Article

pubs.acs.org/cm

Monodisperse Copper Nanocubes: Synthesis, Self-Assembly, and

Large-Area Dense-Packed Films
Hong-Jie Yang, Sheng-Yan He, Hsin-Lung Chen, and Hsing-Yu Tuan*
Department of Chemical Engineering, National Tsing Hua University, 101, Section 2, Kuang-Fu Road, Hsinchu, Taiwan 30013, ROC
*
S Supporting Information

ABSTRACT: In comparison to the well-characterized bot-

tom-up synthesis of Au and Ag nanomaterials, the synthesis of
Cu nanocrystals with well-defined and controllable shapes is
still in need of improvement. Among the many shapes, a cube
covered by six {100} facets can be regarded as a standard
model to study the surface properties of {100} facets. Herein,
we have prepared monodisperse Cu nanoparticles having a
slightly truncated cubic shape with an average edge length of
75.7 nm and a standard deviation of 3.87% by using CuCl as
the precursor, oleylamine as the reaction solvent, and
trioctylphosphine and octadecylamine as shape control agents.
The as-prepared Cu nanocubes tend to self-assemble on transmission electron microscopy grids or silicon substrates. Electron
microscopy and small-angle X-ray scattering reveal that the Cu nanocubes prefer to self-assemble into 2D or 3D rhombohedral
structures (RS). Large-area dense-packed films (1.5 cm × 2.5 cm) composed of monodisperse Cu nanocubes were fabricated by
immersing a Si substrate in a dispersion of dodecanethiol-capped Cu nanocubes in toluene and evaporating the toluene at a
controlled rate while holding the substrate at an angle. The electrical properties of the Cu films with various thickness and
annealing temperatures were studied.

■ INTRODUCTION
Monodisperse metal nanomaterials with well-defined shape
have attracted extensive attention due to their size- and shape-
related properties; thus, their controlled synthesis is a widely
sought goal.1−5 For example, these materials exhibit a unique
optical property known as localized surface plasmon resonance
(LSPR), which occurs when light interacts with free electrons at
the surface, leading to their collective oscillations and
enhancing light absorption and scattering.6 Obtaining uniform
metal nanomaterials is critical to understanding how their size
and shape influence their properties, in order to further
improve their characteristics for specific applications.7−11
Various shapes of silver and gold nanomaterials including
particles,12 cubes,13 wires,14 rods,15 plates,16 and polyhedra17
have been synthesized by chemical methods. The cube shape in
particular has many special properties due to the dominance of
(100) faces on its surface. For example, the catalytic activity of
Pt nanocubes is twice as high as that of commercial Pt catalysts
for oxygen reduction.18 Bulk quantities of metal nanomaterials
with uniform size and shape have been prepared via colloid
synthesis.1 Monodisperse isotropic or anisotropic nanocrystals
may also assemble into 2D or 3D superstructures, which have
been considered as building blocks for self-assembly into
artificial solids.19−22
The self-assembly of nanocrystals into ordered structures
requires a narrow particle size distribution (standard deviation
≤ 5%), uniform shape, compatible surface ligands, and van der
Waals attraction between the nanocrystals.23−28 Usually,
spherical nanocrystals self-assemble into face-centered cubic
(FCC) or hexagonal close packed structures to achieve the
highest packing efficiency, i.e., 74.04%.29−32 Nanocubes with
flat faces and sharp edges tend to assemble in simple cubic
(SC) superlattices.33−35 Recently, Gang and co-workers
reported that nanocubes with rounded corners (truncated)
and nonflat faces (considering the surface ligand shell) can
assemble into rhombohedral structure (RS).23 Dijkstra et al.
report that a system of colloidal hard superballs, with
deformation factor p > 1.1509, prefers to exhibit the C1
phase, which also corresponds to rhombohedral packing.34
The assembly of nanocrystals introduces interesting physical
properties generated from collective interaction between these
nanocrystals.36−38 For example, the absorption spectrum of Au
nanorods shows a red shift of the transverse plasmon band and
a blue shift of the longitudinal plasmon band when the Au
nanorods are assembled in a side-by-side orientation. The
intensity of plasmon coupling is related to the internanorod
distance and the number of nanorods in the assembly.39
Copper (Cu) has the second highest electrical conductivity
of all the metals40 and is more abundant and less expensive
compared to gold and silver. Nanosized Cu exhibits LSPR
because its dielectric function arises from the large negative real
component and small imaginary component.41,42 However,
Received: September 17, 2013
Revised: January 25, 2014
Published: February 11, 2014

© 2014 American Chemical Society 1785 dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
Figure 1. (a, b) TEM images of Cu nanocubes using CuCl as a precursor by hot injection method. (c) UV−visible spectrum of Cu nanocubes
dispersed in toluene at room temperature. The inset shows a photograph of Cu nanocubes solution. (d, e) HRTEM images of a single Cu nanocube.
(f) SAED pattern of the Cu nanocube in (d); the electron beam is perpendicular to the face of the Cu nanocube. (g) Scheme of the Cu nanocube
synthesized by our approach.
control of the shape and size of Cu nanocrystals is still in its
preliminary stages compared to what has been achieved for Ag
and Au in the solution phase. The difficulty in obtaining
uniform Cu nanocrystals arises from their tendency to form
Cu2O or CuO. Shape-controlled synthesis of Cu nanomaterials,
including nanocubes and nanowires, has recently been achieved
using Cu(II) chloride dihydrate in water43 or organic
solvents,44 where the capping ligand plays an important role
in controlling the nucleation and growth of Cu nanostructures.
Controlling the synthesis of Cu nanocubes evidently relies on
finding a suitable capping ligand; further improvement in
monodispersity is still needed.45
Herein, monodisperse Cu nanocubes with an average edge
length of 75.7 nm and a standard deviation of 3.87% were
synthesized by rapidly injecting a mixture of CuCl and
octadecylamine (ODA) in squalane into a hot oleylamine/
trioctylphosphine (TOP) solution at 330 °C. TOP and ODA
are labile capping agents and facilitate the formation of
monodisperse Cu nanocubes. The Cu nanocubes have a
slightly truncated cubic shape, with a deformation factor of p =
2.2. Upon drop casting or evaporation of the solvent from the
dispersion, the monodisperse Cu nanocubes mainly organize
themselves into 2D or 3D supercrystals with RS, as confirmed
by electron microscopy and small-angle X-ray scattering
(SAXS). This method yielded a 1.2-μm thick film of
monodisperse Cu nanocubes over a large area (1.5 cm × 2.5
cm) on a silicon wafer. The optical and electrical properties of
the nanocubes and the nanocube film were studied.
1786
Article
■ EXPERIMENTAL SECTION
Chemicals. All chemicals were used as received. Cu(I) chloride
(99.99%) and trioctylphosphine (TOP, 90%) were purchased from
Alfa Aesar. Oleylamine (OLA, 70%), octadecylamine (ODA, 97%),
squalane (99%), 1-dodecanethiol (98%), and anhydrous toluene
(99.8%) were purchased from Sigma-Aldrich. A stock solution of Cu
was prepared by dissolving 0.1 g (1 mmol) of Cu (I) chloride and 0.2
g (0.74 mmol) of ODA in 2 mL of squalane on a hot plate at 200 °C
for 30 min in an Ar-filled glovebox.
Synthesis of Cu Nanocubes. To synthesize monodisperse Cu
nanocubes, 19 mL of OLA and 1 mL of TOP were mixed in a 50 mL
three-neck flask under inert gas (Ar). The flask was then rapidly heated
to 330 °C. Next, 2 mL of Cu stock solution was quickly injected into a
hot flask. The reaction was held at 330 °C for 3 min. Next, the reaction
mixture was rapidly cooled to room temperature in a water bath to
obtain the Cu nanocubes. The nanocubes were purified by dispersing
the reaction mixture in 20 mL of toluene and then centrifuging at 8000
rpm for 5 min. The supernatant was discarded. A total of 40 mL of
anhydrous toluene was then added to the sediment, and the mixture
was centrifuged at 8000 rpm for 5 min. The washing procedure was
repeated twice to remove unreacted precursors. The isolated Cu
nanocubes were stored in anhydrous toluene before characterization.
Ligand Exchange of Cu Nanocubes. Dodecanethiol-capped Cu
nanocubes were prepared by addition of 1 mL of dodecanethiol into a
Cu nanocubes solution (dispersed in anhydrous toluene) and under
sonication for 10 min. Next, The dodecanethiol-capped nanocubes
were purified by addition of 20 mL of toluene and then centrifuged at
8000 rpm for 5 min. The washing procedure was repeated twice.
Preparation of Cu Nanocube Films and Large Area Cu
Nanocube Superlattice. Large-area Cu nanocube superlattices were
prepared by evaporating the solvent from a dispersion of
dodecanethiol-capped Cu nanocubes in toluene on a silicon substrate
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
at room temperature. To form large area Cu nanocube superlattices,
the Si substrate (1.5 cm × 2.5 cm) was immersed in a vial containing
an appropriate concentration of dodecanethiol-capped Cu nanocubes
dispersed in toluene. The solvent was slowly evaporated by reducing
the system pressure below 150 Torr at 25 °C.46 To make a conductive
Cu nanocube film, a desired amount of dodecanethiol-capped Cu
nanocube solution dispersed in toluene (concentration = 10 mg/mL)
was simply drop-casted on Si or glass substrate (1.5 cm × 2.5 cm) and
dried under air, and the Cu nanocube film was then annealed at 150
°C, 250 °C, or 350 °C (heating rate of 1 °C/min) under a pressure of
0.1 Torr and Ar (50 sccm)/H2 (50 sccm) gas flow for 30 min. The
film thickness can be adjusted by changing the nanocube
concentration.
Characterization. Transmission electron microscopy (TEM),
high-resolution TEM (HRTEM), and selected area electron diffraction
(SAED) were performed on a JEOL JEM 2100F electron microscope
operating at an accelerating voltage of 200 kV. X-ray diffraction (XRD)
measurements were recorded on a Rigaku Ultima IV X-ray
diffractometer using Cu Kα radiation, operating at 40 kV and 20
mA. UV−visible spectra were recorded with a Hitachi U-4100
spectrophotometer at room temperature. Surface topology of the
nanocube films was characterized with a Park Systems XE-70 atomic
force microscope (AFM) in contact mode under ambient conditions.
Scanning electron microscopy (SEM) images were obtained on a
HITACHI-S4700 field-emission SEM with 10 kV accelerating voltage.
TEM, SEM, and XRD samples were prepared by drop casting the Cu
nanocubes dispersed in toluene onto 200 mesh carbon-coated Cu
grids for TEM and SEM or silicon substrates at room temperature for
XRD. SAXS measurements were obtained using monochromatic
radiation of wavelength λ = 1.24 nm on beamline 23A1 at the National
Synchrotron Radiation Research Center (NSRRC) located at Hsin-
Chu, Taiwan. A two-dimensional Mar CCD detector with 512 × 512
pixel resolution was used to record the SAXS pattern. The distance of
sample to the detector is 2423 mm with energy of the X-ray source of
12 keV. The beam center was calibrated using silver behenate powder
standard with the primary reflection peak at 1.076 nm−1. The Fourier-
transform infrared spectra (FTIR) were recorded by Perkin-Elmer
Spectrum RXI FTIR spectrometer with resolution of 4 cm−1. The XPS
measurements were made by a Kratos Axis Ultra DLD using a focused
monochromatic Al X-ray (1486.6 eV) source. The resistivity of the Cu
films (the size of the films was cut into 1 cm × 1 cm) was determined
from sheet resistance using a four-point probe by a Keithley 2400
Source Meter, and the thickness of the Cu films was observed by SEM.
■
RESULTS AND DISCUSSION
Cu nanocubes were synthesized by injecting Cu precursor stock
solution into the mixture of OLA and TOP mixture at 330 °C.
After injection, the color of the solution changed from light
yellow to red-brown, indicating the formation of Cu0 species.
The inset of Figure 1a shows a typical batch of Cu nanocubes
dispersed in toluene; the solution is red-brown in color. Figure
1c shows the UV−vis absorption spectrum of the Cu nanocube
solution. The Cu nanocubes exhibit a major LSPR peak at 588
nm, in the visible region, which agreed with the experimental
observation by other groups.43 Parts a and b of Figure 1 are
TEM images of Cu nanocubes, most of which show truncated
cubic morphology with an average edge length of 75.7 nm and
a standard deviation of about 3.87% (Figure 1b), where a
monodisperse size distribution is defined as having a standard
deviation of ≤5%.47 A schematic of the slightly truncated
nanocube, based on the TEM results, is shown in Figure 1g.
Statistical analysis indicates that 95% of the main product is
cubic in shape, with the remainder of the particles being
spherical or rod-shaped. It is well-known that uniform size and
shape are critical for the self-assembly of structures with long-
range order. The Cu nanocubes self-assemble into ordered 2D
arrays on the TEM grids coated by an amorphous carbon film.
1787
Article
It should be noted that there exist many defects in the 2D
arrays, as observed in the TEM image, where spherical or rod-
like nanocrystals induce defects in the film. Figure 1d shows a
TEM image of a single Cu nanocube, with edge lengths in the
range of 74.8 nm and all corners appearing slightly truncated. A
high-resolution TEM image of a single Cu nanocube (Figure
1e) clearly shows continuous fringes with a lattice distance of
0.18 nm, corresponding to the (200) planes in the face-
centered cubic (FCC) structure that is characteristic of Cu, and
confirms that the cube consists of a single crystal. The
continuous fringes parallel to the edges of the cube indicate that
the Cu nanocubes are bound primarily by facets. The SAED
pattern (Figure 1f) acquired from a single Cu nanocube (Figure
2a) viewed along the [001] zone axis confirms that the
nanocube is a single crystal with its surfaces bounded by {100}
facets.
Figure 2. XRD pattern recorded from Cu nanocubes on silicon
substrates by drop casting process. The eliminated part is the signal
obtained from the silicon substrate.
Figure 2 shows the XRD pattern of the nanocubes. The
peaks were assigned to the (111), (200), and (400) planes of
FCC Cu (JCPDS no. 03-1018), respectively. No adventitious
phases such as CuO and Cu2O were found. Supporting
Information Figure S1 shows that the SAED pattern of the Cu
nanocubes is also consistent with that of a FCC structure. The
intensity of the (200) diffraction is significantly higher than that
of other peaks, (111) and (220), which is the strongest peak
and much greater than the value obtained from a conventional
powder sample. This result indicates that the cubes tend to
preferentially orient parallel to the supporting substrates
(silicon substrate) and are consistent with prior observations
by other researchers.44
The surface of FCC single-crystal particles is usually
composed of low-index crystal planes. In addition, the surface
energy of the low-index crystal planes largely determines the
faceting and crystal growth, which has been estimated to occur
in the following sequence {110} > {100} > {111}.48,49 Taking
into account the surface free energy and surface area, this
implies that a single-crystal seed is expected to exist as
truncated octahedra bound by a mix of {111} and {100} planes.
In solution synthesis, the nucleation and growth stages can be
further controlled by employing capping ligands to selectively
stabilize specific crystal planes. The surface state of Cu
nanocubes was measured by FTIR and XPS. Symmetric and
asymmetric alkane CH2 stretches can be detected at ∼2848 and
∼2908 cm−1, as well as symmetric rocking mode of terminal
methyl group at ∼1454 and 1504 cm−1 and C−P stretching
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
peaks at ∼1108 and 1158 cm−1.50 The N−H stretching region
and NH2 scissoring region are visible at 3156 cm−1 and 1632
cm−1, respectively (Figure 3a).51 Additionally, the N 1s and P
Figure 3. (a) FTIR spectra of Cu nanocubes. XPS spectra of Cu
nanocubes: (b) N 1s; (c) P 2p.
2p peaks (Figure 3b,c) also appeared in the XPS spectrum.
These results indicate that the OLA, ODA, and TOP molecules
do exist on the surface of the Cu nanocube. Previous studies of
Cu nanocubes have found that TOP can promote the
formation of single crystal seeds44 and long carbon chain
primary amines such as hexadecylamine (HDA) can fine-tune
the faceting growth rate,43 although their detained mechanism
is not yet completely clear. On the basis of these previous
cases,43,44 we speculate that, in the present study, TOP and
ODA may play a critical role in the nucleation and growth of
Cu nanocubes to form single crystal seeds and selectively
stabilize the {100} facets.
Direct observation of the nanocube assembly was carried out
via SEM and TEM (Supporting Information Figure S2 and S3).
Figure 4a is a SEM image of a Cu nanocube array, where the
Cu nanocubes self-assembled on the Si substrate with no size
selection upon drop casting the nanocubes from toluene in
ambient environment. The Cu nanocubes are close-packed in
the form of mainly rhombohedral arrays and a small part of the
SC arrangement over a short-range. Figure 4b,c also shows that
the nanocube packing is characteristic of RS. Figure 4d shows
the SAED pattern corresponding to over 20 nanocubes from
the sample in Figure 4c, in the projection direction of [001].
The pattern shows fourfold symmetry, which indicates that the
preferred crystal orientation of self-assembled Cu nanocubes is
with the {100} planes parallel to the substrate. Figure 4e shows
a schematic of a rhombohedral array.
Figure 5 shows the self-assembled patterns of Cu nanocubes
that form upon drop casting the Cu colloid sample in different
positions on the carbon coated Cu grid. Parts a and b of Figure
5 show a typical TEM image and schematic drawing of
nanocubes in a RS assembly. Figure 5c shows the FFT image
1788
Article
Figure 4. (a, b) SEM images of Cu nanocubes self-assembled on
silicon substrate. (c) TEM images of Cu nanocubes self-assembled in
rhombohedral array on a carbon-coated Cu TEM grid. (d) SAED
pattern of the Cu nanocube in rhombohedral array, as show in (c). (e)
Scheme of the Cu nanocubes self-assembled into the rhombohedral
structure (RS).
obtained from Figure 5a. It reveals RS consistent with the
simulation results in Figure 5d that show a lattice angle α = 73°
projected from the [001] direction. Parts e and f of Figure 5
show a TEM image and schematic drawing, respectively, of
nanocubes in a RS assembly. Figure 5g shows the FFT image
obtained from Figure 5e, revealing rhombohedral symmetry
consistent with the simulation results that show RS with a
lattice angle α = 73° projected from the [11̅0] direction (Figure
5h). The RS projected from the [111̅] direction is shown in
Supporting Information Figure S4. In summary, the TEM
images of Cu nanocube superlattices show that the Cu
nanocubes tend to self-assemble into rhombohedral packing
structures.
SAXS was used to further identify the structure of the Cu
nanocube superlattice. The diffraction peaks associated with the
superlattice, observed in the SAXS profile (Figure 6) of the Cu
nanocube film, can be indexed to RS, which is the intermediate
phase in a continuous transformation path from a FCC (α =
60) to SC (α = 90) lattice, where α is the lattice angle. FCC-to-
SC transformation has been observed in several systems23,52
For instance, Zhang et al. observed a continuous transformation
from SC to rhombohedral phases during the assembly of Pd
nanocubes by solvent evaporation.23 The ligand thickness and
the particle sphericity both govern the value of α associated
with RS. According to Zhang et al., α can be determined by first
plotting the ratios of the q positions of the diffraction peaks
over q(100) against α via the equation q(hkl)/q(100) = [(h2 +
l2 + k2) sin2 α + 2(hk + kl + lh)(cos2 α − cos α)]1/2/sin α
(Figure 7a). The value of α is then determined by locating in
the plot the value corresponding to the experimentally observed
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials Article

Figure 5. (a) TEM image of of the [100] projection of RS stacking structure and (b) its stacking structure by the drawing. (c) FFT and (d)
simulation of the TEM image in (a). (e) TEM image of of the [110̅ ] projection of RS stacking structure and (f) its stacking structure by the drawing.
(g) FFT and (h) simulation of the TEM image in (e).

Figure 6. SAXS spectra of the superlattices made of Cu nanocrystals.

Inset shows a schematic of a ligand-capped Cu nanocube with slight
truncation.

reduced peak position. The lattice angle thus obtained from the
experimentally observed value of q(200)/q(100) was 72.5°.
Perfect cube-like nanocrystals tend to assemble into a SC
packing pattern, which is the most energetically favorable.53
The Cu nanocubes method contain a Cu core with some
degree of truncation and surface organic ligands (Figure 6
inset) that cause slight deviations from a perfectly cubic shape.
A wide variety of particle shapes, from octahedron (q = 0.5) to
sphere (q = 1), to cube (q = ∞), can be described by a
superball model, |X|2P + |X|2P + |X|2P ≤ 1, where x, y, and z are Figure 7. (a) Plot of the reduced position of the diffraction peak
Cartesian coordinates and p is the deformation parame- [q(111)/q(100), q(200)/q(100), and q(220)/q(100)] vs α (the
ter.23,34,54 The relationship of lattice vectors and lattice angle relationship of lattice vectors and lattice angle). (b) Plots of 1/p vs
for a RS can be expressed as α(p) = arcos[ei·ej/(|ei||ej|)], i, j = 1, α. The bottom images shows schematic pictures of Cu nanocubes with
2, 3 (i ≠ j), where e1 = −2(s + 2−1/2p)i + 2sj + 2sk, e2 = −2si +2sj obtained p and α.
+2(s + 2−1/2p)k, and e3 = −2si +2(s + 2−1/2p)j + 2sk, where i, j,
and k are unit vectors along the x, y, and z directions and s is and simulation on self-assembled of nonspherical particles are
the smallest positive root of the equation (s + 2−1/2p)2p + 2s2p − still in the early stages of development. But basically, the final
1 = 0. We plotted 1/p against α (Figure 7b) and obtained a self-assembling pattern originates from repulsive and attractive
value of 2.2 for p from the known value α. The RS assembly interactions between adjacent particles. In our case, the
with lattice angle variable is proposed by nanocube sliding nanocubes can be considered as quasispherical shape due to
under a certain extent without rotation.23 The studies of theory geometrical effects (truncated shapes) and soft molecular shells,
1789 dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
Figure 8. (a) SEM and (b) photographic image of the Cu nanocube films prepared by immersion process. (c) AFM images and height profile of Cu
nanocube films fabricated by immersion process on a silicon substrate. (d) Reflectance spectra of the Cu nanocube films (red line) and Cu foil (black
line).
which lead to RS self-assembly. We think that the van der
Waals force is the only force between nanocubes due to their
characteristic of short-range order. For small 1/p < 0.12
(perfect cube or cubelike particles), the van der Waals force is
obvious (>4 kBT); therefore, SC arrangement is favored. When
1/p > 0.25, the influence of the van der Waals force declines
(∼kBT), leading to the RS assembly.23 According to a study by
Torquato and co-workers, the densest packing for all convex
and concave particles is the cube-like superball; the deformed
SC structure, corresponding to RS, becomes the densest when
p > 1.1509.54 A study by Gang et al., on the relationship
between volume fraction and p, shows that the packing of RS
becomes denser packing than that of SC, due to a rapid increase
in the packing density of RS 1/p > 0.12.23 In our system, 1/p is
about 0.4, so it is quite reasonable that these truncated and
ligand coated nanocubes tend to exhibit RS arrangement
1790
Article
Cu is a good alternative material as it possesses high
conductivity and is significantly cheaper than Au and Ag. In
addition, Cu nanocubes can be coated by a layer of organic
molecule which can be dispersed in organic solvent as
conductive inks. In order to obtain better dispersibility, we
perform ligand exchange of Cu nanocubes with dodecanethiol
because the amine is a relatively weak capping agent.55,56 The
shape of Cu nanocubes did not changed after ligand exchange
(Supporting Information Figure S5). Figure 8 shows a SEM
image of a high-quality film of Cu nanocubes formed over a
large area (1.5 cm × 2.5 cm). The Si substrate was immersed
aslant in a solution of dodecanethiol capped Cu nanocubes, and
the solvent (toluene) was slowly evaporated by controlling the
pressure under 150 Torr (Supporting Information Figure S6).46
The image reveals that the surface of the Cu nanocube film is
flat and dense (Figure 8a, Supporting Information Figures S7
and S8). When the Si substrate was slowly dried aslant in the
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
Figure 9. (a) SEM image and (b) cross-sectional morphology of the Cu nanocube film made by drop casting method after heat treatment at 350 °C.
(c) The resistivity of the Cu nanocube film with different thicknesses after heat treatment at 350 °C. (d) The resistivity of the Cu nanocube film
(black) and Cu nanoparticle film (red) with a thickness of 1.2 μm after heat treatment.
solution of dodecanethiol capped Cu nanocubes, the red−
brown solution deposited as a shiny Cu layer on the substrate
(Figure 8b). The reflectance spectra of the Cu nanocube film,
as shown in Figure 8d, exhibits a collective plasmon band at
603 nm. The change in color for the Cu nanocube film can be
attributed to (1) the difference in refractive indices of air and
toluene, because the plasmon resonance is highly sensitive to
the dielectric constant of the surrounding medium,1 and (2)
collective plasmon resonance in self-assembly of Cu nanocubes.
Some studies report that the collective plasmon resonance peak
position is very sensitive to the gap spacing between Cu
nanocubes. For Cu nanocubes solution, it exhibits a unique
plasmon response due to their cube geometries, but for self-
assembling Cu nanocubes, collective plasmon resonance
between Cu nanocubes is expected to control the optical
response; therefore, the color of Cu nanocubes film is
significantly different from Cu nanocubes dispersed in
solution.57,58 We also compare the reflectance spectra of the
Cu nanocube film (red line) and Cu foil (black line) in the
visible region (Figure 8d). The reflectance of the Cu nanocube
film is nearly identical to that of Cu foil in the low-reflectance
region (400−550 nm) except the reflectance dips in Cu
nanocube film. The reflectance of the two materials diverges
above 500 nm; Cu foil exhibits relatively high reflectance in the
550−800 nm wavelength range, while the Cu nanocube film
exhibits relatively low reflectance (reflectance dips) at about
631 nm. These reflectance dips probably were induced from the
interlayer collective plasmon resonance when the layer number
of Cu nanocubes on substrate is large enough. In addition, on
the basis of the results for other thin films of nanocrystals, the
optical properties of Cu nanocube film should be related to the
film thickness (layer number).58,59 Comparing with Cu film
made by Cu nanoparticles synthesized by the polyol process,60
the Cu nanocube film exhibits shiny color which is very similar
1791
Article
to the substrate after polishing, like polishing Cu foil or
polishing silicon substrate. The contact-mode AFM images of
the Cu nanocube films (Figure 8c) also reveals that the surface
roughness of the thin film (the root-mean-square deviation of
its topography) is only 4.6 nm, which means that the Cu
nanocubes can easily be made into smooth Cu nanocube film.
On the basis of Figure 8a, the two-dimensional (2D) packing
density is estimated as ψ2D = N × L2/A, where N and L are the
number of Cu nanocubes in Figure 8a and the average edge
length of Cu nanocubes, respectively. A is the surface area of
substrate in Figure 8a. The calculated two-dimensional packing
density is about 0.723. In addition, we also can calculate the
packing density ϕRS(p) when the truncated nanocube self-
assemble into a RS arrangement with deformation factor p =
2.2 by the following equation, ϕRS(p) = Vsb(p)/VRS(p), where
Vsb is the volume of superballs and VRS = e1 × e2·e3 is the unit
cell volumes of the RS phase. The calculated ϕRS(2.2) is
approximately 0.88.23,54 The packing density of the Cu
nanocube film is reasonably close to the theoretical value. A
metal film composed of nanoparticles with high packing density
has high film conductivity.61 Considering the high conductivity
of natural Cu and the theoretically high packing density of Cu
nanocubes, Cu nanocubes serve as a good material to be made
into conductive film; therefore, the resistivity of the Cu
nanocube film was also measured. The dense Cu nanocube film
prepared by drop casting method has large resistivity before the
annealing process. The surface ligand length is known to play a
critical role in the conductivity of the nanocrystal film. When
the length of the bifunctional cross-linkers < 0.8 nm, the
nanocrystal film becomes very conductive; when the linkers are
longer, the nanocrystal film becomes highly resistant.59 It is
well-known that heat treatment can improve the metal film
conductivity. Figure 9a shows SEM and optical images (inset)
of Cu thin films annealed for 30 min at 350 °C under a pressure
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials
of 0.1 Torr. The flow rates of the Ar and H2 carrier gases were
50 and 50 sccm, respectively. After annealing, the film becomes
dense but still exhibits some cracks. The cross-sectional SEM
image (Figure 9b) shows that the Cu film has a thickness of
around 1.2 μm. Heat treatment reduced the resistivity of the Cu
nanocube films, as shown in Figure 9c,d. First, we compare the
resistivity of Cu film made by Cu nanocrystals and Cu
nanocubes after annealing. According to Figure 9d, the
resistivity (1.33 × 10−7 Ω m) of the Cu nanocube film is
significantly lower than that of the Cu nanocrystals (6.215 ×
10−6 Ω m) film. We think that the packing density is one of the
main reasons leading to the differences in resistivity. In
addition, the contact area between nanocubes should be
significantly greater than that of the nanoparticles due to the
geometry itself. Next, we also compare the effect of resistivity of
Cu nanocube film on film thickness and annealing temperature.
Figure 9c shows the resistivity of Cu nanocube annealed at 350
°C with different thicknesses which are of the same order
(∼10−7 Ω m), and the film resistivity of all samples decreases
with increasing annealing temperature (Figure 9d, from 12.22 ×
10−7 Ω m at 150 °C to 1.33 × 10−7 Ω m at 350 °C). The
highest resistivity of Cu nanocube films is only 1 order of
magnitude higher than that of bulk Cu, whose resistivity is 1.68
× 10−8 Ω m. Heat treatment removes the organic ligands on
the Cu nanocube surface, increasing the extent of interparticle
contact area and enabling sufficient electron transport;
therefore, the Cu film becomes highly conductive.
■ CONCLUSIONS
In summary, monodisperse Cu nanocubes have been
synthesized with well-defined shape and good uniformity
from a hot organic solution. The TOP and ODA ligands play
an important role in selectively forming single seed crystals
stabilizing specific crystal planes, both of which are critical to
achieving monodisperse Cu nanocubes. The nanocubes tend to
self-assemble into square arrays and RS supercrystals; their
assembly is herein confirmed via TEM, SEM, and SAXS. The
XRD patterns suggest that the preferential orientation of the
Cu nanocubes is with their {100} planes parallel to the
substrate. These monodisperse Cu nanocubes can be dispersed
in an organic solvent as a conductive Cu ink, which provides a
convenient and inexpensive method to build dense Cu
nanocube films over a large area. After annealing at 350 °C,
films exhibited electrical resistivity of 1.33 × 10−7 Ω m, one
order higher than that of a bulk Cu film but several orders of
magnitude lower than conducting polymers.62 The high
conductivity of the Cu nanocube films is compatible with
their use in various electronic devices. The successful synthesis
of monodisperse Cu nanocubes is an important step toward
studying and exploiting the properties and self-assembly of Cu
nanomaterials.
■
ASSOCIATED CONTENT
*
S Supporting Information
SAED pattern, TEM and SEM images of Cu nanocubes.
Photography image of the immersion process to make large
area Cu nanocube superlattice. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*(H.-Y.T.) Tel.: + 886-3-572-3661. Fax: + 886-3-571-5408. E-
mail: hytuan@che.nthu.edu.tw.
1792
Article
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
The authors acknowledge the financial support by the National
Science Council of Taiwan (NSC 102-2221-E-007-023-MY3,
NSC 102-2221-E-007-090-MY2, NSC 101-2623-E-007-013-IT,
and NSC 102-2633-M-007-002), the Ministry of Economic
Affairs, Taiwan (101-EC-17-A-09-S1-198), National Tsing Hua
University (102N2051E1, 102N2061E1), and the assistance
from Center for Energy and Environmental Research, National
Tsing-Hua University.
■
REFERENCES
(1) Xia, Y. N.; Xiong, Y. J.; Lim, B.; Skrabalak, S. E. Angew. Chem., Int.
Ed. 2009, 48, 60−103.
(2) Tao, A. R.; Habas, S.; Yang, P. D. Small 2008, 4, 310−325.
(3) Chiu, C. Y.; Chung, P. J.; Lao, K. U.; Liao, C. W.; Huang, M. H. J.
Phys. Chem. C 2012, 116, 23757−23763.
(4) McLellan, J. M.; Li, Z. Y.; Siekkinen, A. R.; Xia, Y. N. Nano Lett.
2007, 7, 1013−1017.
(5) Sun, Y. G.; Xia, Y. N. Science 2002, 298, 2176−2179.
(6) Lu, X.; Rycenga, M.; Skrabalak, S. E.; Wiley, B.; Xia, Y. Annu. Rev.
Phys. Chem. 2009, 60, 167−192.
(7) Henning, A. M.; Watt, J.; Miedziak, P. J.; Cheong, S.;
Santonastaso, M.; Song, M.; Takeda, Y.; Kirkland, A. I.; Taylor, S.
H.; Tilley, R. D. Angew. Chem., Int. Ed. 2013, 52, 1477−1480.
(8) Ma, Y.; Li, W.; Cho, E. C.; Li, Z.; Yu, T.; Zeng, J.; Xie, Z.; Xia, Y.
ACS Nano 2010, 4, 6725−6734.
(9) Sun, J. Y.; Wang, Z. K.; Lim, H. S.; Ng, S. C.; Kuok, M. H.; Tran,
T. T.; Lu, X. ACS Nano 2010, 4, 7692−7698.
(10) Yu, T.; Moon, J.; Park, J.; Park, Y. I.; Na, H. B.; Kim, B. H.;
Song, I. C.; Moon, W. K.; Hyeon, T. Chem. Mater. 2009, 21, 2272−
2279.
(11) Watt, J.; Yu, C.; Chang, S. L. Y.; Cheong, S.; Tilley, R. D. J. Am.
Chem. Soc. 2013, 135, 606−609.
(12) Tao, A. R.; Ceperley, D. P.; Sinsermsuksakul, P.; Neureuther, A.
R.; Yang, P. Nano Lett. 2008, 8, 4033−4038.
(13) Zhang, Q.; Li, W.; Moran, C.; Zeng, J.; Chen, J.; Wen, L.-P.; Xia,
Y. J. Am. Chem. Soc. 2010, 132, 11372−11378.
(14) Hu, L.; Kim, H. S.; Lee, J.-Y.; Peumans, P.; Cui, Y. ACS Nano
2010, 4, 2955−2963.
(15) Pietrobon, B.; McEachran, M.; Kitaev, V. ACS Nano 2009, 3,
21−26.
(16) Lai, Y.; Pan, W.; Zhang, D.; Zhan, J. Nanoscale 2011, 3, 2134−
2137.
(17) Henzie, J.; Grünwald, M.; Widmer-Cooper, A.; Geissler, P. L.;
Yang, P. Nat. Mater. 2012, 11, 131−137.
(18) Wang, C.; Daimon, H.; Lee, Y.; Kim, J.; Sun, S. J. Am. Chem. Soc.
2007, 129, 6974−6975.
(19) Demortiere, A.; Launois, P.; Goubet, N.; Albouy, P. A.; Petit, C.
J. Phys. Chem. B 2008, 112, 14583−14592.
(20) An, K.; Lee, N.; Park, J.; Kim, S. C.; Hwang, Y.; Park, J.-G.; Kim,
J.-Y.; Park, J.-H.; Han, M. J.; Yu, J.; Hyeon, T. J. Am. Chem. Soc. 2006,
128, 9753−9760.
(21) Quan, Z. W.; Fang, J. Y. Nano Today 2010, 5, 390−411.
(22) Goodfellow, B. W.; Patel, R. N.; Panthani, M. G.; Smilgies, D.-
M.; Korgel, B. A. J. Phys. Chem. C 2011, 115, 6397−6404.
(23) Zhang, Y. G.; Lu, F.; van der Lelie, D.; Gang, O. Phys. Rev. Lett.
2011, 107, 135701.
(24) LaGrow, A. P.; Ingham, B.; Cheong, S.; Williams, G. V. M;
Dotzler, C.; Toney, M. F.; Jefferson, D. A.; Corbos, E. C.; Bishop, P.
T.; Cookson, J.; Tilley, R. D. J. Am. Chem. Soc. 2012, 134, 855−858.
(25) Bian, K.; Choi, J. J.; Kaushik, A.; Clancy, P.; Smilgies, D.-M.;
Hanrath, T. ACS Nano 2011, 5, 2815−2823.
(26) Singh, A.; Dickinson, C.; Ryan, K. M. ACS Nano 2012, 6, 3339−
3345.
dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793
Chemistry of Materials Article

(27) Choi, J. J.; Bealing, C. R.; Bian, K.; Hughes, K. J.; Zhang, W.; (61) Woo, K.; Kim, D.; Kim, J. S.; Lim, S.; Moon, J. Langmuir 2009,
Smilgies, D.-M.; Hennig, R. G.; Engstrom, J. R.; Hanrath, T. J. Am. 25, 429−433.
Chem. Soc. 2011, 133, 3131−3138. (62) Long, Y.; Chen, Z. J.; Wang, N. L.; Ma, Y. J.; Zhang, Z.; Zhang,
(28) Hanrath, T.; Choi, J. J.; Smilgies, D.-M. ACS Nano 2009, 3, L. J.; Wan, M. X. Appl. Phys. Lett. 2003, 83, 1863−1865.
2975−2988.
(29) Bodnarchuk, M. I.; Kovalenko, M. V.; Heiss, W.; Talapin, D. V.
J. Am. Chem. Soc. 2010, 132, 11967−11977.
(30) Smith, D. K.; Goodfellow, B.; Smilgies, D.-M.; Korgel, B. A. J.
Am. Chem. Soc. 2009, 131, 3281−3290.
(31) Bian, K.; Wang, Z.; Hanrath, T. J. Am. Chem. Soc. 2012, 134,
10787−10790.
(32) Shevchenko, E. V.; Talapin, D. V.; Kotov, N. A.; O’Brien, S.;
Murray, C. B. Nature 2006, 439, 55−59.
(33) Chen, M.; Kim, J.; Liu, J. P.; Fan, H. Y.; Sun, S. H. J. Am. Chem.
Soc. 2006, 128, 7132−7133.
(34) Ni, R.; Gantapara, A. P.; Graaf, J.; Roij, R.; Dijkstra, M. Soft
Matter 2012, 8, 12135−12135.
(35) Zhang, J.; Kumbhar, A.; He, J. B.; Das, N. C.; Yang, K. K.; Wang,
J. Q.; Wang, H.; Stokes, K. L.; Fang, J. Y. J. Am. Chem. Soc. 2008, 130,
15203−15209.
(36) Singh, A.; English, N. J.; Ryan, K. M. J. Phys. Chem. B 2013, 117,
1608−1615.
(37) Singh, A.; Coughlan, C.; Laffir, F. R.; Ryan, K. M. ACS Nano
2012, 6, 6977−6983.
(38) Singh, A.; Geaney, H.; Laffir, F. R.; Ryan, K. M. J. Am. Chem.
Soc. 2012, 134, 2910−2913.
(39) Jain, P. K.; Eustis, S.; El-Sayed, M. A. J. Phys. Chem. B 2006, 110,
18243−18253.
(40) Dixit, P.; Miao, J. M.; Preisser, R. Electrochem. Solid-State Lett.
2006, 9, L16−L16.
(41) Kam, Z. Nature 1983, 306, 625−625.
(42) Zeman, E. J.; Schatz, G. C. J. Phys. Chem. 1987, 91, 634−643.
(43) Jin, M. S.; He, G. N.; Zhang, H.; Zeng, J.; Xie, Z. X.; Xia, Y. N.
Angew. Chem., Int. Ed. 2011, 50, 10560−10564.
(44) Guo, H. Z.; Chen, Y. Z.; Ping, H. M.; Jin, J. R.; Peng, D. L.
Nanoscale 2013, 5, 2394−2402.
(45) Wang, Y.; Chen, P.; Liu, M. Nanotechnology 2006, 17, 6000−
6006.
(46) Shevchenko, E. V.; Talapin, D. V.; Murray, C. B.; O’Brien, S. J.
Am. Chem. Soc. 2006, 128, 3620−3637.
(47) Murray, C. B.; Kagan, C. R.; Bawendi, M. G. Annu. Rev. Mater.
Sci. 2000, 30, 545−610.
(48) Zhang, H.-X.; Wang, H.; Re, Y.-S.; Cai, W.-B. Chem. Commun.
2012, 48, 8362−8364.
(49) Wang, Z. L. J. Phys. Chem. B 2000, 104, 1153−1175.
(50) Chen, S.; Zhang, X.; Zhang, Q.; Tan, W. Nanoscale Res. Lett.
2009, 4, 1159−1165.
(51) Cooper, J. K.; Franco, A. M.; Gul, S.; Corrado, C.; Zhang, J. Z.
Langmuir 2011, 27, 8486−8493.
(52) Eggiman, B. W.; Tate, M. P.; Hillhouse, H. W. Chem. Mater.
2006, 18, 723−730.
(53) Yamamuro, S.; Sumiyama, K. Chem. Phys. Lett. 2006, 418, 166−
169.
(54) Jiao, Y.; Stillinger, F. H.; Torquato, S. Phys. Rev. E 2009, 79,
041309.
(55) Ye, E.; Zhang, S.-Y.; Liu, S.; Han, M.-Y. Chem.Eur. J. 2011, 17,
3074−3077.
(56) Kanninena, P.; Johans, C.; Merta, J.; Kontturi, K. J. Colloid
Interface Sci. 2008, 318, 88−95.
(57) Su, K. H.; Wei, Q. H.; Zhang, X.; Mock, J. J.; Smith, D. R.;
Schultz, S. Nano Lett. 2003, 3, 1087−1090.
(58) Lin, M.-H.; Chen, H.-Y.; Gwo, S. J. Am. Chem. Soc. 2010, 132,
11259−11263.
(59) Musick, M. D.; Keating, C. D.; Lyon, L. A.; Botsko, S. L.; Pena,
D. J.; Holliway, W. D.; McEvoy, T. M.; Richardson, J. N.; Natan, M. J.
Chem. Mater. 2000, 12, 2869−2881.
(60) Park, B. K.; Jeong, S.; Kim, D.; Moon, J.; Lim, S.; Kim, J. S. J.
Colloid Interface Sci. 2007, 311, 417−424.

1793 dx.doi.org/10.1021/cm403098d | Chem. Mater. 2014, 26, 1785−1793