Aqueous phase synthesis of Au-Ag core-shell nanocrystals with tunable shapes and their optical and catalytic properties

Article abstract of DOI:10.1021/ja410663g

In this study, rhombic dodecahedral gold nanocrystals were used as cores for the generation of Au-Ag core-shell nanocrystals with cubic, truncated cubic, cuboctahedral, truncated octahedral, and octahedral structures. Gold nanocrystals were added to an aq

Full text of DOI:10.1021/ja410663g

Article
pubs.acs.org/JACS
Aqueous Phase Synthesis of Au−Ag Core−Shell Nanocrystals with
Tunable Shapes and Their Optical and Catalytic Properties
Yu-Chi Tsao, Sourav Rej, Chun-Ya Chiu, and Michael H. Huang*
Department of Chemistry, National Tsing Hua University, Hsinchu 30013, Taiwan
S
* Supporting Information
ABSTRACT: In this study, rhombic dodecahedral gold nanocrystals were used as
cores for the generation of Au−Ag core−shell nanocrystals with cubic, truncated cubic,
cuboctahedral, truncated octahedral, and octahedral structures. Gold nanocrystals were
added to an aqueous mixture of cetyltrimethylammonium chloride (CTAC) surfactant,
AgNO3, ascorbic acid, and NaOH to form the core−shell nanocrystals. The
nanocrystals are highly uniform in size and shape, and can readily self-assemble into
ordered packing structures on substrates. Results from observation of solution color
changes and variation in the reaction temperature suggest octahedra are produced at a
higher growth rate, while slower growth favors cube formation. The major localized
surface plasmon resonance (LSPR) band positions for these nanocrystals are red-
shifted compared to those for pristine silver particles with similar dimensions due to the LSPR effect from the gold cores. By
increasing the concentrations of reagents, Au−Ag core−shell cubes and octahedra with tunable sizes were obtained. Au−Ag
cubes with body diagonals of 130, 144, and 161 nm and octahedra with body diagonals of 113, 126, and 143 nm have been
prepared, allowing the investigation of size effect on their optical properties. Au−Ag octahedra with thinner Ag shells (12−16.5
nm) exhibit a blue-shifted major LSPR band relative to the LSPR band at 538 nm for the gold cores. For Au−Ag octahedra and
cubes with thicker shells (22.5−37 nm), the major LSPR band is progressively red-shifted from that of the gold cores with
increasing shell thickness and particle size. The Au−Ag octahedra show higher catalytic activity than cubes toward reduction of 2-
amino-5-nitrophenol by NaBH4 at 30 °C, but both particle shapes display significantly enhanced catalytic efficiency at 40 °C.
from cubic to octahedral morphologies. An alternative strategy
to achieving this goal is by making Au−Ag core−shell
heterostructures. While various Au−Ag core−shell nanostruc-
tures can be synthesized using a polyol process in the presence
of PVP,11−13 formation of Au−Ag core−shell nanocrystals in
aqueous solution has also been achieved.14−22 However, mainly
nanocubes were produced with the use of polyhedral gold cores
in aqueous solution. Other shapes such as Au−Ag core−shell
octahedra and tetrahedra can be obtained by a light-induced
synthetic approach.21 Interestingly, use of decahedral gold
nanocrystals as cores can yield Au−Ag nanorods, while gold
nanoplate cores give elongated Au−Ag structures.23−26 Clearly,
the growth of Au−Ag core−shell nanocrystals with systematic
shape evolution from cubic to octahedral structures in aqueous
solution using polyhedral gold cores has not been achieved yet.
Furthermore, it is highly desirable that the synthesized Au−Ag
nanocrystals show excellent size and shape control for their
large-scale assembly and comparison of their optical and
catalytic properties.
INTRODUCTION
■
Gold and silver nanocrystals are an important class of
nanomaterials due to their strong size- and shape-dependent
LSPR properties and ability to bind a wide variety of organic
and biological molecules for potential applications in chemical
sensing, photothermal therapy, and as sensitive substrates for
surface-enhanced Raman scattering (SERS) of adsorbed
molecules.1−5 Catalytic activity of silver nanostructures
exposing different crystal faces for the oxidation of styrene
has also been examined.6 For the synthesis of silver
nanocrystals with tunable shapes from cubic to octahedral
structures, a polyol process has generally been employed. Silver
precursor added to a solution of ethylene glycol or 1,5-
pentanediol in the presence of capping polymer such as
poly(vinyl pyrrolidone), or PVP, is normally heated to 150−
180 °C.7−9 The initially produced silver nanocubes are
transformed into truncated cubes, cuboctahedra, and octahedra
by controlling the reaction time.7,8 The polyol synthetic
approach suffers from elaborate PVP removal steps, high
reaction temperature, and large sizes of the synthesized
octahedra (250−300 nm). Development of aqueous phase
synthesis of silver nanocrystals using simple surfactant should
be quite important. However, silver polyhedra formed in
aqueous solutions are mostly cubes as far as a series of particle
shape transition from cubic to octahedral morphologies is
concerned.10 It remains challenging to synthesize silver
nanocrystals in aqueous solution with systematic shape tuning
In this study, we have successfully synthesized Au−Ag core−
shell nanocubes, truncated cubes, cuboctahedra, truncated
octahedra, and octahedra in aqueous solution using rhombic
dodecahedral gold nanocrystals as cores. The particles are
highly uniform in size and shape that they spontaneously self-
Received: October 18, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. (a) SEM image of the synthesized rhombic dodecahedral gold nanocrystals. (b) Size distribution histogram and a drawing of the particle.
The arrow indicates the particle size. (c) UV−vis absorption spectrum of the rhombic dodecahedral gold nanocrystals.
were used for the growth of truncated octahedra and octahedra,
respectively. After further reaction for 50 min, the solution was
centrifuged at 3500 rpm for 10 min. An illustration of the synthetic
procedure used is given in Scheme S1.
assemble on substrates. The relative growth rates in the
formation of nanocubes and octahedra have been evaluated by
visual observation of solution color changes and adjustment of
reaction temperature. Sizes of the nanocubes and octahedra can
also be tuned over a range. Optical property characterization of
the particles has been performed. Interestingly, both blue-shifts
and red-shifts of the core−shell plasmonic band relative to
those of the Au cores have been recorded. Facet-dependent
catalytic activity of the Au−Ag nanocubes and octahedra was
also examined through borohydride reduction of 2-amino-5-
nitrophenol.
Synthesis of Cubic and Octahedral Au−Ag Core−Shell
Nanocrystals with Tunable Size. The above procedures described
yield Au−Ag core−shell nanocubes and nanooctahedra with edge
lengths of 75 and 81 nm, respectively. To make larger nanocubes and
octahedra, the same volume of Au rhombic dodecahedra solution was
used (40 μL), but the amounts of other reagents used were increased.
For the synthesis of Au−Ag nanocubes with edge lengths of 83 and 93
nm, the amount of CTAC and the concentrations of AgNO3 solution,
ascorbic acid solution and NaOH solution added were 1.5 and 2 times
those used to make 75 nm nanocubes, respectively. Similarly, to obtain
octahedra with edge lengths of 90 and 102 nm, the amount of CTAC
and the concentrations of AgNO3 solution, ascorbic acid solution and
NaOH solution added were 1.5 and 2 times those used to make 81 nm
octahedra, respectively. A scheme summarizing the experimental
procedure is available in Scheme S2.
Catalytic Reduction of 2-Amino-5-nitrophenol. For a typical
catalysis reaction, 1 mL of 1.0 × 10−3 M 2-amino-5-nitrophenol
solution was added to 8 mL of deionized water and stirred thoroughly.
After that, different volumes of the Au−Ag nanocrystal solutions were
added to the reaction mixture (i.e., 300 μL for nanocubes and 148 μL
for octahedra), keeping the total surface area of the core−shell
particles approximately the same at around 10.78 cm2 (see Table S1).
The reaction mixture was maintained at 30 °C using a water bath.
Finally, 2.4 mL of 0.1 M ice-cold NaBH4 solution was added and the
solution was mixed well to start the reaction. Next, 800 μL of the
reaction mixture was transferred to a narrow-slit cuvette for spectral
measurements without dilution. The solution was discarded after
taking a spectrum and fresh solution from the same reaction mixture
was used for the next measurement. UV−vis absorption spectra were
recorded at different time periods depending on the reaction rate.
After the completion of the reduction reaction, the yellowish 2-amino-
EXPERIMENTAL SECTION
■
Synthesis of Cubic, Truncated Cubic, and Cuboctahedral
Au−Ag Core−Shell Nanocrystals. The rhombic dodecahedral gold
nanocrystals with an average size of 53 nm were prepared following
our reported method.27 In a typical synthesis of Au−Ag core−shell
nanocubes, 0.064 g of CTAC surfactant, 9.46 mL of deionized water,
and 40 μL of the Au rhombic dodecahedra solution were introduced.
The vial was kept in a water bath set at 30 °C. Then 50 μL of 0.01 M
AgNO3 solution was added into the sample vial. For the growth of
truncated cubic and cuboctahedral Au−Ag core−shell nanocrystals, 40
and 30 μL, respectively, of 0.01 M AgNO3 solution were introduced
into the reaction mixture. Finally, 150 μL of 0.0175 M ascorbic acid
solution and 300 μL of 0.0175 M NaOH solution were added.
Immediately after the addition of NaOH, the solution turned from
pink to yellow within 1 min. After 50 min, the solution was centrifuged
at 3500 rpm for 10 min.
Synthesis of Truncated Octahedral and Octahedral Au−Ag
Core−Shell Nanocrystals. Truncated octahedral and octahedral
Au−Ag core−shell nanocrystals were synthesized using the same
amounts of CTAC and Au rhombic dodecahedra solution. However,
150 μL of 0.05 M ascorbic acid solution and 300 μL of 0.05 M NaOH
solution were added. Here 35 and 50 μL of 0.01 M AgNO3 solution
B
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. SEM images of Au−Ag core−shell (a1−a3) nanocubes, (b1−b3) truncated cubes, (c1−c3) cuboctahedra, (d1−d3) truncated octahedra,
and (e1−3) octahedra viewed at different magnifications. All scale bars are equal to 100 nm.
5-nitrophenol solution turned colorless due to the formation of 2,5-
RESULTS AND DISCUSSION
■
diaminophenol product.
Our goal in the preparation of Au−Ag core−shell nanocrystals
Instrumentation. Scanning electron microscopy (SEM) images of
is to achieve their systematic shape evolution from cubic to
the samples were obtained using a JEOL JSM-7000F electron
octahedral structures for possible assembly and facet-dependent
microscope. Transmission electron microscope (TEM) character-
property investigation.28,29 The relative growth rates needed to
ization was performed on a JEOL JEM-2100 microscope with an
form different particle morphologies can also be studied. Here
rhombic dodecahedral gold nanocrystals were employed for the
preparation of Au−Ag core−shell nanocrystals. The particles
operating voltage of 200 kV. Elemental mapping images were acquired
by energy-dispersive X-ray spectroscopy (EDS) using the same JEOL
JEM-2100 electron microscope equipped with a STEM unit and an
were synthesized following a seed-mediated growth method in
the presence of a tiny amount of NaBr.27 Figure 1 illustrates the
Inca Energy 250 detector from Oxford Instruments. Powder X-ray
diffraction (XRD) patterns were recorded on a Shimadzu XRD-6000
diffractometer with Cu Kα radiation. UV−vis absorption spectra were
taken using a JASCO V-670 spectrophotometer.
SEM image of the synthesized rhombic dodecahedral Au
nanocrystals with an average edge length of 33 or 77 nm by
measuring the opposite corner-to-corner distance. Because of
C
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 3. TEM images, corresponding SAED patterns, and representative drawings of single Au−Ag core−shell (a1−a3) cube, (b1−b3)
cuboctahedron viewed along the [100] direction, (c1−c3) truncated octahedron viewed along the [110] direction, and (d1−d3) octahedron viewed
along the [111] direction. All scale bars are equal to 20 nm. The triangular diffraction spots seen in panel d2 may be due to double diffraction
because of the polyhedral core and shell nature. Split diffraction spot effect is also visible in panel b2.
their uniform size and shape, these particles readily self-
assemble into a multilayered packing structure with their ⟨111⟩
axis perpendicular to the substrate plane. They have been
demonstrated to pack into supercrystals with micrometer
dimensions.30 The rhombic dodecahedra give a sharp LSPR
band at 538 nm. The same rhombic dodecahedral gold
nanoparticles can also be obtained by introducing KI instead of
NaBr.31
To make the Au−Ag core−shell nanocrystals, an aqueous
solution of CTAC surfactant, rhombic dodecahedral gold
nanocrystals, AgNO3, ascorbic acid, and NaOH was prepared
and reacted at 30 °C for 50 min to obtain the products. When
the volume of 0.01 M AgNO3 solution introduced was reduced
from 50 to 40 and 30 μL, Au−Ag nanocubes were transformed
into truncated cubes and cuboctahedra. Raising the concen-
trations of ascorbic acid and NaOH from those employed for
the synthesis of nanocubes, Au−Ag octahedra and truncated
octahedra were produced with the use of 50 and 35 μL of 0.01
M AgNO3 solution, respectively. Lower ascorbic acid and
NaOH concentrations favor the formation of Au−Ag cubes,
while higher ascorbic acid and NaOH concentrations facilitate
the growth of octahedra. However, it is difficult to simply tune
the concentrations of ascorbic acid and NaOH to achieve a
single particle shape; a mixture of cuboctahedra and truncated
octahedra can form. We found an easier way to obtain the
intermediate particle morphologies is to fix the reaction
conditions for making Au−Ag cubes and octahedra, and adjust
the volume of AgNO3 slightly to yield truncated cubes,
cuboctahedra, and truncated octahedra. Figure 2 presents SEM
images of the synthesized Au−Ag core−shell cubes, truncated
cubes, cuboctahedra, truncated octahedra, and octahedra taken
at different magnifications. Highly uniform Au−Ag nanocrystals
have been generated in entire five samples, such that they all
spontaneously self-assemble into ordered packing arrange-
ments. The cubes, truncated cubes, and cuboctahedra form
roughly face-to-face alignment. Alternatively, their alignment
can shift within the same layer and between upper and lower
layer as have been observed before for Au and PbS
nanocubes.30,32 The octahedra and truncated octahedra can
form monolayer with their {111} faces contacting the substrate
plane and multilayer packing structures oriented to show their
{110} edges. The average particle sizes in terms of body
D
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
diagonals for the Au−Ag core−shell cubes, truncated cubes,
cuboctahedra, truncated octahedra, and octahedra are 130, 103,
100, 89, and 113 nm (or edge lengths of 75, 73, 71, 80, and 81
nm) with standard deviations of 3.5% or less, respectively (see
Figures S1 and S2 and Table S2, Supporting Information).
Silver polyhedra with these dimensions have not been
demonstrated to show such large-scale assembly, proving the
success of this synthetic strategy to yield silver nanocrystals
with excellent shape control and tunability.
the shoulder band at 480 nm for the nanocubes presumably is
due to their larger sizes than particles of other shapes. Ag
nanocubes with the same edge length of 75 nm have been
reported to show LSPR bands at 345, 395, 450, and 515 nm,
while Ag octahedra with an edge length of 75 nm display a
major LSPR band at 500 nm.33 Clearly the major LSPR band of
the Au−Ag core−shell nanocrystals is red-shifted compared to
pristine Ag particles of the same size and shape. The
contribution of the gold cores to the major LSPR band
position of silver nanocrystals is obvious. The spectral red-shift
toward the LSPR band position of the Au core is reasonable
because its LSPR property remains despite the presence of a
metal shell, but the band position is tuned by the shell
composition and thickness.15,16
Further structural analysis of these nanocrystals was
performed. XRD patterns of the Au−Ag core−shell cubes
and truncated cubes showed only the (200) reflection peak (see
Figure S3). Au and Ag have nearly identical lattice constants, so
the XRD patterns only give one set of reflection peaks. The
cuboctahedra exhibited a strong (200) peak and a weak (111)
peak. The relative intensities of the (200) and (111) peaks are
reversed for the truncated octahedra. For the octahedra, only
the (111) peak was observed. These relative peak intensity
changes were recorded as a result of preferential deposition of
the nanocrystals on their {100} or {111} faces. Figure 3
presents TEM images, the corresponding selected-area electron
diffraction (SAED) patterns, and the drawings of single Au−Ag
cube, cuboctahedron, truncated octahedron, and octahedron.
The gold cores can be seen. These SAED patterns match to
that of face-centered-cubic silver. The single-crystalline nature
of the silver shell is evident. High-angle annular dark-field
scanning TEM (HAADF-STEM) images of the Au−Ag core−
shell cubes give better contrast of the core and shell
components (see Figure S4). In addition, EDS line scans
confirm the composition of the core and shell as Au and Ag.
UV−vis absorption spectra of these Au−Ag core−shell
nanocrystals were taken (Figure 4). Since body diagonals are
We have shown previously through nanocrystal synthesis
with systematic shape evolution that the formation of different
particle shapes is related to their different growth rates.28 The
extent of particle growth can be followed by simply monitoring
the solution color changes. Figure S5 presents photographs of
the nanocrystal solutions at different time points in the
synthesis of Au−Ag core−shell cubes and octahedra. As silver
shells are formed, the purplish pink gold nanocrystal solution
gradually turns more brightly yellowish. After only 15 s, the
solution color remained the same in the formation of octahedra,
while it took 39 s for the solution color to become unchanged
in the synthesis of cubes. This visual observation suggests that
octahedra are formed at a faster growth rate than that for cubes.
Mirkin et al. have also found that the formation of different
silver particle shapes is related to their different reaction
rates.34,35 The Au−Ag octahedra were produced by introducing
higher concentrations of ascorbic acid and NaOH solution than
those used for the growth of cubes, implying that a higher
reduction or growth rate favors the formation of octahedra,
while a slower reduction rate facilitates the generation of cubes.
To test this relationship, a reaction condition for the
preparation of Au−Ag octahedra was used, but the reaction
temperature was decreased from 30 to 20 and 7 °C to reduce
the particle growth rate. Remarkably, Au−Ag core−shell
cuboctahedra and cubes were respectively synthesized at 20
and 7 °C (see Figure 5). The nanocrystals are also highly
uniform in size and shape that they readily self-assemble. The
results support the general observation that polyhedral particles
with different morphologies are formed at different growth
rates, and that a slower rate favors the production of Au−Ag
cubes. Thus, nanocrystals with different shapes can be obtained
by either varying the ascorbic acid and NaOH concentrations
or changing the reaction temperature to tune the particle
growth rate. The role of NaOH in enhancing the reducing
power of ascorbic acid is explained. Because of the relatively
high concentration of CTAC added, AgNO3 should first react
with chloride ions to form AgCl. Although ascorbic acid can
reduce AgCl to Ag, the reaction rate is too slow. Addition of
NaOH can accelerate the reduction reaction and enables better
control of the resulting particle shapes. The following equation
is the redox reaction taking place between ascorbic acid and Ag+
ions.
Figure 4. UV−vis-NIR absorption spectra of various synthesized Au−
Ag core−shell nanocrystals. Body diagonals are used for the particle
sizes.
more appropriate and useful than edge lengths to reflect the
actual nanocrystal sizes for particles with different morpholo-
gies, these numbers are given in Figure 4. The characteristic
multiple absorption band feature of silver due to the dipole and
quadruple plasmon resonance has been recorded. The
nanocubes show four LSPR bands at 346, 403, 481, and 555
nm. Particles with other shapes display only three bands with
the major band centered at 521, 524, 526, and 529 nm for
truncated octahedra, cuboctahedra, truncated cubes, and
octahedra, respectively. The complete peak positions for
these particles are available in Table S3. The emergence of
ascorbic acid + 2Ag+ + H2O
⇌ dehydroascorbic acid + 2H+ + 2Ag
Adding NaOH to the reaction mixture for the growth of silver
shells decreases [H+] and shift the equilibrium to the right,
which increases the reducing power of ascorbic acid. Hence,
E
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. SEM images showing the product shape changes by synthesizing Au−Ag core−shell nanocrystals at different temperatures. The scale bars
are equal to 100 nm.
Figure 6. SEM images of the Au−Ag core−shell nanocubes synthesized with body diagonals varying from (a) 130 to (b) 144 and (c) 161 nm and
octahedra with body diagonals of (d) 113, (e) 126, and (f) 143 nm. Scale bars are all equal to 100 nm.
Figure 7. UV−vis absorption spectra of Au−Ag core−shell (a) nanocubes and (b) octahedra with different sizes. Body diagonals are used for particle
size.
introduction of different concentrations of NaOH can adjust
the redox reaction rate to control the shapes of Au−Ag core−
shell nanocrystals.
core−shell octahedra with average body diagonals of 126 and
143 nm. Figure 6 is the SEM images of synthesized Au−Ag
core−shell cubes and octahedra in this study with three
different sizes. Again the larger cubes and octahedra maintain
excellent size and shape uniformity that they spontaneously
self-assembly into ordered arrays, particularly for octahedra.
This high degree of particle size and shape control is extremely
important for the accurate examination of size effect on their
spectral shifts.
Figure 7 gives UV−vis spectra of the Au−Ag core−shell
cubes and octahedra shown in Figure 6. Again body diagonals
are provided as this dimension reflects the true particle size
better than edge length. The exact peak positions and particle
It is also interesting to make Au−Ag core−shell nanocrystals
over a size range by increasing the shell thickness, and see how
their LSPR peaks are tuned. When the amount of CTAC and
concentrations of AgNO3, ascorbic acid, and NaOH solutions
were increased 1.5 and 2 times higher than those used to make
Au−Ag core−shell cubes with 130 nm body diagonal, while
keeping the volume of rhombic dodecahedral gold nanocrystal
solution added constant, Au−Ag core−shell cubes with
respective average body diagonals of 144 and 161 nm were
prepared. Use of the same synthetic conditions yielded Au−Ag
F
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Major LSPR Peak Positions and Shifts of Au−Ag Core−Shell Octahedra and Cubes with Different Sizes Relative to the
LSPR Peak of the Gold Cores at 538 nm
Oh volume (nm3)
2.5 × 105
113
3.4 × 105
126
5.0 × 105
143
Cube volume (nm3)
4.2 × 105
130
5.7 × 105
144
8.0 × 105
161
body diagonal (nm)
body diagonal (nm)
shell thickness (nm)
12
16.5
22.5
shell thickness (nm)
24.5
30
37
major LSPR peak (nm)
LSPR peak shift to Au core peak (nm)
529
534
543
major LSPR peak (nm)
LSPR peak shift to Au core peak (nm)
555
560
564
−9
−4
+5
+17
+22
+26
Figure 8. (a and b) Time-dependent UV−vis absorption spectra for the borohydride reduction of 2-amino-5-nitrophenol at 30 °C using Au−Ag
core−shell (a) cubes and (b) octahedra. (c) ln[C0/Ct] versus time plot using Au−Ag core−shell nanocubes and octahedra as catalysts for the
reduction of 2-amino-5-nitrophenol carried out at 30 °C.
sizes in terms of edge length, body diagonal, and single particle
volume are available in Table S4. At larger sizes, the cubes show
four clearly identifiable peaks, while the shoulder band for the
octahedra becomes more pronounced, further indicating the
appearance of the shoulder band (or peak 3 for the cubes) in
particles with larger dimensions. The spectral shifts of the major
LSPR band relative to that of the gold nanocrystal cores are
most interesting and insightful to analyze for the size effects.
For the Au−Ag core−shell cubes, the major LSPR band red-
shifts from 555 to 560 and 564 nm with increasing particle size.
Similarly, the major LSPR band red-shifts from 529 to 534 and
543 nm as octahedra become larger. Progressive spectral red-
shift with increasing particle size is normal and expected.9
However, a more careful analysis reveals how the shell thickness
can tune the major LSPR band position. Table 1 lists the
directions and magnitudes of spectral shifts of the major LSPR
band for these Au−Ag core−shell octahedra and cubes relative
to the LSPR peak position of the gold cores with consideration
of the silver shell thickness (see Figure S6 for the determination
of Ag shell thickness). For Au−Ag cubes with Ag shell
thicknesses increasing from 24.5 to 30 and 37 nm, the particles
show significant red-shifts of 17−26 nm from the absorption
band of the Au cores at 538 nm. However, for Au−Ag
octahedra with thinner Ag shell thicknesses of 12−22.5 nm,
both spectral blue-shift and red-shift were recorded. The 113-
nm Au−Ag octahedra with a shell thickness of 12 nm exhibit a
blue-shift of 9 nm. Hence, both spectral blue-shift and red-shift
can be observed in Au−Ag nanocrystals, and likely in other
bimetallic core−shell particles, depending on the shell thick-
ness. Normally the presence of Ag shells shifts the LSPR band
of polyhedral Au−Ag core−shell nanocrystals to the blue to
reflect the SPR character of silver. However, a large shell
thickness results in a significant increase in overall particle size,
so absorption bands of the core−shell particles become more
red-shifted with growing particle dimensions. It is also
interesting to note that the extents of spectral red-shift differ
G
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
substantially for 143-nm octahedra and 130-nm cubes (5 vs 17
nm), even though the shell thickness difference seems small.
This is likely attributed to the fact that the Au core is actually
closer to the nanocrystal surface at some points of the core−
shell cube than the Au core is in an octahedron (see Figure 3).
Hence, the SPR character of the Au core is more pronounced
and this leads to a greater peak red shift. Such spectral analysis
enhances our understanding of SPR behavior in bimetallic
core−shell nanocrystals, and the data are more reliable when
the particles have excellent size and shape control.
CONCLUSION
■
Rhombic dodecahedral gold nanocrystals were employed as
cores for the growth of Au−Ag core−shell nanocrystals with
systematic shape evolution from cubic to truncated cubic,
cuboctahedral, truncated octahedral, and octahedral structures.
The nanocrystals are highly uniform in size and shape that they
readily self-assemble on substrates. Silver polyhedra assembly
using small particles with edge lengths of 80 nm or less is rare.
Thus, synthesis of polyhedral silver nanocrystals, which are
quite challenging to prepare in aqueous solution using a simple
capping agent such as CTAC surfactant, has been achieved
through the core−shell growth strategy. The major LSPR band
positions for these nanocrystals are red-shifted compared to
pristine silver particles with similar dimensions due to the SPR
effect of the gold cores. When the reagent amounts are
increased, Au−Ag core−shell cubes and octahedra with tunable
sizes can be obtained. With relatively thin silver shells, the
major LSPR band of Au−Ag octahedra is blue-shifted relative
to that of the gold cores, but becomes progressively red-shifted
with increasing shell thickness. Au−Ag octahedra are more
catalytically active than cubes toward NaBH4 reduction of 2-
amino-5-nitrophenol at 30 °C, but both particle shapes increase
their catalytic efficiency significantly at 40 °C. It is envisioned
that these particles can be utilized to catalyze many other
reactions, assembled to form superlattice structures, and serve
as excellent SERS substrates.
The Au−Ag core−shell cubes and octahedra exposing
exclusively {100} and {111} facets can be used to examine
the facet-dependent catalytic activity of silver nanocrystals.
Although Ag and Au−Ag core−shell nanoparticles have been
tested for catalytic reduction of 4-nitrophenol by NaBH4 in
water, facet-dependent catalytic activity investigation was not
possible.36,37 We have previously compared the catalytic activity
of gold nanocubes, octahedra, and rhombic dodecahedra
toward 4-nitroaniline reduction.38 Gold rhombic dodecahedra
showed the best catalytic activity at all the temperatures
examined. Gold nanocubes displayed better catalytic efficiency
than octahedra in the temperature range of 25−32 °C, but
interestingly octahedra can be somewhat better catalysts than
cubes at higher reaction temperatures (36−40 °C). Here,
different volumes of the nanocrystal solutions having the same
total surface area were used for the comparative catalytic
activity experiments (see Supporting Information for the
calculations). Figure 8 offers time-dependent UV−vis absorp-
tion spectra for the NaBH4 reduction of 2-amino-5-nitrophenol
at 30 °C using Au−Ag core−shell cubes and octahedra to
mediate electron transfer. The band at 460 nm from 2-amino-5-
nitrophenol descreases gradually as the reaction proceeds, while
two new bands at 285 and 330 nm from the product 2,5-
diaminophenol increase continuously. However, the reaction is
still far from completion after 135 min. A plot of ln[Co/Ct]
versus time for these spectra gives straight lines, indicating the
reaction follows first-order kinetics. The Au−Ag octahedra and
cubes have respective rate constants of 9.69 × 10−3 min−1 and
5.49 × 10−3 min−1, so octahedra are more catalytically active at
30 °C. For the reaction to take place, both BH4− and 2-amino-
5-nitrophenolate ion should be adsorbed on the particle surface
ASSOCIATED CONTENT
■
S
* Supporting Information
Additional experimental procedures, calculations for the
catalysis experiments, particle size distribution histograms,
XRD patterns, HAADF-TEM image, EDS line scans, and
additional SEM images and UV−vis spectra. This material is
available free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
hyhuang@mx.nthu.edu.tw
Notes
The authors declare no competing financial interest.
−
to facilitate electron transfer from BH4 to 2-amino-5-
ACKNOWLEDGMENTS
■
nitrophenolate. The surface atom densities of Ag (111) and
Ag (100) planes are 13.8 and 12.0 atoms/nm2, respectively. An
octahedron with a higher surface silver atom density can have
more reactive sites than a cube for ionic adsorption, so the Au−
Ag octahedra are more catalytically active than the cubes. Of
course, binding energies of the molecule to different surface
planes of a metal can also be determined to evaluate the facet-
dependent catalytic activity.38 Remarkably, the same reaction
takes place so rapidly at 40 °C for both the Au−Ag octahedra
and cubes that 2-amino-5-nitrophenol is mostly converted after
10 min of reaction (see Figure S7). In fact, the 460 nm band
drops more rapidly for cubes than for octahedra after 5 min of
reaction, and clearly identifiable product bands are better
recorded for the cubes. However, the product band rises more
quickly for octahedra. Reversal of relative reactivity is possible
for the two particle shapes depending on the reaction
temperature. Mostly importantly, both Au−Ag cubes and
octahedra become highly efficient catalysts for this reaction at
higher temperatures.
This work was funded by National Science Council of Taiwan
(Grant 101-2113-M-007-018-MY3).
REFERENCES
■
(1) Mayer, K. M.; Hafner, J. H. Chem. Rev. 2011, 111, 3828−3857.
(2) McFarland, A. D.; Van Duyne, R. P. Nano Lett. 2003, 3, 1057−
1062.
(3) Hu, K.-W.; Liu, T.-M.; Chung, K.-Y.; Huang, K.-S.; Hsieh, C.-T.;
Sun, C.-K.; Yeh, C.-S. J. Am. Chem. Soc. 2009, 131, 14186−14187.
(4) Mulvihill, M. J.; Ling, X. Y.; Henzie, J.; Yang, P. J. Am. Chem. Soc.
2010, 132, 268−274.
(5) Wu, H.-L.; Tsai, H.-R.; Hung, Y.-T.; Lao, K.-U.; Liao, C.-W.;
Chung, P.-J.; Huang, J.-S.; Chen, I-C.; Huang, M. H. Inorg. Chem.
2011, 50, 8106−8111.
(6) Xu, R.; Wang, D.; Zhang, J.; Li, Y. Chem. Asian J. 2006, 1, 888−
893.
(7) Xia, X.; Zeng, J.; Oetjen, L. K.; Li, Q.; Xia, Y. J. Am. Chem. Soc.
2012, 134, 1793−1801.
(8) Tao, A.; Sinsermsuksakul, P.; Yang, P. Angew. Chem., Int. Ed.
2006, 45, 4597−4601.
H
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(9) Zhang, Q.; Li, W.; Moran, C.; Zeng, J.; Chen, J.; Wen, L.-P.; Xia,
Y. J. Am. Chem. Soc. 2010, 132, 11372−11378.
(10) (a) Yu, D.; Yam, V. W.-W. J. Am. Chem. Soc. 2004, 126, 13200−
13201. (b) Yu, D.; Yam, V. W.-W. J. Phys. Chem. B 2005, 109, 5497−
5503.
(11) (a) Tsuji, M.; Miyamae, N.; Lim, S.; Kimura, K.; Zhang, X.;
Hikino, S.; Nishio, M. Cryst. Growth Des. 2006, 6, 1801−1807.
(b) Tsuji, M.; Matsuo, R.; Jiang, P.; Miyamae, N.; Ueyama, D.; Nishio,
M.; Hikino, S.; Kumagae, H.; Kamarudin, K. S. N.; Tang, X.-L. Cryst.
Growth Des. 2008, 8, 2528−2536.
(12) Park, G.; Seo, D.; Jung, J.; Ryu, S.; Song, H. J. Phys. Chem. C
2011, 115, 9417−9423.
(13) Seo, D.; Yoo, C. I.; Jung, J.; Song, H. J. Am. Chem. Soc. 2008,
130, 2940−2941.
(14) Fan, F.-R.; Liu, D.-Y.; Wu, Y.-F.; Duan, S.; Xie, Z.-X.; Jiang, Z.-
Y.; Tian, Z.-Q. J. Am. Chem. Soc. 2008, 130, 6949−6951.
(15) Gong, J.; Zhou, F.; Li, Z.; Tang, Z. Langmuir 2012, 28, 8959−
8964.
(16) Ma, Y.; Li, W.; Cho, E. C.; Li, Z.; Yu, T.; Zeng, J.; Xie, Z.; Xia, Y.
ACS Nano 2010, 4, 6725−6734.
(17) Gom
Carbo-
N.; Pastoriza-Santos, I.; Per
Marzan
́
ez-Grana, S.; Goris, B.; Altantzis, T.; Fernan
́
dez-Lop
́
ez, C.;
ez,
ez-Juste, J.; Bals, S.; Van Tendeloo, G.; Liz-
̃
́
Argibay, E.; Guerrero-Martínez, A.; Almora-Barrios, N.; Lop
́
́
́
, L. M. J. Phys. Chem. Lett. 2013, 4, 2209−2216.
(18) Cho, E. C.; Camargo, P. H. C.; Xia, Y. Adv. Mater. 2010, 22,
744−748.
(19) Okuno, Y.; Nishioka, K.; Kiya, A.; Nakashima, N.; Ishibashi, A.;
Niidome, Y. Nanoscale 2010, 2, 1489−1493.
(20) Yo, H.; Millstone, J. E.; Li, S.; Jang, J.-W.; Wei, W.; Wu, J.;
Schatz, G. C.; Mirkin, C. A. Nano Lett. 2009, 9, 3038−3041.
(21) Langille, M. R.; Zhang, J.; Personick, M. L.; Li, S.; Mirkin, C. A.
Science 2012, 337, 954−957.
(22) Xue, C.; Millstone, J. E.; Li, S.; Mirkin, C. A. Angew. Chem., Int.
Ed. 2007, 46, 8436−8439.
(23) Langille, M. R.; Zhang, J.; Mirkin, C. A. Angew. Chem., Int. Ed.
2011, 50, 3543−3547.
(24) Yang, Y.; Wang, W.; Li, X.; Chen, W.; Fan, N.; Zou, C.; Chen,
X.; Xu, X.; Zhang, L.; Huang, S. Chem. Mater. 2013, 25, 34−41.
(25) Li, C.; Sun, L.; Sun, Y.; Teranishi, T. Chem. Mater. 2013, 25,
2580−2590.
(26) Hong, S.; Choi, Y.; Park, S. Chem. Mater. 2011, 23, 5375−5378.
(27) Wu, H.-L.; Kuo, C.-H.; Huang, M. H. Langmuir 2010, 26,
12307−12313.
(28) Chiu, C.-Y.; Huang, M. H. J. Mater. Chem. A 2013, 1, 8081−
8092.
(29) Huang, M. H.; Lin, P.-H. Adv. Funct. Mater. 2012, 22, 14−24.
(30) Liao, C.-W.; Lin, Y.-S.; Chanda, K.; Song, Y.-F.; Huang, M. H. J.
Am. Chem. Soc. 2013, 135, 2684−2693.
(31) Chung, P.-J.; Lyu, L.-M.; Huang, M. H. Chem.−Eur. J. 2011, 17,
9746−9752.
(32) Wu, J.-K.; Lyu, L.-M.; Liao, C.-W.; Wang, Y.-N.; Huang, M. H.
Chem.−Eur. J. 2012, 18, 14473−14478.
(33) Xia, X.; Zeng, J.; Oetjen, L. K.; Li, Q.; Xia, Y. J. Am. Chem. Soc.
2012, 134, 1793−1801.
(34) Langille, M. R.; Personick, M. L.; Zhang, J.; Mirkin, C. A. J. Am.
Chem. Soc. 2012, 134, 14542−14554.
(35) Zhang, J.; Langille, M. R.; Mirkin, C. A. J. Am. Chem. Soc. 2010,
132, 12502−12510.
(36) Jiang, H.-L.; Akita, T.; Ishida, T.; Haruta, M.; Xu, Q. J. Am.
Chem. Soc. 2011, 133, 1304−1306.
(37) Solanki, J. N.; Murthy, Z. V. P. Ind. Eng. Chem. Res. 2011, 50,
7338−7344.
(38) Chiu, C.-Y.; Chung, P.-J.; Lao, K.-U.; Liao, C.-W.; Huang, M. H.
J. Phys. Chem. C 2012, 116, 23757−23763.
I
dx.doi.org/10.1021/ja410663g | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX