in a flask containing 30 mL oleylamine
and kept under inert argon atmosphere and magnetic
stirring during the procedure. The mixture was
degassed for 20 min before heated to 130 C and
maintained at this temperature for 20 min in order to
evacuate the mixture from the water molecules. After
that period the mixture was further heated at 290 C
at a rate of 3 C/min, remained at this point for 1 h
and allowed to cool at room temperature. The solid
sediment obtained by centrifugation was washed
several times with a mixture of acetone and hexane in
order to remove the byproducts and the excess of
oleylamine. The final product was dispersed and
stored into hexane. The quantities of the surfactants
(mmol TOPO/mmol ACA) ranged from 0.4/1 up to
3.2/1 for the samples Ni1–Ni4 and 1/1.6 and 1/3.2 for
the samples Ni5, Ni6, respectively.
In order to investigate the modification in the
morphology caused by the synergistic action of TOPO
and ACA, two reference samples were synthesized in
the absence of both surfactants (Ni0) and exclusively in
the presence of 3.2 mmol of ACA (Ni7). For comparison
reasons, samples Ni8 and Ni9 were grown at a
lower temperature (225 C) to establish a cubic Ni
structure. Table 1 summarizes the synthesis conditions
and the main characteristics of the samples used in this
study.
Reagents
Nickel acetate tetrahydrate, oleylamine, TOPO, and
ACA were purchased from Aldrich. All chemicals
were used as received without further purification.
Characterization
The crystalline structure of the particles was analyzed
using X-ray powder diffraction (XRD) with a Philips
PW1820 diffractometer, using CuKa radiation. The 2h
angular range was 30–90 , the step size 0.05 and the
step time 3 s. To investigate the morphology, the size,
and the arrangement of the nanoparticles, samples for
transmission electron microscopy (TEM) were prepared
by drop-casting the colloidal dispersions onto
carbon-coated copper grids. TEM images were
acquired on a JEOL 100 Cx microscope, operating at
an acceleration voltage of 100 kV. The size
distribution was based in the statistical analysis of at
least 100 particles for each sample. The microstructure
of the particle was examined by a high-resolutionTEM
(HRTEM) JEOL 2011 microscope working at 200 kV.
Magnetic properties were measured using a superconducting
quantum interference device (SQUID, Quantum
Design MPMS-5), between 5 and 300 K.
Magnetization measurements as a function of temperature
were recorded as follows: Zero-Field-Cooling
(ZFC): the sample was cooled from room temperature
RT down to 5 Kwithout applying a magnetic field, and
then heated back to RT under a magnetic field of
100 Oe; Field Cooling (FC): immediately following
the ZFC measurement the samples were cooled down
to 5 K without removing the field. Thermogravimetric
analysis (TG) and FTIR spectroscopy were used to
examine qualitatively and quantitatively the role of the
surfactant molecules in the self-assembly procedures.
Infrared (IR) spectra (400–4,000 cm-1) were recorded
on a Nicolet FT-IR 6700 spectrometer with samples
prepared as KBr pellets. Thermogravimetric analysis
(TGA) of powder particles were carried out (SETARAM
SetSys-1200) in the range from room temperature
to 700 C at a heating rate of 10 C min-1 under
N2 atmosphere.
Results and discussion
Structural characterization
The crystalline structure of all the samples was
determined by powder XRD. Figure 1 shows a
Table 1 Synthetic parameters and main characteristics of
samples Ni0–Ni9
Sample TOPO
mmol/ACA
mmol
Phase
structure
Reaction
temperature
( C)
Size
(nm)
SD
(%)
Ni0 –/– hcp 290 5.5 14.8
Ni1 0.4/1 hcp 290 81 15.0
Ni2 0.8/1 hcp 290 75 15.9
Ni3 1.6/1 hcp 290 63 15.8
Ni4 3.2/1 hcp 290 58 22.3
Ni5 1/1.6 hcp 290 69 18.0
Ni6 1/3.2 hcp 290 57 18.8
Ni7 –/3.2 hcp 290 108 26.3
Ni8 0.8/1 fcc 225 147 16.0
Ni9 1/3.2 fcc 225 59 27.3
J Nanopart Res
123
comparison diagram of these XRD patterns. Samples
Ni0–Ni7 show a clear high crystalline structure, where
eight different peaks corresponding to the hcp nickel
are observed. On the other hand, the structure of
samples Ni8 and Ni9 was identified as cubic with three
different peaks, corresponding to (111), (200), and
(220) fcc crystalline planes. The high purity of each
sample is verified by the absence of coexisting hcp and
fcc reflections. Another evidence of the sole presence
of hcp structure and the lack of fcc ferromagnetic phase
in samples Ni0–Ni7 was that not even a small amount
of particles could be magnetically decanted from the
reaction mixture, while Ni8 and Ni9 were very easily
separated by approaching a permanent magnet. No
oxidic species were found within the detection limit of
the XRD measurements.
The similarity between the hcp Ni (a = 0.265 nm)
and the hexagonal nickel hydride Ni2H
(a = 0.266 nm) lattice may be considered as a source
of uncertainty. However, the reaction temperature
(290 C) is far higher than the reported limit (160 C)
for the stabilization of nickel hydrides (Jeon et al.
2006) and the complete dehydrogenation of possible
intermediate hydride products should be safely
assumed. Apart from this, the reaction environment
(oleylamine and surfactants) does not favor the
extended hydride formation in any stage of the
procedure.
Based on the observed crystal structures of
Ni0–Ni9, the crystalline structure hcp or fcc hinge
on the reaction temperature and no apparent dependence
appears from the ratio of the ligands. Higher
reaction temperatures favor the fast decomposition of
the Ni precursor, producing high Ni atoms concentrations
and promoting the hexagonal phase formation.
Additionally, the production of the hcp structure
can be considered as a mater of kinetics, under the
rapid formation conditions of thermal decomposition
(Jeon et al. 2006). Thus, the exact control of the phase
structure of Ni nanoparticles can be achieved by
adjusting both thermodynamic and kinetic factors of
the process. These results are in agreement with our
previous report that at high temperatures hcp Ni
nanoparticles are formed, whereas at relative lower
temperatures fcc ones are favored (Mourdikoudis
et al. 2009). Moreover, this conclusion is in accordance
with other scientists (Chen et al. 2007b; Han
et al. 2007) who, by using long-chain amines as
solvents above 260 C prepared the hcp structure,
while at about 240 C they received the cubic
structure. In addition, (Luo et al. 2009) have prepared
hcp Ni nanoparticles with size 89.5 nm by the
reduction of Ni(ac)2, using octadecene as solvent
and TOPO (1 mmol) as ligand.
To get more information about the nanoparticles
structure and morphology samples were examined by
TEM. The comparative study of samples Ni0 and Ni7
was used in order to understand the role of oleylamine,
TOPO, and ACA in the size of the formed
particles. Figure 2a shows that in the absence of
surfactants the pure oleylamine inhibits the growth of
Ni nanoparticles which are stabilized at 5.5 nm. This
is explained by the reduction in surface energy
caused by the considerable amount of oleylamine.
Apart from the dense covering and the strong binding
of the linear oleylamine molecules on the surface of
Fig. 1 XRD patterns of samples Ni0–Ni9. Vertical lines
denote expected diffraction peak positions for hcp Ni (open
triangles) and fcc Ni (full triangles)
Fig. 2 TEM images of samples Ni0 (a) and Ni7 (b),
illustrating the significant size increase in the case of inclusive
ACA use
J Nanopart Res
123
Ni nanoparticles, the formation of a Ni-oleylamine
complex that increases the energy demand for
decomposition and delays the growth stage is another
possibility (Park et al. 2005). The addition of ACA
significantly changes the size of the obtained particles
that reaches 110 nm (Fig. 2b). When ACA is used as
the only surfactant, the large volume of its molecules,
selectively bounded to the growing particle surface,
increases the percentage of unoccupied surface and as
a consequence allows the growth of larger particles.
The particle size is controlled between these two
extreme cases by the addition of TOPO in varying
amounts. TOPO has also bulky end groups and its
coordination with Ni determines the growth rate of
nanoparticles (Hou et al. 2005). Figure 3a–h illustrates
the TEM images of samples Ni1–Ni6 and Ni8–
Ni9 which reveal the influence of the TOPO and
ACA quantity at size, shape, and dispersity of the hcp
and fcc Ni nanoparticles. The mean sizes of the
samples under study are summarized in the graph of
Fig. 4. Generally, hcp Ni samples consist of well
isolated and monodispersed particles with their size
ranging from 58 to 81 nm (Ni1–Ni4) following the
Fig. 3 TEM images of
samples Ni1 (a), Ni2 (b),
Ni3 (c), Ni4 (d), Ni5 (e),
Ni6 (f), Ni8 (g), and Ni9
(h). Electron diffraction
patterns represent a hcp
(Ni1) and a fcc (Ni9)
structure (i)
Fig. 4 Distribution of hcp Ni particle sizes derived by TEM
images for different quantities of TOPO and ACA added
J Nanopart Res
123
decrease of TOPO quantity (3.2–0.4 mmol) for
constant ACA concentration (1 mmol). It is interesting
to mention that the final size is also determined by
the total amount of both surfactants. Keeping TOPO
concentration steady (1 mmol in samples Ni5–Ni6)
the increase of ACA leads to smaller particles. This
fact indicates that TOPO acts in association with
ACA in the growth and the stabilization of the
nanoparticles. It should be noted that in the absence
of ACA, the strong binding between Ni and TOPO
resulted in the formation of a stable complex that
prevented particle production at 290 C.
A similar trend appears in the fcc Ni samples as
well, although the presence of strong magnetic
interactions promotes aggregation effects. Higher
total surfactant concentrations (4.2 mmol for Ni6
and Ni9) bring similar sizes (58 nm) while lower
amounts (1.8 mmol for Ni2 and Ni8) are not sufficient
to overcome the magnetic forces of fcc Ni
samples. All Ni nanoparticles exhibited a polyangular
shape with nummular angles. Such particles’ shape
suggests a possible final coalescence stage between
smaller particles that results in the formation of the
observed polycrystals. The corresponding electron
diffraction patterns verify the XRD conclusions
concerning the purity of hcp or fcc Ni structure in
each sample (Fig. 3i). The size dispersion ranges
between 15 and 27% (Table 1).
The refinement of XRD patterns by the Rietveld
method (Rietveld 1969) was used in order to calculate
the unit cell parameters. Fitting was carried by the
Fullprof software (Rodrı´guez-Carbajal 2009) importing
the hcp Ni (PDF #45-1027) and the fcc Ni (PDF
#04-0850) structures as reference. The observed shift
in most of the XRD spectra indicates the variation of
the lattice parameters compared to the reference
appearing also in the fitting results of Table 2. The
lattice volume reduction, due to internal strain,
reaches 0.48% in the large hcp nanoparticles and
0.15% in the cubic ones. By analyzing the average
peak broadening, the mean crystal size and the level of
internal strain were also calculated. The total broadening,
b, after the removal of the instrument contribution,
can be described (Williamson and Hall 1953):
b2 ¼ ð0:9 k=DxcoshÞ2þð4e tanhÞ2 ð1Þ
where k = 1.54178 A ° is the radiation wavelength, Dx
the mean size of the nanocrystallites, e the internal
strain in % and h the diffraction angle. The experimentally
observed broadening of several peaks in
each spectrum was used to compute the average
particle size Dx and the strain e simultaneously by
using the least squares method (Table 2). The large
deviation of the crystallite size compared to the TEM
observed dimensions (Dt) for the large hcp and fcc
nanoparticles indicates the existence of multiple
crystalline regions into each particle. A similar case
of polycrystallinity was also described on iron oxide
particles[60 nm by Justin Joseyphus et al. 2007. It
should be mentioned that the values calculated by this
method are systematically higher than those resulted
by the application of Scherrer’s equation, indicating
that the particles are not stress free. The internal
strain is reversely proportional to the mean crystal
size. In sample Ni0 intense stress is induced by the
high surface/volume ratio, while in large particles the
strain is a result of multicrystallinity.
FTIR and TG analysis
In order to detect the surface state of Ni nanoparticles
and estimate the organic content in dry samples, IR
spectroscopy and thermogravimetric analysis were
applied. All IR spectra of the samples were recorded
as dry powders and showed similar absorption peaks
but were not so illustrative for the role of the different
ratio of the surfactants. A representative spectrum
(Ni2) is shown in Fig. 5. The weak absorption peak at
2,961 cm-1 is the characteristic of the cis –HC=CH–
arrangement in the oleylamine. The well-defined
Table 2 Crystal parameters based on XRD analysis of samples
Ni0–Ni9
Sample a,b (A°) c(A° ) Dx (Dt) (nm) e (%)
hcp-Ni 2.6515 4.343
Ni0 2.653 (0) 4.328 (2) 7.2 (5.5) 0.0098 (9)
Ni1 2.648 (5) 4.333 (1) 43 (81) 0.0027 (6)
Ni2 2.648 (0) 4.333 (7) 46 (75) 0.0025 (5)
Ni3 2.648 (9) 4.333 (7) 42 (63) 0.0028 (7)
Ni4 2.648 (9) 4.333 (6) 37 (58) 0.0034 (0)
Ni5 2.648 (4) 4.333 (5) 47 (69) 0.0024 (4)
Ni6 2.648 (9) 4.333 (6) 41 (57) 0.0029 (8)
Ni7 2.648 (5) 4.333 (7) 58 (108) 0.0019 (1)
fcc-Ni 3.5238
Ni8 3.522 (6) 26 (147) 0.0057 (4)
Ni9 3.521 (9) 30 (59) 0.0046 (8)
J Nanopart Res
123
bands at 2924.5 and 2851.2 cm-1 are attributed to the
vibrations of asymmetric and symmetric stretching of
methylene groups of the aliphatic chains. The presence
of the strong band at 1,630 cm-1 is due to the
asymmetric stretching of mCOO– of the deprotonated
form of ACA while the symmetric stretching of
mCOO– is considered at about 1,430 cm-1. The low
frequency of these bands compared to the frequency
of the free carboxylic group of pure ACA, 1,693 and
1,452 cm-1, respectively, indicates that ACA molecules
are bounded covalently on the surface of the
nanoparticles and that there are no free ACA
molecules. A set of peaks at about 1,150 cm-1 can
be attributed to TOPO and ACA taking into account
the pure spectra of these two ligands.
The TG curves of Ni2, Ni6, and Ni9 have been
included in Fig. 5 for comparison. It can be seen that
for Ni9 and Ni6 the TG curves are superimposed.
Considering that these samples have been prepared
with the same ratio of surfactants and that their size is
very similar this behavior is expected. In both cases
the total weight loss of organic components is about
6% and occurred up to 300 C. This loss is taking
place in three stages, two of them are about 1.5%
while the third one is about 3% indicative the
presence of three different surfactants. Considering
TG analysis of Ni2 three drops of approximately 3%
can be observed as before up to 300 C. The lower
weight loss in this case certifies the presence of lower
absorbed surfactants than in Ni9 and Ni6.
Magnetic characterization
The magnetic properties were studied for the hcp
samples Ni1, Ni2, Ni4, Ni6, and the fcc Ni9.
Magnetic hysteresis loops of the hcp Ni samples
(Fig. 6) show a lack of saturation, even at 5 T and
coercivity values close to zero. The Langevin
behavior of magnetization in the low-field region of
the 300 K measurement indicates the existence of a
dominant superparamagnetic contribution. At low
temperatures, magnetization can be described as the
sum of an irreversible contribution responsible for the
ferromagnetism and a reversible one corresponding to
the unsaturated behavior at high fields (Tzitzios et al.
2006). The appearance of hysteresis cycle in samples
Ni1 and Ni6 at 5 K reveals ferromagnetic features in
other words a higher degree of order compared to
samples Ni2 and Ni4.
Saturation magnetization, defined after the subtraction
of the linear term, depends on the size of
nanoparticles and aggregation effects as well. The
low values obtained, show that hcp Ni even in
relative large dimensions where nanoscale effects are
not important, is a very feeble ferromagnetic material
that lies in the threshold of antiferromagnetism. This
is a possible explanation for the ambiguous conclusions
of relevant studies (Papaconstantopoulos et al.
1989; Chen et al. 2007a; Park et al. 2005). As shown
in Table 3, magnetization at the maximum field (5 T)
seems to decrease with diameter in samples Ni1, Ni2,
and Ni4. This is probably due to the existence of an
augment percentage of surface spins at smaller
nanoparticles, in comparison with bigger ones, which
Fig. 5 FTIR spectrum of sample Ni2 (a) and TG curve of
samples Ni2, Ni6, and Ni9 (b) compared to the corresponding
measurements for pure oleylamine, TOPO, and ACA
J Nanopart Res
123
reduces the magnetism of each particle. But the
gradual increase of TOPO concentration at Ni1, Ni2,
and Ni4 is also an interpretation, as long as the pacceptor
TOPO has been reported to reduce the
surface magnetism of metal nanoparticles (Cordente
et al. 2001). Accordingly, saturation magnetization
and coercivity were much higher at 5.3 nm hcp Ni
nanoparticles prepared only using oleylamine (Mourdikoudis
et al. 2009), in a sample similar to Ni0.
More reliable conclusions concerning the relation
between the structure and the magnetism of hcp Ni
nanoparticles arise from the temperature versus
magnetization curves (Fig. 6a–d). Samples Ni1 and
Ni6 seem to have much different behavior than Ni2
and Ni4. ZFC curves of the former ones present a
very sharp peak at about 12 K followed, at higher
temperatures, by the immediate diminishing of magnetization
near zero. This is a typical pattern for a
Fig. 6 Magnetic hysteresis
loops at 5 and 300 K and
FC-ZFC curves of the hcp
samples Ni1 (a, e), Ni2 (b,
f), Ni4 (c, g), and Ni6 (d,
h), respectively. The insets
at 300 and 5 K are
expanded around the origin
J Nanopart Res
123
superparamagnetic material showing a mean blocking
temperature at the maximum of ZFC curve. The fact
that the FC branch continues to increase below the
ZFC peak has been reported as another evidence of
superparamagnetism contrary to spin-glass effects
that generally lead to the saturation in this range
(Bitoh et al. 1995). In this case, the low blocking
temperature (TB) and the sharpness of the peak is due
to the low anisotropy of hcp Ni nanoparticles despite
their large size. The anisotropy constant K, calculated
by equation K = 25kBTB/V, where kB is Boltzmann’s
constant, TB the blocking temperature, and V the
volume of one nanoparticle assuming a spherical
shape, is extremely small and its value hardly reaches
102 J m-3 for the two samples. Compared to the
anisotropy constant of cubic Ni (5 9 103 J m-3) this
value is indicative of the very weak magnetic
moments of Ni atoms in hexagonal structure as a
result of the non-equilibrium conditions required for
its formation and the longer distances between the
interacting atoms in this crystallographic system.
Blocking temperature values appear to be independent
of the particles diameter and distribution, as in
the case of smaller hcp Ni nanoparticles with sizes
over 15 nm (Tzitzios et al. 2006; Jeon et al. 2006).
The M–T curves of samples Ni2 and Ni4 are more
complicated. The TB, ascribed to the ferromagnetic–
superparamagnetic transition of hcp Ni structure,
appears around 10 K, the same as for samples Ni1
and Ni6. However, the splitting between ZFC and FC
magnetization curves occurs above 200 K, while in
smaller nanoparticles (Ni4) a kink at 270 K and an
up-turn of ZFC branch below 7 K also exist. The
observed behavior was attributed to the contribution
of another relaxing phase. Macroscopically there is
no evidence for the presence of another phase except
pure hcp Ni and a possible reason for the unexpected
result would be some regions with lattice defects and
vacancies formed during the synthetic procedure or
even the structural stress. But HRTEM images, like
those of Fig. 7 for Ni4, indicate the presence of a
shell surrounding each Ni nanoparticles. The
observed lattice fringes correspond to (200) planes
of cubic NiO as a result of partial Ni oxidation.
Surface oxidation of the exposed external layer is
anticipated for Ni nanoparticles due to its direct
proximity to oxygen atoms contained in surfactant
molecules. Such disordered antiferromagnetic phase
may result in random magnetic anisotropy and
frustration as well as in the appearance of net
ferromagnetic moment in regions with imperfect
stoichiometry (Del Bianco et al. 2008). The latter
explains the irreversibility indicated by ZFC–FC
curves far over the blocking temperature.
The continuous increase of ZFC curve as the
temperature decreases is also typical for an antiferromagnetic
material in these dimensions. As the
relative thickness of the oxidized layer becomes
higher at smaller particle diameters, the discussed
decrease of saturation magnetization could be also
attributed to an increase of the NiO shell thickness.
Nevertheless, the large particles size and the low hcp
Ni anisotropy exclude possible exchange bias effects.
Finally, the kink in both ZFC and FC curves at 270 K
might suggest the reduction of Neel temperature (TN)
Table 3 Magnetic properties of samples Ni1, Ni2, Ni4, Ni6,
and Ni9
Sample Mmax (emu/gNi) Hc (mT) TB (K)
300 K 5 K 300 K 5 K
Ni1 0.33 5.13 11 95 12
Ni2 0.22 2.86 5 7 6
Ni4 0.18 2.03 18 18 13
Ni6 0.30 4.89 24 71 14
Ni9 53 57 12 38 [300
Fig. 7 High magnification TEM image of sample Ni4
illustrating a thin NiO shell
J Nanopart Res
123
of cubic NiO (524 K) possibly due to finite-size
effects. A slight variation in structure could alter the
d-orbitals splitting and therefore the Neel temperature,
as demonstrated in (Sidorov 1998). In addition,
the comparison with the stoichiometric oxide is not
always safe since Ni–O–Ni bond distances and angles
(that determine the antiferromagnetic superexchange
interactions) could be different when NiO arises from
hcp or fcc Ni oxidation. However, other type of phase
transition can also excite an antiferromagnetic order
at temperatures higher than TN. It is interesting to
note that in the hexagonal layered Ni dioxide
Ag2NiO2 a structural phase transition has been
reported at 270 K concerning the distance between
crystallographic planes (Nozaki et al. 2008). Accordingly,
the abrupt change of magnetization on Fig. 6g
at the same temperature range may be explained by a
similar effect.
Under this prism, the magnetic properties of hcp
Ni nanoparticles depend on the size and on the
synthetic procedure followed. Our results show that
the presence of small quantities from different phases
can induce serious magnetic contributions in the
weak magnetic features of hcp Ni. Considering the
studied nanoparticles as relative large ([50 nm) it
can be expected that similar effects could become
dominant in the magnetic behavior of smaller nanoparticles
(\20 nm) and combined with nanoscale
effects may cause deviations from the actual hcp Ni
magnetic character. In this case, TOPO and ACA
ratio not only determines the final particle diameter
but also influences the extent of surface oxidation.
More specifically, in samples Ni1 and Ni6, where the
magnetic behavior does not support the presence of a
NiO layer, the TOPO-to-ACA ratio lies under 0.4
while for the surface-oxidized samples Ni2 and Ni4
reaches the values of 0.8 and 3.2. It seems that the
strong binding of ACA to the particle surface
provides surface protection against natural oxidation.
In contrast, high TOPO percentage favors disordered
NiO layer formation.
Intrinsically different magnetic properties are
observed in sample Ni9 prepared under similar
conditions and having the same size with Ni6 but
consisting of cubic Ni. As shown in Fig. 8, it displays
net ferromagnetism even at room temperature and
magnetization reaches saturation very easily, even at
1 T. Particularly, the saturation magnetization (Ms)
has a value of 53 Am2/kg at room temperature, very
close to the corresponding of bulk Ni (55 Am2/kg)
and slightly increases at 5 K. The low coercivity
value of 0.018 T measured at 300 K reaches 0.038 T
at 5 K and is consistent with the weak magnetic
anisotropy of nickel compared to Fe or Co. The
ferromagnetic behavior of Ni9 is verified by the shape
of ZFC–FC curves. The blocking temperature stands
over 300 K which is the measurement limit. It should
be mentioned that the kink in ZFC branch at 20 K
could be attributed to the presence of uncompensated
magnetic moments at the surface or to a fraction of
smaller nanoparticles undetectable by TEM means
that undergo a superparamagnetic transition.
Conclusions
The controllable crystal structure (hexagonal or cubic
symmetry) and size of relatively large ([50 nm)
nickel nanoparticles were achieved by varying the
reaction temperature and the proportion of ACA and
TOPO in the thermal decomposition of nickel acetate
tetrahydrate in oleylamine. High temperatures
(290 C) favor the formation of hcp Ni nanoparticles
that present with superparamagnetic features at room
Fig. 8 Magnetic hysteresis
loops at 5 and 300 K and
FC-ZFC curves for Ni9
J Nanopart Res
123
temperature and very weak ferromagnetism under the
blocking temperature of about 12 K. By reducing the
temperature to 225 C cubic Ni nanoparticles with
strong ferromagnetic characteristics even at room
temperature were prepared. The amount of ACA and
especially TOPO was found to determine the growth
mechanism and therefore the observed particle size.
In addition, the ratio of the two surfactants is critical
in order to obtain pure hcp Ni nanoparticles and avoid
surface oxidation influencing their magnetic properties.
Thus, the percentage of TOPO during synthesis
should be kept low. We suggest that even slight
differences in synthetic conditions explain the contradictive
literature results regarding the magnetic
behavior of hcp Ni nanoparticles.
References
Bitoh T, Ohba K, Takamatsu M, Shirane T, Chikazawa S
(1995) Field-cooled and zero-field-cooled magnetization
of superparamagnetic fine particles in Cu97Co3 alloy:
comparison with spin-glass Au96Fe4 alloy. J Phys Soc Jpn
64:1305–1310. doi:10.1143/JPSJ.64.1305
Chen L, Chen J, Zhou H, Zhang D, Wan H (2007a) Synthesis
of dodecanethiol monolayer-stabilized nickel nanoparticles.
Mater Sci Eng A 452–453:262. doi: 10.1016/
j.msea.2006.10.140
Chen Y, Peng DL, Lin D, Luo X (2007b) Preparation and
magnetic properties of nickel nanoparticles via the thermal
decomposition of nickel organometallic precursor in
alkylamines. Nanotechnology 18:505703. doi:10.1088/
0957-4484/18/50/505703
Chen Y, Luo X, She H, Yue GH, Peng DL (2009) Size- and
structure-controlled synthesis and characterization of
nickel nanoparticles. J Nanosci Nanotecnol 9:5157–5163.
doi:10.1166/jnn.2009.1206
Chinnasamy CN, Jeyadevan B, Shinoda K, Tohji K, Narayanasamy
A, Sato K, Hisano S (2005) Synthesis and
magnetic properties of face-centered-cubic and hexagonal-
close-packed Ni nanoparticles through polyol process.
J Appl Phys 97:10J309. doi:10.1063/1.1851951
Cordente N, Respaud M, Senocq F, Casanove MJ, Amiens C,
Chaudret B (2001) Synthesis and magnetic properties of
nickel nanorods. Nano Lett 1:565–568. doi:10.1021/
nl0100522
Cordente N, Amiens C, Chaudret B, Raspaud M, Senocq F,
Casanove MJ (2003) Chemisorption on nickel nanoparticles
of various shapes: influence on magnetism. J Appl
Phys 94:6358. doi:10.1063/1.1621081
Del Bianco L, Boscherini F, Tamisari M, Spizzo F, Vittori
Antisari M, Piscopiello E (2008) Exchange bias and
interface structure in the Ni/NiO nanogranular system. J
Phys D Appl Phys 41:134008–134015. doi:10.1088/0022-
3727/41/13/134008
Guo D, Wu C, Hu H, Wang X, Li X, Chen B (2009) Study on
the enhanced cellular uptake effect of daunorubicin on
leukemia cells mediated via functionalized nickel nanoparticles.
Biomed Mater 4:25013–025021. doi:10.1088/
1748-6041/4/2/025013
Han M, Liu Q, He Z, Song Y, Xu Z, Zhu J (2007) Controllable
synthesis and magnetic properties of cubic and hexagonal
phase nickel nanocrystals. Adv Mater 19:1096–1100.
doi:10.1002/adma.200601460
Hou Y, Gao S (2003) Monodisperse nickel nanoparticles prepared
from a monosurfactant system and their magnetic
properties. J Mater Chem 13:1510. doi:10.1039/b303226d
Hou YL, Gao S (2004) Solvothermal reduction synthesis and
magnetic properties of polymer protected iron and nickel
nanocrystals. J Alloys Compd 365:112. doi:10.1016/
S0925-8388(03)00651-0
Hou Y, Kondoh H, Ohta T, Gao S (2005) Size-controlled
synthesis of nickel nanoparticles. Appl Surf Sci 241:218–
222. doi:10.1016/j.apsusc.2004.09.045
Jeon Y, Lee GH, Park J, Kim B, Chang Y (2005) Magnetic
properties of monodisperse NiHx nanoparticles and
comparison to those of monodisperse Ni nanoparticles. J
Phys Chem B 109:12257–12260. doi:10.1021/jp050489o
Jeon YT, Moon JY, Lee GH, Park J, Chang Y (2006) Comparison
of the magnetic properties of metastable hexagonal
close-packed Ni nanoparticles with those of the stable
face-centered cubic Ni nanoparticles. J Phys Chem B
110:1187–1191. doi:10.1021/jp054608b
Justin Joseyphus R, Kodama D, Matsumoto T, Sato Y, Jeyadevan
B, Tohji K (2007) Role of polyol in the synthesis
of Fe particles. J Magn Magn Mater 310:2393–2395.
doi:10.1016/j.jmmm.2006.10.1132
Luo X, Chen Y, Yue GH, Peng DL, Luo X (2009) Preparation
of hexagonal close-packed nickel nanoparticles via a
thermal decomposition approach using nickel acetate tetrahydrate
as a precursor. J Alloys Compd 476:864–868.
doi:10.1016/j.jallcom.2008.09.117
Maksimovic GD, Vukajlovic FR (1992) Magnetic properties of
hexagonal-close-packed Co and Ni. Physica B 176:227.
doi:10.1016/0921-4526(92)90010-P
Mi Y, Yuan D, Liu Y, Zhang J, Xiao Y (2005) Synthesis of
hexagonal close-packed nanocrystalline nickel by a thermal
reduction process. Mater Chem Phys 89:359.
doi:10.1016/j.matchemphys.2004.09.012
Mourdikoudis S, Simeonidis K, Vitalta-Clemente A, Tuna F,
Tsiaoussis I, Angelakeris M, Dendrinou-Samara C, Kalogirou
O (2009) Controlling the crystal structure of Ni
nanoparticles by the use of alkylamines. J Magn Magn
Mater 321:2723–2728. doi:10.1016/j.jmmm.2009.03.076
Nozaki H, Sugiyama J, Janoschek M, Roessli B, Pomjakushin
V, Keller L, Yoshida H, Hiroi Z (2008) Neutron diffraction
study of layered Ni dioxides: Ag2NiO2. J Phys
Condens Matter 20:104236–104239. doi:10.1088/0953-
8984/20/10/104236
Papaconstantopoulos DA, Fry JL, Brener NE (1989) Ferromagnetism
in hexagonal-close-packed elements. Phys Rev
B 39:2526–2528. doi:10.1103/PhysRevB.39.2526
Park J, Kang E, Son SU, Park HM, Lee MK, Kim J, Kim KW,
Noh HJ, Park JH, Bae CJ, Park JG, Hyeon T (2005)
Monodisperse nanoparticles of Ni and NiO: synthesis,
characterization, self-assembled superlattices, and
J Nanopart Res
123
catalytic applications in the Suzuki coupling reaction. Adv
Mater 17:429–434. doi:10.1002/adma.200400611
Rietveld HM (1969) A profile refinement method for nuclear
and magnetic structures. J Appl Crystallogr 2:65–71.
doi:10.1107/S0021889869006558
Rodrı´guez-Carbajal J (2009) FULLPROF Program: Rietveld,
profile matching & integrated intensity refinement of Xray,
Version 4.70
Sidorov VA (1998) Differential thermal analysis of magnetic
transitions at high pressure: neel temperature of NiO up to
8 GPa. Appl Phys Lett 72:2174. doi:10.1063/1.121312
Singla ML, Negi A, Mahajan V, Singh KC, Jain DVS (2007)
Catalytic behavior of nickel nanoparticles stabilized by
lower alkylammonium bromide in aqueous medium. Appl
Catal Gen 323:51. doi:10.1016/j.apcata.2007.01.047
Tzitzios V, Basina G, Gjoka M, Alexandrakis V, Georgakilas
V, Niarchos D, Boukos N, Petridis D (2006) Chemical
synthesis and characterization of hcp Ni nanoparticles.
Nanotechnology 17:3750–3755. doi:10.1088/0957-4484/
17/15/023
Vergara J, Mandurga V (2002) Structure and magnetic properties
of Ni films obtained by pulsed laser ablation
deposition. J Mater Res 17:2099–2104. doi:10.1557/
JMR.2002.0310
Wang H, Jiao X, Chen D (2008) Monodispersed nickel nanoparticles
with tunable phase and size: synthesis, characterization,
and magnetic properties. J Phys Chem C
112:18793–18797. doi:10.1021/jp805591y
Williamson GK, Hall WH (1953) X-ray line broadening from
filed aluminum and wolfram. Acta Metall 1:22–31.
doi:10.1016/0001-6160(53)90006-6
Winnischofer H, Rocha TCR, Nunes WC, Socolovsky LM,
Knobel M, Zanchet D (2008) Chemical synthesis and
structural characterization of highly disordered Ni colloidal
nanoparticles. ACSnano 2:1313–1319. doi:10.1021/
nn700152w
J Nanopart Res
123