Nickel nanocrystals: Fast synthesis of cubes, pyramids and tetrapods

Article abstract of DOI:10.1039/c2ra22024e

We report the use of organometallic precursors in microwave and microfluidic reactors to synthesize Ni nanoparticles of various shapes: spheres, cubes, trigonal pyramids and tetrapods. Each route has the ability to yield a preferential shape based on kinetic control of crystal growth. Tetrapods are obtained only in a microfluidic reactor. Trigonal pyramids crystallize in the metastable hexagonally close-packed (HCP) phase while spheres, cubes and tetrapods crystallize in the thermodynamically stable face-centred cubic (FCC) phase. The magnetic properties of nanocubes (in particular the magnetic coercive field) are found to be greatly enhanced compared to bulk and previous examples of nanostructured Ni materials.

Full text of DOI:10.1039/c2ra22024e

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Nickel nanocrystals: fast synthesis of cubes, pyramids
and tetrapods3
Cite this: RSC Advances, 2013, 3,
1380
Meital Shviro and David Zitoun*
We report the use of organometallic precursors in microwave and microfluidic reactors to synthesize Ni
nanoparticles of various shapes: spheres, cubes, trigonal pyramids and tetrapods. Each route has the ability
to yield a preferential shape based on kinetic control of crystal growth. Tetrapods are obtained only in a
microfluidic reactor. Trigonal pyramids crystallize in the metastable hexagonally close-packed (HCP) phase
while spheres, cubes and tetrapods crystallize in the thermodynamically stable face-centred cubic (FCC)
phase. The magnetic properties of nanocubes (in particular the magnetic coercive field) are found to be
greatly enhanced compared to bulk and previous examples of nanostructured Ni materials.
Received 3rd September 2012,
Accepted 19th November 2012
DOI: 10.1039/c2ra22024e
www.rsc.org/advances
ods.10,11 Another approach combines soft template with
inorganic precursors and generally refers to surfactant assisted
Introduction
Nickel nanoparticles (NPs) have attracted a large amount of
interest for their magnetic1 and catalytic2 properties due the
abundance of Ni compared to other Pt group metals, which
inspires research on size and shape control of nanostruc-
tures.3 Numerous physical and chemical methods have been
used to produce nickel NPs, such as metal evaporation-
condensation, electrochemical methods, sonochemical synth-
esis, metal salt reduction and organometallic precursor
decomposition.4 The solution processes succeed in producing
monodisperse colloidal NPs with controllable sizes by redu-
cing or decomposing metal precursors.
Transition metal NPs from the Pt group show a tunable
catalytic activity as a function of nanocrystals size and shape in
hydrogenation reactions.5 Catalytic activity and selectivity of
metal NPs are strongly dependent on their morphology due to
the different catalytic activity of exposed planes.6 A model
system for well controlled shape of metal nanocrystals by
solution synthesis has been demonstrated with Pt7 and Pd.8
These noble metals have a large number of shapes which have
been generated and the concept has been extended to Rh.9 The
principles that direct Pd or Rh nanocrystals growth have yet to
be extended to catalytically active 3d metals, especially to Ni.
Indeed, 3d transition metals like Ni are slightly more difficult
to handle compared to Pd, considering the need for a stronger
reducing agent and the occurrence of surface oxidation.
Several strategies have been used to control the shape of
nickel NPs, the hard template approach provides an effective
route to submicron polycrystalline nanowires and nanor-
solvothermal reactions in presence of ligands, ionic surfac-
tants or polymers.12 Ni(II) salt (mainly Ni(acac)2) yields
spherical NPs in the presence of hydride,13,14 long chain
amine,15,16 polyol17,18 or amino-borane reducing agent.19
Heterogeneous nucleation has yielded Ni NPs with shape
control as reported for the synthesis of hexagonal and
triangular Ni nanoplates from the catalytic addition of Fe
(CO)5,20,21 seeding of Ni on Mn powders22 and more generally
by alloying Ni with other transition metals like Pt.23,24
On the other hand, the use of organometallic precursors
has proved to be very efficient in order to synthesize metallic
NPs with a clean metallic surface. Organometallic precursors
like Ni(cyclooctadiene)2 (Ni(COD)2 or Ni(g4-C8H12)2) have
successfully yielded Ni NPs with a broad range of sizes.25–28
Hydrogen assisted decomposition of Ni(g4-C8H12)2 yields a
mixture of Ni faceted NPs and nanorods in presence of an
excess of hexadecylamine ligand.29 Recently, hydrogen
assisted decomposition of metal–organic Ni(acac)2 has also
enabled the formation of Ni nanocubes.30
Reports on size and shape controlled synthesis of nickel
NPs are scarce and worthy of investigation, in particular
compared to the degree of complexity reached with Pd NPs, Ni
nearest neighbor. In order to control the shape of Ni NPs, we
chose to investigate the rapid reaction of organometallic Ni
precursors to stabilize non-spherical shapes. The use of
advanced synthesis tools such as microwave and microfluidic
reactors affords reliable and finely tuned synthesis parameters
to achieve a high level of control on the kinetics of the
reaction. Following this methodology, we report the synthesis
of Ni cubes, pyramids, tetrapods. All syntheses are achieved
with soft amine ligands which do not quench the surface
magnetism and could be removed to yield an active metallic
surface.
Bar Ilan University, Department of Chemistry and Bar Ilan Institute of
Nanotechnology and Advanced Materials (BINA), Ramat Gan 52900, Israel.
E-mail: David.Zitoun@biu.ac.il
Electronic supplementary information (ESI) available: DSC, TGA, NMR, XRD
3
and TEM measurements. See DOI: 10.1039/c2ra22024e
1380 | RSC Adv., 2013, 3, 1380–1387
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
SQUID
Experimental
Magnetic properties were measured using a Superconducting
Quantum Interference Design (SQUID) magnetometer MPMS
XL7, in the range of temperature 2–300 K and of field 0–5 T.
The temperature-dependent susceptibility was measured using
DC procedure. The sample was transferred under nitrogen to
the SQUID chamber to prevent any oxidation. The sample was
cooled to 2 K under zero magnetic field, low magnetic field
(5.0 mT) was applied and data collected from 2 K to 300 K
(zero-field cooled, ZFC). Field Cooled (FC) measurements were
performed from 2 K to 300 K with an applied field during the
cooling. Hysteresis loop was measured at 2 K.
Materials
Ni(COD)2 (STREM, 98%) was used as received.
1-Hexadecylamine (HDA) (ACROS, 90%), dodecylamine
(Acros, 98%) and triphenylphosphine (STREM, 99%) have
been dried, degassed and stored in a glove box. Benzylamine
(ACROS, 99%), mesitylene (ACROS, 97%), oleylamine (ACROS,
80–90%), tri-n-octylphosphine (Acros, 97%) and octylamine
(ACROS, 99%) were dried, degassed and stored on molecular
sieve in the glove box.
Synthesis on a heating plate
In a round flask bottle Ni(COD)2 (0.03 mol L21) was dissolved
in 1-hexadecylamine (10 equiv) and mesitylene, the mixture
was heated to 170 uC for 30 min. The same procedure has been
carried on changing the temperature, reaction time and
ligand. ICP measurements were carried on the powders to
measure Ni content.
Results and discussion
Reaction mechanism
We have studied the decomposition of Ni (g4-C8H12 2
by
)
Differential Scanning Calorimetry (DSC), Thermal Gravimetric
Analysis (TGA) and 1H NMR of the solution after nanoparticles
Microwave oven
filtration (see Fig. S1–S3 ). The DSC data displays an
3
Ni(COD)2 (0.03 mol L21) was dissolved with 1-hexadecylamine
(10 equiv) in mesitylene in a glovebox. The solution was then
transferred to a special sealed vial for microwave reaction
(CEM Discover) and heated at 170 uC (run time 20 min, hold
time from 2 to 60 min). The same procedure has been done
while replacing 1-hexadecylamine with different amines chain,
oleylamine, octylamine, dodecylamine and benzylamine.
exothermic peak, corresponding to the precursor decomposi-
tion, at 80 uC and an endothermic peak, corresponding to the
vaporization of cyclooctadiene, at 150 uC. TGA data confirms
that the complex is decomposed in Ni metal and two
equivalent of cyclooctadiene; the obtained weight loss is very
close to the theoretical 78.6% weight loss (21.4% metal
content) which is consistent with a complete reaction (Fig.
S2 ). When the thermal decomposition is achieved in
3
Microfluidic reactor
1-octadecene, 1H NMR spectrum shows that the by-product
Ni(COD)2 (0.03 mol L21) was dissolved with HDA (10 equiv) in
mesitylene. The solution was injected into a microfluidic cell
(Syrris glass reactor 250 mL, three input) using an Asia pump
(Syrris) specially designed for flow chemistry with a typical rate
of 25 mL min21. The reactor was heated to 170 uC on a
homemade heating stage equipped with two thermocouples.
The same procedure has been done at different injections
rates and temperatures.
consists only in cyclooctadiene (Fig. S3 ). The nickel NPs are
3
therefore synthesized by thermal decomposition of Ni(COD)2
at a temperature above 80 uC in different ligands in order to
control size and shape of the NPs. Moreover, three different
types of reactors are used in order to investigate the effect of
reaction time on the size, shape and crystallographic structure
of the NPs. Apart from the standard flask in a heating mantle,
the reaction is also performed in a microwave oven and in a
microfluidic reactor.
DSC, TGA, NMR, XRD
10 mg of Ni(COD)2 powder was placed in a crucible and the
measurement was performed under nitrogen from 300 K to
Size-controlled nickel nanoparticles by ligand screening
Several sets of experiments allow us to understand the roles of
the ligands in order to control the size of Ni NPs. For these
studies, the concentrations of Ni (g4-C8H12)2 and amine ligand
are held constant, and all the reactions are carried out under
N2 on a Schlenk line or in the glove box to avoid any oxidation.
In a typical hotplate reaction or in a microwave oven, nickel
NPs are synthesized in the presence of 1-hexadecylamine
(HDA) as ligand and mesitylene as solvent. All the reagents are
placed in a round bottom flask and heated at a temperature set
from 80 uC (decomposition of Ni(g4-C8H12)2) to 170 uC, boiling
point of mesitylene) for different reaction times.
The microwave technique has the potential to greatly
contribute to all areas of synthetic chemistry. In particular,
the synthesis of NPs and nanostructures, whose growth is
highly sensitive to the reaction conditions like temperature
and reaction time.31,32 Microwave reactions in a nitrogen filled
tube allow screening of the influence of different parameters
700 K following a heating slope of 10 uC min21 1H NMR
.
spectroscopy: after thermal decomposition of Ni(COD)2 in
1-octadecene, the solution was filtered and diluted in CDCl3.
1H NMR was processed on a 300 MHz Bruker spectrometer.
XRD is collected on a Rigaku Ultima IV XRD System. Ni NPs
were drop casted on a glass slide.
Electron microscopy
Transmission electron microscopy (TEM) images were
obtained from a JEOL-JEM 100SX with 80–100 kV accelerating
voltage, high-resolution TEM (HRTEM) images using a JEOL
JEM 2100 microscope operating at 200 kV. Samples were
prepared by placing a drop of diluted sample on a 400-mesh
carbon-coated copper grid.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 1380–1387 | 1381

View Article Online
Paper
RSC Advances
Fig. 2 TEM and SAED image of Ni nanoparticles in standard flask reaction
(sample 2c: HDA/Ni(COD)2 = 10, T = 170 uC, 2 min).
Fig. 1 Transmission electron microscopy (TEM) images of Ni nanoparticles
synthesized in different ligands with a constant ratio precursor/ligand: 1/2 in a
microwave at T = 170 uC for 30 min a) octylamine, b) dodecylamine, c)
oleylamine and d) hexadecylamine.
Shape-controlled nickel nanoparticles with higher
hexadecylamine content
Particle shape is kinetically rather than thermodynamically
controlled. Therefore, the reaction time and temperature have
been monitored from very short (2 min) to long reaction time
(240 min) at temperatures ranging from 80 to 170 uC in a
standard heating mantle with 10 equivalents of hexadecyla-
on the size and shape of the nanoparticles. The ligands are
chosen for their affinity to Ni (amine group) and the rather
bulky alkyl or aryl group allow an effective stabilization of the
NPs. Trialkylphosphines are also used as bulky ligands to
severely limit particle size. Fig. 1 shows the TEM images of Ni
NPs obtained with 2 equivalents of ligand per Ni organome-
tallic precursor for a 30 min reaction at 170 uC. In all cases, the
colloidal black dispersions are stable for weeks after filtration
of the excess of ligand and redispersion in hexane or toluene.
The NPs crystallize in the face-centred cubic (FCC) phase with
lattice parameters corresponding to bulk Ni (Fm-3m, a =
3.52 Å) (JCPDS-04-0850) as shown from X-ray diffractometry
mine. Some typical results are shown in Fig. 2 and Fig. S5 ,
3
while all the data about synthetic parameters are collected in
Table 1.
The results shine a light on the importance of temperature
and time of reaction as crucial parameters for controlling size
and shape of particles. At low temperatures (below 120 uC), the
syntheses yield spherical Ni NPs, a shape already obtained
with Ni(g4-C8H12)2, as reported in the literature.28 The NPs are
slightly polydispersed (D = 5 ¡ 3 nm) and do crystallize in the
(Fig. S4a ) and electron diffraction (inset Fig. 1b). Particle size
3
depends primarily on the nature of the functional group. Very
small NPs (D = 2 ¡ 1 nm) are obtained in the presence of
triphenylphosphine while the synthesis with trioctylphosphine
FCC lattice (Fig. S5a ). Intermediate temperature (T = 150–160
3
uC) yields elongated NPs, nanorods, as the reaction time is
extended from 30 to 45 min (Fig. S5b ). The nanorods present
3
yields a stable complex (Fig. S4b ).
3
an aspect ratio of 5 6 15 nm and crystallize also in the FCC
lattice. Previous research29 has shown the formation of nickel
nanorods along with the formation of several other shapes.
In the case of our study, temperature and time window have
a crucial role in the formation of nanorods. A prolonged
heating (240 min) at the same temperature leads to the
formation of spherical NPs. Long reaction time involves the
reshaping into spherical NPs of nearly the same volume, which
Ni NPs stabilized with primary amines give the following
sizes: benzylamine (D = 4 ¡ 2 nm), octylamine (D = 3 ¡ 2 nm),
dodecylamine (D = 5 ¡ 3 nm), hexadecylamine (D = 6 ¡ 2 nm)
and oleylamine (D = 4 ¡ 2 nm) (Fig. 1 and Fig. S4 ).
3
Alkylamines yield faceted NPs while increasing the amine
chains from octyl to hexadecylamine allows an increase of
particle size.
While using hexadecylamine, square and triangular cross
section NPs are present along with more spherical NPs
(Fig. 1d). The NP shape being clearly linked to the presence
of ligand, the amine/precursor ratio has been increased and
set to 10/1 in the following experiments.
is expected from thermodynamic considerations (Fig. S5d ).
3
At even higher temperature (T = 170 uC), short time reaction
yields square, triangular and round shape particles (Fig. 2) as
already observed in the microwave experiments at lower ligand
content (Fig. 1d).
1382 | RSC Adv., 2013, 3, 1380–1387
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Table 1 Summary of reaction conditions and results in a standard heating mantle
Sample
HDA/Ni(COD)2
T/uC
Reaction time (min)
Morphology
Size (nm)
2a
2b
2c
2d
10
10
10
10
80–120
150–160
170
30
45
2
Spheres
5 ¡ 3
5*15; 7 ¡ 3
5 ¡ 2
Rods, spheres
Triangles, cubes and spheres
spheres
150–170
240
4 ¡ 3
thermal treatment leads to the same result since shape
controlled particles NPs heated to 150 uC for 30 min under
Kinetic control on the yield of shape controlled NPs in
microwave and microfluidic reactors
inert atmosphere reshape into spherical particles (Fig. S7 ).
3
The investigation of Ni NPs synthesis has raised the issue of
accurately controlling reaction time in order to improve the
yield of a particular shape, which could be achieved in a
microwave oven. The reaction under microwave irradiation is
conducted at 170 uC with Ni(g4-C8H12)2 in the presence of 10
equivalents of HDA in mesitylene. Temperature and pressure
are monitored during the reaction and Fig. 3 presents TEM
images for particles after 2 to 60 min long reaction. Particle
size increases slightly with reaction time: D = 4 ¡ 2 nm (2
min), D = 5 ¡ 2 nm (20 min), D = 6 ¡ 2 nm (20 min), D = 6 ¡ 3
nm (60 min).
TEM images show the effect of reaction time on the amount
of triangular cross-sectioned, cubic and spherical NPs forma-
tion. While time reaction is gradually increased from 2 to 30
min, the amount of the triangular cross-sectioned particles
increases (from 20% to 40%) with a majority of cubic NPs (80
In order to investigate the control on the shape of the
nanoparticles, the reactions have been carried out in a
microfluidic reactor. The flow chemistry in the microfluidic
reactor offers a variety of advantages over conventional
macroscale chemical processes for the synthesis of metal
NPs,33–35 including enhanced mass and heat transfer,36,37
reproducibility,38 potential for sensor integration for in situ
reaction monitoring39, rapid screening of parameters, low
reagent consumption during optimization, and synthesis
parameters independent of the process scale.40 The high
surface-to-volume ratio of the micro reactor channels enables
precise temperature control.
In a typical synthesis, a solution of Ni(g4-C8H12 2
) and 10
equivalents of HDA in mesitylene is injected into the glass
microfluidic reactor at different injection rates and tempera-
tures. Fig. 4 shows a typical TEM image showing the formation
of branched nanoparticles. The synthetic details and summary
of results are collected in Table 2.
The reaction time is determined by the injection rate. Since
the glass reactor is approximately 250 mL, the reaction time can
be compared with the conditions used in the microwave
to 60%) (Fig. 3d). Low magnification is reported in the ESI
3
(Fig. S6). When the reaction is carried out for 60 min, shape
controlled particles transform into spherical NPs as already
observed on a standard heating plate (Fig. 3c). Post synthesis
Fig. 3 TEM images of Ni nanoparticles microwave synthesized in 10 equivalents
of HDA a) 2 min, b) 30 min c) 60 min reaction, d) plot showing the time
dependent shape frequency (& for cubic NPs, m for triangular cross-sectioned
NPs, $ for spherical NPs).
Fig. 4 TEM image of Ni NPs synthesis in microfluidic cell (T = 170 uC for 30 min,
sample 4d from Table 2).
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 1380–1387 | 1383

View Article Online
Paper
RSC Advances
Table 2 Reaction conditions in microfluidic reactor
Sample
Injection rate (mL min21
)
Reaction time (min)
T/uC
Morphology
Size (nm)
4a
4b
4c
4d
100
25
25
8
2.5
10
10
170
170
150
170
Triangles, cubes and spheres
Triangles, cubes and spheres
Triangles, cubes, spheres, tetrapods
Triangles, cubes, spheres, tetrapods
7 ¡ 2
8 ¡ 4
10 ¡ 4
8 ¡ 4 (pod diameter)
30
reaction. For fast reactions (t = 2 min) at 170 uC, the reaction
Triangular shape particles have been synthesized with the
yield cubes and triangular NPs as in MW reaction (Fig. S8a ).
best yield from microwave reaction (40% triangles, T = 170 uC,
30 min). Fig. 6 shows an HRTEM image of triangular nickel
NPs with Fast Fourier Transform (FFT) of a single particle.
(100) and (1–10) reflections correspond only to the hexagonally
close-packed (HCP) phase with standard lattice parameters
(P63/mmc, a = 2.622, c = 4.321) (JCPDS-45-1027). HRTEM
shows a large contrast from the centre to the edge, suggesting
the NPs are trigonal pyramids rather than flat triangles. Most
of the pyramids are truncated on the vertices.
3
Lowering the reaction temperature to 150 uC yields branched
NPs along with smaller triangular NPs (Fig. 4). A majority of
branched NPs present a three-fold symmetry and the higher
contrast in the centre suggests the presence of fourth out-of-
plane arm, resulting in
a tetrahedral symmetry. These
tetrapods coexist with triangular cross-section NPs as already
obtained from microwave reaction. Nevertheless, an extended
reaction time results in the reshaping of these branched NPs
and tetrapods to yield more spherical NPs (Fig. S8d ).
Among all shapes obtained from microfluidic reaction,
branched structures are the most frequent shape. The
branched structure usually consists in three pods of the same
length. TEM image (Fig. 7c) shows a darker region at the
centre suggesting the presence of a fourth pod. Tetrapods
crystallize in the FCC lattice (Fig. 7b). Each tetrapod is found
to be single crystalline and grows along the [111] axis of the
FCC lattice (Fig. 7a and inset). A growth process will be
proposed in the discussion.
Magnetic measurements have been performed both as dry
powders and as frozen dispersion. The latter allow collecting
magnetic properties of diluted samples where all magnetic
interactions (exchange and dipolar) are not significant. In
diluted samples, magnetism arises from a collection of
individual NPs and reflects the behaviour of a single average
NP. We report below the magnetic properties of samples
3
HRTEM and magnetic measurement of the shape controlled
NPs
HRTEM has been performed on representative samples of the
different shapes observed: spherical, square, triangular and
tetrapod shape. Nickel crystallizes in two common compact
lattices: most common is the FCC phase and the HCP phase,
which has been stabilized in low dimensionality solids like
thin films and nanoparticles.
The best yield of nanocubes has been achieved in the
microwave (80% cubes, T = 170 uC, 2 min, with 100 equivalents
of oleylamine) (Fig. 5a). The cubes self-organize in a square
lattice even if the polydispersity prevents long range ordering.
Cubic NPs crystallize in the face-centred cubic (FCC) phase
with lattice parameters corresponding to bulk Ni (Fm-3m, a =
3.52 Å) (JCPDS-04-0850) (electron diffraction, Fig. 5b). SAED
does not show a noticeable decrease of the (111) diffraction
ring, expected for aligned nanocubes, which could be
explained by the shape polydispersity resulting in the disorder.
HRTEM of a single nanocube shows the orientation along the
[200] direction. Faces of the cubes consist in the (200) planes.
Fig. 5 Ni cubic NPs from a 2 min MW reaction: TEM (a); SAED (b) and HRTEM of
Fig. 6 HRTEM of Ni trigonal pyramids; Inset: power spectrum of the area in the
a single NP showing (200) planes (c).
square.
1384 | RSC Adv., 2013, 3, 1380–1387
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
The same measurements have been performed on shape
controlled NPs. The results displayed are representative of
frozen dispersions. 4 nm cubes from microwave reactions
display a strong ferromagnetic behavior at low temperature
(Fig. 8) with a coercive field of 152 mT, nearly 10 times higher
than the bulk values and nearly 4 times higher than the
spherical NPs of equivalent size. The magnetic moment is
found to be 29 A m2 kgNi at T = 2 K, calculated from the
21
volume fraction of Ni NPs obtained from ICP measurements.
ZFC maximum (usually called blocking temperature, TB) is
found to be TB = 23 K and fitting the ZFC/FC curve using a
model of single domain NPs, assuming a magnetic size of 4 ¡
2 nm as obtained from TEM, gives a very large anisotropy
constant of Keff = 7.5 104 J m23
.
Fig. 7 HRTEM of a single tetrapod, inset: respective power spectrum (a), SAED
Discussion on the growth mechanism and magnetic
properties of Ni NPs
(b) and TEM image of a single tetrapod (c).
FCC 4d and 5d transition metals like Pd, Pt, Rh, Ag and Au
have been successfully synthesized with different shapes at the
nanoscale. In all cases, the structure remains FCC and shape
controlled is achieved by controlling the surface energy of the
metal using a chemical grafting on the surface of the NPs,
density of structural defects and redox potential of the
chemical reducing agent. In the present case, the ligand
nature (hexadecylamine) and concentration (2 or 10 equiva-
lents relative to the Ni precursor) have been kept constant
while temperature is the only relevant parameter towards
decomposition of the organometallic precursor. The ligand
effectively controls particle size but does not account for the
stabilization of different shape (cubes, triangles, tetrapods,
rods).
synthesized in hexadecylamine with spherical and cubic
shape.
Magnetic properties of 4 nm spherical NPs are typical of Ni
NPs with a coercive field of 40 mT and a saturation
magnetization of 44 A m2 kgNi (bulk values 17 mT, 58 A
21
m2 kg21) (Fig. 8). This high magnetic moment has been
obtained thanks to the use of organometallic precursor and
soft ligands (amine) which lead to NPs with a magnetic core
and shell.
FCC cubic NPs present [200] facets while HCP truncated
pyramids mainly have exposed [111] facets. Both shapes
transform into spherical NPs upon annealing of colloidal
dispersion.
The cubic shape corresponds to previous research on Pd
nanocubes structures and to a recent work on Ni nanocubes
synthesized by reducing Ni(acac)2 under mild hydrogen
pressure.30 As pointed out in the latest study, the growth
mechanism of the NP is controlled by the kinetics of the
metal–organic precursor.
Tetrapod shape has been obtained only while the reaction is
performed in a microfluidic reactor. The pods crystallize in the
FCC lattice with growth along the [111] axis. Tetrapods are
found to be single-crystalline. Multipods at the nanoscale have
been observed for ionic solids like chalcogenides. The growth
mechanism of CdSe tetrapods proceeds through the nuclea-
tion of a polyhedron (trigonal pyramids, zinc-blende, cubic)
which seeds the overgrowth of wurtzite pods on the exposed
facets leading to tetrapods.41 In the case of oxides, the nucleus
is a truncated octahedra and the growth of hexapods follows
an oriented attachment mechanism.42
Metallic multipods have been synthesized with Rh,9 Au43
and Pt44 and formation of metallic multipods and controlled
platonic shapes has been associated with the competitive
growth between [111] and [100] planes in the FCC lattice. The
growth rate of each facet can be adjusted through the
adsorption of a chemical species on the orthogonal facet
(usually a metal ion).
Fig. 8 ZFC/FC (m0H = 5.0 mT) (a) and hysteresis (T = 2 K) (b) curves of NPs
dispersed in toluene (& cubic NPs, $ spherical NPs).
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 1380–1387 | 1385

View Article Online
Paper
RSC Advances
anisotropy of FCC nickel. Our experimental results are
consistent with the enhanced anisotropy displayed by Ni
nanorods synthesized from hydrogen assisted decomposition
of Ni(g4-C8H12)2 in hexadecylamine.47
Compared to a previous report on cubic Ni NPs, the
coercivity field is greatly increased from m0Hc = 10.5 mT for
12 nm Ni nanocubes30 to m0Hc = 151 mT in our measurement
on 4 nm Ni nanocubes. The same trend is observed on the
magnetic measurements with a magnetic moment of 29 A m2
kgNi21 compared to the reported value of 12.4 A m2 kgNi21. Ni
nanocubes display large anisotropy and magnetic moment
consistent with the use of Ni organometallic precursor which
yields large surface magnetism. Therefore, the anisotropy and
magnetic measurement result from core and surface atoms.
Anisotropy is found to be the sum of magnetocrystalline,
shape and surface magnetism. In our study, two shapes
(spherical and cubic) nanoparticles are prepared with the
same organometallic precursor and ligands, so that the
magnetic properties could be assigned to the NPs shape.
Fig. 9 Schematic diagram of the cubes, truncated trigonal pyramids and
tetrapods.
In our synthetic conditions, the formation mechanism of Ni
tetrapods is different from the suggested mechanism for other
metallic nanostructures since no metallic cation is present in
the solution. The formation of tetrapods only occurs in the
microfluidic reactor after the formation of pyramidal NPs. One
may assume pods growth is promoted by the very fast heating
rate, compared to heating plate and microwave oven. The pods
grow in the [111] direction, which corresponds to the exposed
facets of pyramidal NPs (Fig. 9). The growth mechanism goes
through the preferential overgrowth of pods on the vertices of
the pyramids. The most plausible growth mechanism consists
of the formation of pyramidal NPs follows by preferential
overgrowth on the vertices of the structure. Ligands on the
vertices are thought to be more labile and allow mass
transport from the precursor solution to the vertices, resulting
in the growth of pods. The branched NPs then recrystallize in a
single crystalline FCC lattice. According to annealing experi-
ments, this morphology is only stable for a short reaction time
and rearrange into spheres as reported for Pt multipods.43
In the case of Ni, the change in shape is accompanied with a
change in crystallographic phase. Bulk nickel can crystallize
into two phases with relatively close energy; the most stable
FCC phase and metastable HCP phase. Low dimensionality Ni
(thin films and nanoparticles) do crystallize into both lattices
and leads to stable HCP phase. NPs have been found to
crystallize in the HCP lattice when the reaction is performed at
high temperature.45,46
In our study, the structural and magnetic measurements of
HCP pyramidal NPs could not be properly measured since the
highest yield of pyramids achieved is 40%. XRD and SQUID
measurements do not show significant evidence of the
presence of HCP phase, which has been observed only by
HRTEM on multiple samples. The discrepancy could be due to
the nucleation of HCP Ni on a very small scale which prevents
macroscopic observation. HCP Ni is observed only for
pyramidal NPs and not on any other shape. Therefore, we do
believe that the pyramidal NPs obtained result from the
nucleation of HCP Ni whereas cubic NPs are obtained from
FCC nuclei.
Conclusions
Ni NPs have been synthesized from the thermal decomposi-
tion of the organometallic Ni(g4-C8H12)2. Spherical NPS are
formed through addition of alkylamines or triphenylpho-
sphine ligands with a particle size from 3 to 6 nm. Alkylamine
ligands can effectively be used for tailoring the shape of NPs
quickly using a microwave reactor. The reaction yields cubes
up to 80% yield for reactions below 20 min and trigonal
pyramids with a yield of 40% for a 30 min reaction. Pyramidal
NPs crystallize into the metastable HCP phase as evidenced
from HRTEM study. Shape control has also been achieved
through microfluidic reaction leading to the formation of
single crystalline FCC tetrapods which form in the micro-
reactor through the overgrowth on pyramidal NPs. Tetrapods
are stable at room temperature while they rearrange into
spheres upon annealing at 150 uC. The magnetic properties of
4 nm nanocubes are greatly enhanced compared to spherical
NPs with a very high coercive field of 151 mT and anisotropy
constant of 7.5 105 J m23, respectively 10 times and 50%
higher than the bulk values.
Acknowledgements
The authors thank Dr Andras Paszternak for fruitful discus-
sion and critical reading of the manuscript and Dr Judith
Grinblat for HRTEM analysis. The authors thank the EU
through FP7-IEF mobility fellowship for its financial support.
In any case, shape control allows a strong enhancement (ten
times) of the coercivity of Ni in the case of cubes (80%)
compared with bulk values. 4 nm cubic NPs display a coercive
field 4 times larger than the corresponding 4 nm spherical
NPs. The enhancement of magnetic anisotropy (from Keff = 5.0
104 J m23 to Keff = 7.5 104 J m23) is consistent with the shape
anisotropy of cubic NPs, in addition to the magnetocrystalline
References
1 I. S. Lee, N. Lee, J. Park, B. H. Kim, Y.-W. Yi, T. Kim, T.
K. Kim, I. H. Lee, S. R. Paik and T. Hyeon, J. Am. Chem. Soc.,
2006, 128, 10658.
1386 | RSC Adv., 2013, 3, 1380–1387
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
2 J. Park, E. Kang, S. U. Son, H. M. Park, M. K. Lee, J. Kim, K.
W. Kim, H.-J. Noh, J.-H. Park, C. J. Bae, J.-G. Park and
T. Hyeon, Adv. Mater., 2005, 17, 429.
3 M. Kruszynska, H. Borchert, J. Parisi and J. K. Olesiak, J.
Am. Chem. Soc., 2010, 132, 15976.
4 B. L. Cushing, V. L. Kolesnichenko and C. J. O’Connor,
Chem. Rev., 2004, 104, 3893.
5 (a) Y.C. Liu and Y.W. Chen, Ind. Eng. Chem. Res., 2006, 45,
2973.
6 G. A. Somorjai, in Introduction to Surface Chemistry and
Catalysis, Wiley, 2010.
7 J. Chen, T. Herricks and Y. Xia, Angew. Chem., Int. Ed., 2005,
44, 2589.
8 Y. Xiong, H. Cai, B. J. Wiley, J. Wang, M. J. Kim and Y. Xia,
J. Am. Chem. Soc., 2007, 129, 3665.
9 A. J. Biacchi and R. E. Schaak, ACS Nano, 2011, 5, 8089.
10 C.-M. Liu, Y.-C. Tseng, C. Chen, M.-C. Hsu, T.-Y. Chao and
Y.-T. Cheng, Nanotechnology, 2009, 20, 415703.
11 O. Rabin, P. R. Herz, Y. M. Lin, A. I. Akinwande, S.
B. Cronin and M. S. Dresselhaus, Adv. Funct. Mater., 2003,
13, 393.
12 H. Nakamura, Y. Yamaguchi, M. Miyazaki, H. Maeda,
M. Uehara and P. Mulvaney, Chem. Commun., 2002, 2844.
13 Y. Hou and S. Gao, J. Mater. Chem., 2003, 13, 1510.
14 D. S. Sidhaye, T. Bala, S. Srinath, H. Srikanth, P. Poddar,
M. Sastry and B. L. V. Prasad, J. Phys. Chem. C, 2009, 113,
3426.
15 J. Park, E. Kang, S. U. Son, H. M. Park, M. K. Lee, J. Kim, K.
W. Kim, H.-J. Noh, J.-H. Park, C. J. Bae, J.-G. Park and
T. Hyeon, Adv. Mater., 2005, 17, 429.
16 H. Wang, X. Jiao and D. Chen, J. Phys. Chem. C., 2008, 112,
18793.
17 H. Winnischofer, T. C. R. Rocha, W. C. Nunes, L.
M. Socolovsky, M. Knobel and D. Zanchet, ACS Nano,
2008, 2, 1313.
18 L. Bai, J. Fan, Y. Cao, F. Yuan, A. Zuo and Q. Tang, J. Cryst.
Growth, 2009, 311, 2474.
19 O. Metin, V. Mazumder, S. Ozkar and S. Sun, J. Am. Chem.
Soc., 2010, 132, 1468.
20 Y. Leng, Y. Li, X. Li and S. Takahashi, J. Phys. Chem. C,
2007, 111, 6630.
24 J. Zhang, H. Yang, J. Fang and S. Zou, Nano Lett., 2010, 10,
638.
25 (a) D. De Caro and J. S. Bradley, Langmuir, 1997, 13, 3067;
(b) M. Shviro and D. Zitoun, Nanoscale, 2012, 4, 762.
26 T. O. Ely, C. Amiens, B. Chaudret, E. Snoeck, M. Verelst,
M. Respaud and J.-M. Broto, Chem. Mater., 1999, 11, 526.
27 P. Migowski, G. Machado, S. R. Texeira, M. C. M. Alves,
J. Morais, A. Traverse and J. Dupont, Phys. Chem. Chem.
Phys., 2007, 9, 4814.
´
28 E. Ramırez-Meneses, I. Betancourt, F. Morales, V. Montiel-
´
Palma, C. C. Villanueva-Alvarado and M. E. Hernandez-
Rojas, J. Nanopart. Res., 2011, 13, 365.
29 N. Cordente, M. Respaud, F. Senocq, M.-J. Casanove,
C. Amiens and B. Chaudret, Nano Lett., 2001, 1, 565.
30 A. P. Lagrow, B. Ingham, S. Cheong, G. V. M. Williams,
C. Dotzler, M. F. Toney, D. a. Jefferson, E. C. Corbos, P.
T. Bishop, J. Cookson and R. D. Tilley, J. Am. Chem. Soc.,
2012, 134, 855.
31 B. He, J. J. Tan, K.Y. Liew and H. Liu, J. Mol. Catal. A:
Chem., 2004, 221, 121.
32 F. K. Liu, Y. C. Chang, F. H. Ko and T. C. Chu, Mater. Lett.,
2004, 58, 373.
33 Y. Song, J. Hormes and C. S. S. R. Kumar, Small, 2008, 4,
698.
34 C. X. Zhao, L. He, S. Z. Qiao and A. P. J. Middelberg, Chem.
Eng. Sci., 2011, 66, 1463–1479.
35 K. F. Jensen, Chem. Eng. Sci., 2001, 56, 293.
36 K. F. Jensen, MRS Bull., 2006, 31, 101.
37 T. Gervais and K. Jensen, Chem. Eng. Sci., 2006, 61, 1102.
38 H. Nakamura, Y. Yamaguchi, M. Miyazaki, H. Maeda,
M. Uehara and P. Mulvaney, Chem. Commun., 2002, 2844.
39 N. D. Mas, A. Gu, T. Kraus, M. A. Schmidt and K. F. Jensen,
Ind. Eng. Chem. Res., 2005, 44, 8997.
40 H. Nakamura, Y. Yamaguchi, M. Miyazaki, H. Maeda,
M. Uehara and P. Mulvaney, Chem. Commun., 2002, 2844.
41 L. Manna, E. C. Scher and A. P. Alivisatos, J. Am. Chem. Soc.,
2000, 122, 12700.
42 D. Zitoun, N. Pinna, N. Frolet and C. Belin, J. Am. Chem.
Soc., 2005, 127, 15034.
43 S. Chen, Z. L. Wang, J. Ballato, S. H. Foulger and D.
L. Carroll, J. Am. Chem. Soc., 2003, 125, 16186.
44 X. Teng and H. Yang, Nano Lett., 2005, 5, 885.
45 M. Han, Q. Liu, J. He, Y. Song, Z. Xu and J. M. Zhu, Adv.
Mater., 2007, 19, 1096.
21 Y. Leng, Y. Wang, X. Li, T. Liu and S. Takahashhi,
Nanotechnology, 2006, 17, 4834.
22 G. Zhang, S. Sun, R. Li, Y. Zhang, M. Cai and X. Sun, Chem.
Mater., 2010, 22, 4721.
23 J. Zhang and J. Fang, J. Am. Chem. Soc., 2009, 131, 18543.
46 J. Vergara and V. Madurga, J. Mater. Res., 2002, 17, 2099.
47 N. Cordente, C. Amiens, B. Chaudret, M. Respaud,
F. Senocq and M.-J. Casanove, J. Appl. Phys., 2003, 94, 6358.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 1380–1387 | 1387