Letter
pubs.acs.org/NanoLett
Synthesis of Pt−Ni Octahedra in Continuous-Flow Droplet Reactors
for the Scalable Production of Highly Active Catalysts toward
Oxygen Reduction
Guangda Niu,†,‡ Ming Zhou,† Xuan Yang,† Jinho Park,§ Ning Lu,∥ Jinguo Wang,∥ Moon J. Kim,∥
Liduo Wang,‡ and Younan Xia*,†,§
†
The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Atlanta,
Georgia 30332, United States
‡
Key Lab of Organic Optoelectronics and Molecular Engineering of Ministry of Education, Department of Chemistry, Tsinghua
University, Beijing 100084, China
§
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332, United States
∥
Department of Materials Science and Engineering, University of Texas at Dallas, Richardson, Texas 75080, United States
S Supporting Information
*
ABSTRACT: A number of groups have reported the syntheses of nanosized Pt−Ni
octahedra with remarkable activities toward the oxygen reduction reaction (ORR), a
process key to the operation of proton-exchange membrane fuel cells. However, the
throughputs of those batch-based syntheses are typically limited to a scale of 5−25 mg
Pt per batch, which is far below the amount needed for commercial evaluation. Here
we report the use of droplet reactors for the continuous and scalable production of
Pt−Ni octahedra with high activities toward ORR. In a typical synthesis, Pt(acac)2,
Ni(acac)2, and W(CO)6 were dissolved in a mixture of oleylamine, oleic acid, and
benzyl ether, and then pumped into a polytetrafluoroethylene tube. When the solution
entered the reaction zone at a temperature held in the range of 170−230 °C, W(CO)6
quickly decomposed to generate CO gas, naturally separating the reaction solution
into discrete, uniform droplets. Each droplet then served as a reactor for the
nucleation and growth of Pt−Ni octahedra whose size and composition could be
controlled by changing the composition of the solvent and/or adjusting the amount of Ni(acac)2 added into the reaction
solution. For a catalyst based on Pt2.4Ni octahedra of 9 nm in edge length, it showed an ORR mass activity of 2.67 A mgPt−1 at 0.9
V, representing an 11-fold improvement over a state-of-the-art commercial Pt/C catalyst (0.24 A mgPt−1).
KEYWORDS: droplet reactors, continuous production, scale up, Pt−Ni octahedra, oxygen reduction reaction
P
ORR.5−7 By introducing a solvent to modify the protocol, our
group demonstrated the synthesis of Pt−Ni 9 nm octahedra
with specific and mass activities of 10.6 mA cm−2 and 3.3 A
mgPt−1 at 0.9 V vs reversible hydrogen electrode (RHE), almost
45 and 17 times of those of a state-of-the-art Pt/C catalyst,
respectively.8 Later on, we further demonstrated the synthesis
of Pt−Ni octahedra of 6 and 12 nm in edge length by adjusting
the amount of oleylamine (OAm) introduced into the reaction
system.10 Most recently, Huang and co-workers reported the
synthesis of Mo-doped Pt3Ni octahedra with record-setting
specific and mass activities of 10.3 mA cm−2 and 6.98 A mgPt−1
at 0.9 V.9 On the other hand, Yang, Stamenkovic, and their coworkers demonstrated the fabrication of Pt−Ni nanoframes
with specific and mass activities of 8.48 mA cm−2 and 5.7 A
mgPt−1 at 0.9 V.11
roton-exchange membrane fuel cells (PEMFCs) offer a
practical solution to zero-emission engines, but it is still
challenging to market this technology on an industrial scale
primarily due to the cost issue. The cost of a PEMFC can be
largely attributed to the Pt catalyst deposited on the cathode to
mitigate the sluggish kinetics of oxygen reduction reaction
(ORR).1 Combining Pt with a 3d8 transition metal such as Fe,
Co, or Ni has been demonstrated as one of the most effective
strategies for enhancing the specific activity of Pt and thus
reducing its loading in an ORR catalyst.2−11 In particular, it was
reported in 2007 by Stamenkovic and co-workers that
Pt3Ni(111) single-crystal substrates had a remarkable specific
activity of more than 90-fold greater than that of a commercial
Pt/C catalyst.3 Motivated by this incredible improvement,
many synthetic protocols have been developed for generating
Pt−Ni nanocrystals with different sizes, shapes, and elemental
compositions.4−10 Specifically, Fang, Zou, Yang, and their coworkers reported some of the earliest syntheses of Pt−Ni
nanosized octahedra and cubes using a solution-phase method,
albeit their samples only showed modest activities toward
© 2016 American Chemical Society
Received: March 30, 2016
Revised: April 24, 2016
Published: May 2, 2016
3850
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
colloidal nanocrystals.17 This technique offers many attractive
features, including relatively fast thermal/mass transfer in in
individual droplets18 and the ability to rapidly screen the
reaction parameters19 and run multiple syntheses in parallel.20
We have demonstrated the use of this technique for the scalable
production of Pd, Au, and Pd−M (M = Au, Pt, and Ag)
nanocrystals with well-controlled sizes, shapes, and compositions.21 In such a system, the total volume of production (Vtotal)
can be expressed as Vtotal = Vd × f × t, where Vd is the volume of
each droplet, f is the frequency at which the droplets are
generated, and t is the duration of time during which the
synthesis is conducted. The linear relationship between Vtotal
and the duration of synthesis allows for consistent, scalable
production without compromising the quality and reproducibility of the products. Either an immiscible liquid or a gas can
serve as the carrier phase to separate the reaction solution into
consecutive droplets in a fluidic channel.22 In practice, it is
more attractive to use a gas rather than a liquid as the carrier
phase as it will allow easy collection of the products without
involving any separation. It would be even more advantageous
to design a reaction system that can automatically evolve into
discrete droplets in a fluidic channel during the synthesis
without the use of any external gas.
In this paper, we report a technique for the continuous and
scalable production of Pt−Ni octahedral nanocrystals in droplet
reactors with CO gas generated in situ as the carrier phase.
Prior to the introduction into a fluidic device, all the reagents,
including Pt(acac)2, Ni(acac)2, and W(CO)6, were dissolved in
a mixture of oleic acid (OAc), OAm, and benzyl ether (BE).
When the reaction solution entered the reaction zone, W(CO)6
immediately decomposed to generate CO gas, which could
serve as both the reducing agent and carrier phase. The CO gas
could separate the reaction solution confined in a fluidic
channel into uniform droplets, with the size of each droplet
being largely determined by the amount of W(CO)6 added into
the reaction solution. We could obtain Pt−Ni octahedral
nanocrystals with different compositions and sizes by adjusting
the amount of Ni(acac)2 added, as well as the amounts of OAm
and OAc in the reaction mixture. Using a fluidic device based
on polytetrafluoroethylene (PTFE) tube of 1.58 mm in inner
diameter, we could produce 9 nm Pt−Ni octahedra at a
throughput of 20 mg per hour. The throughput was increased
to 160 mg per hour by switching to PTFE tube with an inner
diameter of 3.0 mm. When operated continuously and with
multiple fluidic devices in parallel, it will be feasible to scale up
the production of Pt−Ni ORR catalyst to hundreds of grams on
a daily basis. The Pt−Ni/C catalyst based on 9 nm octahedra
showed an ORR mass activity of 2.67 A mgPt−1 at 0.9 V, an
improvement of 11-fold relative to a commercial Pt/C catalyst
(0.24 A mgPt−1).
For the batch synthesis of Pt−Ni octahedra reported in
literature, W(CO)6 was introduced into the reaction solution as
a solid powder.8 However, it is not a good idea to use solid
powder with a fluidic device because the powder may
precipitate out from the reaction solution, clogging the fluidic
channel. In the present work, we addressed this issue by
dissolving all chemicals, including Pt(acac)2, Ni(acac)2, and
W(CO)6, in a mixture of OAc, OAm, and BE at a carefully
selected temperature of 70 °C. Because the decomposition
temperature of W(CO)6 is around 170 °C, this compound
should not decompose at 70 °C. Meanwhile, Pt(acac)2 and
Ni(acac)2 should not be reduced by OAm either at this
relatively low temperature.
Figure 1. (a) Schematic illustration of the fluidic device used for
generating droplets in situ, followed by the formation of Pt−Ni
octahedral nanocrystals within each droplet. (b) Photograph showing
the formation of droplets in a PTFE tube during the continuous
pumping of a precursor solution containing W(CO) 6 at a
concentration of 2.0 mg mL−1.
Despite those impressive results, there still exists a major gap
in moving the catalysts based on Pt−Ni octahedra from
academia to industry. The gap can be mainly attributed to the
inability to produce the bimetallic nanocrystals at an industrially
relevant scale while still maintaining a tight control over their
physicochemical properties. For a batch-based protocol, the
throughput is typically limited to a scale of 5−25 mg Pt per
batch,9,12 which is barely enough for a single test based on
membrane electrode assembly (MEA), not to mention the
much larger quantity needed for the commercial use at an
industrial scale. 13 Due to the lack of batch-to-batch
reproducibility, it is impractical to increase the quantity of
catalysts by combining the products from multiple runs of
batch syntheses.14 An alternative approach is to increase the
volume of reaction solution used in each run of synthesis by
switching to a larger reactor. However, due to the intrinsic
difficult in achieving thermal and compositional homogeneity
over a large volume, it has proven challenging to obtain
products with high quality and good uniformity when the
reaction volume is scaled up.15 In a new approach, Peng and coworkers reported a scalable method for the production of Pt−
Ni octahedra by impregnating both the Pt and Ni precursors in
a carbon support, followed by their coreduction in a mixture of
H2 and CO gases to generate Pt−Ni/C catalysts.16 However,
the nanoparticles in the products still need to be improved in
terms of size and shape uniformity in order to maximize their
catalytic activity toward ORR. In general, the use of H2 and CO
may increase the level of risk due to the flammability and
toxicity of these two gases. As such, there is still an urgent need
to develop a new technique for the continuous and scalable
production of Pt−Ni nanocrystals with well-defined shapes, as
well as tunable sizes and compositions.
Synthesis in continuous-flow droplet reactors has recently
emerged as a practical route to the scalable production of
3851
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
Figure 2. (a−c) TEM images of the products obtained using the same protocol developed for the 9 nm Pt−Ni octahedral nanocrystals, except for
the introduction of W(CO)6 into the precursor solution at different concentrations: (a) 2.0, (b) 1.0, and (c) 5.0 mg mL−1. The insets illustrate the
formation of droplets or plugs in the PTFE tube. When W(CO)6 was used at 5.0 mg mL−1, the solution in the PTFE tube would jet out from the
outlet. (d) TEM image of Pt−Ni nanocrystals formed when Ar gas was supplied as the carrier phase, with all other conditions being kept the same as
in (a).
After the reaction solution had cooled to room temperature,
it was injected into a fluidic device made of PTFE tubes (1.58
mm in inner diameter) at a rate of 0.15 mL min−1. A portion of
the tube (the reaction zone) was immersed in an oil bath held
at 230 °C. The residence time, determined by the flow rate and
the length of PTFE tube, has to be kept at 30 min or longer to
ensure the complete development of an octahedral shape for
the Pt−Ni nanocrystals. As long as this requirement is met, the
length of the PTFE tube should not affect the quality of the
Pt−Ni octahedra. Figure 1a and b, shows a schematic of the
setup and a photograph of the actual device. As soon as the
solution entered the reaction zone, CO was quickly released
due to the decomposition of W(CO)6, separating the flowing
solution into a train of multiple droplets. The color of the
solution in each droplet quickly turned from yellow to dark
brown, indicating the formation of nanocrystals. The separation
distance between adjacent droplets could be kept the same to
maintain a steady flow of uniform droplets. Figure 1b shows a
photograph of the droplets when 2.0 mg mL−1 W(CO)6 was
introduced into the reaction solution. This concentration was
adequate to separate the solution into discrete droplets, and at
the same time, to ensure the formation of Pt−Ni octahedral
nanocrystals with a uniform size and shape (see Figure 2a for a
typical sample). Under this set of experimental conditions, we
could routinely produce 20 mg of Pt−Ni octahedra on an
hourly basis using a single fluidic device.
We also investigated the effect of W(CO)6 concentration in
determining the morphology of Pt−Ni nanocrystals in a set of
control experiments. When the concentration of W(CO)6 was
reduced to 1.0 mg mL−1, plugsinstead of dropletswere
formed in the PTFE tube, as shown in Figure 2b (see the inset
for an illustration) and Figure S1. The resultant Pt−Ni
nanocrystals had a broad size distribution, with a majority of
truncated octahedral shapes (Figure 2b). According to the
nucleation and growth theory proposed by LaMer and coworkers, the polydispersed size can be attributed to the overlap
of nucleation and growth processes during the synthesis, as a
result of the reduced amount of CO gas and thus a slow
reduction rate.23 When 5.0 mg mL−1 W(CO)6 was used (the
same as in a batch synthesis), the excessive CO generated in the
PTFE tube led to jetting for the reaction solution (Figure 2c
and the inset). The resultant particles showed irregular shapes.
On the other hand, when Ar was introduced as the carrier
phase, with 2.0 mg mL−1 W(CO)6 in the reaction phase, the
products were also poorly defined in terms of shape. For Pt−Ni
octahedral nanocrystals, CO can serve as a capping agent for
the {111} facets.5 Upon Ar was employed as the carrier phase,
the amount of CO dissolved into the solution became
inadequate in promoting the formation of {111} facets on
3852
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
Figure 3. TEM images of octahedral nanocrystals with different sizes and compositions: (a) 9 nm Pt1.7Ni, (b) 9 nm Pt3.0Ni, (c) 6 nm Pt2.1Ni, and (d)
6 nm Pt2.7Ni. All the compositions were determined using ICP-MS, after the nanocrystals has been treated with HAc.
Table 1. Control over the sizes and compositions of Pt−Ni octahedral nanocrystals by adjusting the amounts of precursors and
solvents
experimental conditions
edge length
Pt(acac)2 (mmol)
Ni(acac)2 (mmol)
OAm (mL)
OAc (mL)
BE (mL)
Pt/Ni atomic ratio (before HAc)
Pt/Ni atomic ratio (after HAc)
9 nm
0.051
0.051
0.051
0.051
0.051
0.051
0.039
0.031
0.019
0.088
0.058
0.029
2
2
2
0.5
0.5
0.5
1
1
1
2
2
2
7
7
7
7
7
7
1.3
1.5
2.8
0.8
1.4
2.1
1.7
2.4
3.0
2.1
2.7
2.8
6 nm
the resultant nanocrystals. Taken together, W(CO)6 has to be
used at an optimal concentration of 2.0 mg mL−1 in order to
ensure the generation of a steady flow of droplets and the
production of Pt−Ni octahedral nanocrystals with both
uniform size and well-defined shape. We further optimized
the reaction temperature of the fluidic system (Figure S2). With
reaction temperature at 170 and 200 °C, the Pt/Ni atomic
ratios after HAc treatments were 3.41 and 2.78, respectively,
higher than that at 230 °C (Table S1).24 It might be caused by
the slower reducing rate of Ni(acac)2 at lower temperatures.
Moreover, for the sample obtained at 170 °C, there were some
irregularly shaped nanocrystals rather than well-defined
octahedra, resulting from the deficient amount of Ni during
reaction.10
Once the concentration of W(CO)6 and the reaction
temperature were optimized, the technique could be readily
applied to the production of Pt−Ni octahedra with tunable
compositions and sizes. In a standard synthesis, we constructed
the fluidic devices from PTFE tunes with an inner diameter of
1.58 mm. We obtained 9 nm Pt−Ni octahedra by codissolving
0.051 of mmol Pt(acac)2, 0.031 mmol of Ni(acac)2, and 20 mg
of W(CO)6 in a mixture of 2 mL of OAm, 1 mL of OAc, and 7
mL of BE. The as-obtained Pt−Ni octahedra were then treated
with acetic acid (HAc) at 60 °C for 2 h to remove most
surfactants on the surface. Figure 2a shows a typical TEM
image of the products after treatment with HAc, which can be
assigned as 9 nm Pt2.4Ni octahedra according to the TEM and
ICP-MS analyses. The Pt/Ni atomic ratio of the octahedra
could be tuned by adjusting the amount of Ni(acac)2 in the
3853
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
Figure 4. (a) TEM image of the 9 nm Pt2.4Ni octahedral nanocrystals supported on carbon. (b) HAADF-STEM image for an individual octahedral
nanocrystal. (c) HAADF-STEM image and (d) EDS line scan showing Pt and Ni distributions along the arrow marked in (c).
Prior to catalytic characterization, we used attenuated total
reflectance Fourier transform infrared spectroscopy (ATRFTIR) to evaluate the surface cleanness of the Pt−Ni octahedra
before and after HAc treatment (Figure S4). For the asobtained sample, we observed the C−H vibration peaks at 2922
and 2854 cm−1, whereas the peak at 1071 cm−1 can be assigned
to C−N vibration. These peaks suggest the presence of OAm
on the surface of the as-obtained Pt−Ni octahedra.26 After HAc
treatment, those peaks disappeared, indicating the improvement in terms of surface cleanness. In addition, we used X-ray
photoelectron spectroscopy (XPS) to study the electronic
structure of the Pt−Ni octahedra. For the Ni 2p XPS spectrum
(Figure S5a), in addition to zerovalent Ni metal, we could also
find Ni2+ in the as-obtained Pt−Ni nanocrystals, probably due
to the inclusion of Ni(OH)2.11 The HAc treatment could easily
remove Ni2+ on the surface of the nanocrystals. As a result, the
signals from Ni2+ dramatically decreased for the HAc-treated
sample, accompanied by an increase in the Pt/Ni atomic ratio
(Table 1). The Pt 4f spectrum (Figure S5b) indicates the
dominance of zero-valent Pt in both the as-obtained and HActreated samples. Compared to pure Pt metal, the binding
energy of Pt 4f peaks for the Pt−Ni nanocrystals was increased
by 0.4 eV.27 Meanwhile, the W 4f spectrum (Figure S5c)
suggests the presence of WO3 in the as-obtained sample of Pt−
Ni octahedra, confirming the hypothesis proposed by Fang and
co-workers that W decomposed from W(CO)6 could also serve
as a reductant for the reaction.5 However, the WO3 could be
removed by repeatedly washing the sample with toluene or
acetic acid. This result indicates that W was not incorporated
reaction solution. We obtained 9 nm Pt1.7Ni (Figure 3a) and
Pt3.0Ni (Figure 3b) octahedra when 0.039 and 0.019 mmol of
Ni(acac)2 were added in the reaction solution, respectively. The
size of the Pt−Ni octahedra could be varied by adjusting the
amount of OAm added into the reaction solution. According to
the understanding achieved from batch synthesis, OAm not
only serves as a surface stabilizer, but also forms coordination
complexes with the Pt ions.10,25 The coordination complexes
were more stable than Pt(acac)2, slowing down the reduction
process. When the amount of OAm was reduced from 2 to 0.5
mL, the reduction rate for the Pt precursor during the
nucleation process increased to form more seeds, resulting in
Pt−Ni octahedra of 6 nm in edge length. In addition, OAc
could form complexes with Ni precursor, leading to a
considerably slower reduction rate for the Ni precursor,
whereas BE served as a solvent to reduce the coverage density
of OAc and OAm on the surface of Pt−Ni octahedra.10,24 We
successfully synthesized 6 nm Pt2.1Ni octahedra (Figure 3c)
using a reaction solution containing 0.051 mmol of Pt(acac)2,
0.088 mmol of Ni(acac)2, 20 mg of W(CO)6, 0.5 mL of OAm,
2 mL of OAc, and 7 mL of BE. When the amount of Ni(acac)2
was reduced to 0.058 mmol while keeping all other parameters
unchanged, 6 nm Pt2.7Ni octahedra were obtained (Figure 3d).
We also conducted a synthesis with a fluidic device assembled
from PTFE tubes of 3.0 mm in inner diameter. As
demonstrated by Figure S3, the octahedral shape, as well as
the uniformity in terms of shape and size, was essentially
maintained.
3854
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
Figure 5. Electrocatalytic properties of the octahedral Pt2.4Ni/C catalyst for oxygen reduction reaction. (a) Cyclic voltammetry curves of the Pt2.4Ni/
C (9 nm, after treatment with HAc), and commercial Pt/C catalysts (20 wt % Pt) in a N2-saturated 0.1 M HClO4 solution. Scanning rate = 100 mV
s−1. (b) Comparison of positive-going ORR polarization curves of the catalysts in an O2-saturated 0.1 M HClO4 solution. Scanning rate = 10 mV s−1.
Rotation speed = 1600 rpm. The currents were normalized to the geometric area of the rotating disk electrode (0.196 cm2). (c) ECSA and (d) mass
activity at 0.9 V vs RHE of the catalysts before and after the accelerated durability tests.
recorded at room temperature in a N2-saturated 0.1 M HClO4
solution at a sweeping rate of 100 mV s−1 in the potential range
of 0.08−1.1 V. A broad pair of hydrogen adsorption/desorption
(HUPD) peaks appeared between 0.05 and 0.4 V for 9 nm
Pt2.4Ni/C. We calculated the electrochemically active surface
area (ECSAs) of the catalysts from the charges associated with
the average of hydrogen adsorption and desorption, and the
values were 31.4 and 63.6 for the Pt−Ni/C and Pt/C catalysts,
respectively. We also derived ECSAs from Cu underpotential
deposition (CuUPD, Figure S6b and e) by assuming a charge
density of 480 μC cm−2 Pt.28 The ECSA of the Pt2.4Ni/C
catalyst was 46.4 m2 gPt−1, also smaller than that of the
commercial Pt/C catalyst (69.4 m2 gPt−1). Figure 5b shows the
positive-going ORR polarization curves of the catalysts. The
curves were recorded in an O2-saturated HClO4 solution at
room temperature with a scanning rate of 10 mV s−1. The
kinetic currents of the ORR polarization curves were calculated
by following the Koutecky−Levich equation and then
normalized against the ECSA and Pt mass to obtain the
specific and mass activities (ik,specific and ik,mass), respectively.29
As listed in Table 2, the 9 nm Pt2.4Ni/C catalyst had a
specific activity of 5.74 mA cm−2 at 0.9 V, 17 times greater than
that of the commercial Pt/C catalyst (0.34 mA cm−2). The
specific activity was lower than what was reported in our
previous publication (7.0 mA cm−2), which employed catalyst
obtained through a batch synthesis.8 The difference was
possibly due to the higher exposure of {100} facets on the
catalyst in this work, which show a lower specific activity than
{111} facets in aqueous HClO4 solutions.3 The mass activity of
Table 2. Comparison of the Specific Electrochemically
Active Surface Area, ECSA, Specific Activity, SA, and Mass
Activity, MA, toward ORR for the Commercial Pt/C and 9
nm Pt2.4Ni/C Catalysts
Pt/C
9 nm
Pt2.4Ni/C
specific ECSA (m2
gPt−1)
SA at 0.9 V (mA
cm−2)
MA at 0.9 V (A
mgPt−1)
69.4
46.4
0.34
5.74
0.24
2.67
into the crystal lattice of Pt−Ni octahedra as atoms under the
experimental conditions used.5
Figure 4a shows TEM image of a Pt−Ni/C catalyst 9 nm
Pt2.4Ni octahedra supported on carbon after the treatment with
HAc. The high-angle annular dark-field scanning TEM
(HAADF-STEM) image in Figure 4b was taken from an
individual octahedron along the [011] zone axis, demonstrating
a single-crystal structure with {111} facets on the side faces and
{100} facets on the truncated corners. The lattice spacing
between the {111} planes of Pt−Ni was 2.24 Å, smaller than
that (2.27 Å) of face-centered cubic (fcc) Pt.8 The energydispersive X-ray spectroscopy (EDS) line scan across an
octahedron confirms that the Pt and Ni elements were
uniformly distributed throughout the nanocrystal (Figure 4c
and d).
We then used the rotating disk electrode (RDE) technique
to evaluate the electrochemical properties of the 9 nm Pt2.4Ni/
C catalyst by benchmarking against a commercial Pt/C catalyst.
Figure 5a shows their cyclic voltammograms (CVs), which were
3855
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Nano Letters
■
the 9 nm Pt2.4Ni/C catalyst was 2.67 A mgPt−1, 11 times higher
than that of the commercial Pt/C catalyst (0.24 A mgPt−1). We
also characterized the 6 nm Pt2.7Ni/C catalyst (Figure S7),
which gave an ECSA of 52.7 m2 gPt−1, together with a specific
activity of 3.34 mA cm−2 and a mass activity of 1.76 A mgPt−1 at
0.9 V. The dependence of specific activity on the particle size is
consistent with our previous report.10 This trend is probably
due to the increased coordination number for the surface atoms
and thus the decreased binding energies of oxygen species as
the catalytic particles became larger.30
The durability of the catalysts was evaluated through
accelerated tests by applying linear potential sweeps in the
range of 0.6−1.1 V at a rate of 100 mV s−1 in an O2-saturated
HClO4 solution at room temperature. As shown in Figure 5c,
the specific ECSA of the 9 nm Pt2.4Ni/C catalyst only dropped
by 7.7% to 42.8 m2 gPt−1 after 5000 cycles and by 13.4% to 40.2
m2 gPt−1 after 10 000 cycles. In comparison, the specific ECSA
of the commercial Pt/C catalyst showed about 47.0% loss after
5000 cycles and 59.7% loss after 10 000 cycles. At 0.9 V, the
mass activity of the 9 nm Pt−Ni/C catalyst dropped to 1.13 A
mgPt−1 after 5000 cycles and 1.07 A mgPt−1 after 10 000 cycles
(Figure 5d), still far exceeding the pristine Pt/C catalyst. We
found that after the durability test, the Pt−Ni/C catalyst was
composed of spherical nanoparticles with a diameter of 8−9 nm
(Figure S8). Compared to the sample before durability test, the
percentage of {111} facets on the surface decreased, leading to
the reduced specific and mass activity. Moreover, the Pt/Ni
atomic ratio of the catalyst had increased from 2.4 to 3.4 due to
the leaching of Ni during the durability test. This deviation of
Pt/Ni atomic ratio from the optimal value for the highest ORR
activity also accounted for the loss of the specific and mass
activity.10
In summary, we have demonstrated the synthesis of Pt−Ni
octahedra with controllable sizes and compositions in
continuous-flow droplet reactors. The use of droplet reactors
allows us to linearly increase the throughput of production.
Different from prior demonstrations, the carrier phase of CO
gas could be generated in situ through the decomposition of
W(CO)6 in the reaction phase. The W(CO)6 has to be used at
an optimal concentration to ensure the generation of uniform
droplets as a steady flow, as well as the formation of octahedral
nanocrystals with well-defined shape and uniform size. The
composition of the Pt−Ni octahedra could be manipulated by
adjusting the amount of Ni(acac)2 added into the reaction
solution, whereas the particle size could be tuned by varying the
volumes of OAm and OAc involved. This work offers a
practical approach to the continuous and scalable production of
Pt−Ni octahedral nanocrystals to be used as ORR catalysts for
large-scale commercialization of the PEMFC technology.
■
Letter
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected].
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
This work was supported by startup funds from the Georgia
Institute of Technology. As visiting students from Tsinghua
University and Chongqing University, respectively, G.N. and
M.Z. were also partially supported by the China Scholarship
Council (CSC).
■
REFERENCES
(1) Steele, B. C.; Heinzel, A. Nature 2001, 414, 345−352.
(2) Guo, S.; Li, D.; Zhu, H.; Zhang, S.; Markovic, N. M.;
Stamenkovic, V. R.; Sun, S. Angew. Chem., Int. Ed. 2013, 52, 3465−
3468.
(3) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G.; Ross, P.
N.; Lucas, C. A.; Marković, N. M. Science 2007, 315, 493−497.
(4) Cui, C.; Gan, L.; Heggen, M.; Rudi, S.; Strasser, P. Nat. Mater.
2013, 12, 765−771.
(5) Zhang, J.; Yang, H.; Fang, J.; Zou, S. Nano Lett. 2010, 10, 638−
644.
(6) Wu, J.; Zhang, J.; Peng, Z.; Yang, S.; Wagner, F. T.; Yang, H. J.
Am. Chem. Soc. 2010, 132, 4984−4985.
(7) Wu, J.; Gross, A.; Yang, H. Nano Lett. 2011, 11, 798−802.
(8) Choi, S.-I.; Xie, S.; Shao, M.; Odell, J. H.; Lu, N.; Peng, H.-C.;
Protsailo, L.; Guerrero, S.; Park, J.; Xia, X.; Wang, J.; Kim, M. J.; Xia, Y.
Nano Lett. 2013, 13, 3420−3425.
(9) Huang, X.; Zhao, Z.; Cao, L.; Chen, Y.; Zhu, E.; Lin, Z.; Li, M.;
Yan, A.; Zettl, A.; Wang, Y. M.; Duan, X.; Mueller, T.; Huang, Y.
Science 2015, 348, 1230−1234.
(10) Choi, S.-I.; Xie, S.; Shao, M.; Lu, N.; Guerrero, S.; Odell, J. H.;
Park, J.; Wang, J.; Kim, M. J.; Xia, Y. ChemSusChem 2014, 7, 1476−
1483.
(11) Chen, C.; Kang, Y.; Huo, Z.; Zhu, Z.; Huang, W.; Xin, H. L.;
Snyder, J. D.; Li, D.; Herron, J. A.; Mavrikakis, M.; Chi, M.; More, K.
L.; Li, Y.; Marković, N. M.; Somorjai, G. A.; Yang, P.; Stamenkovic, V.
R. Science 2014, 343, 1339−1343.
(12) Wu, J.; Qi, L.; You, H.; Gross, A.; Li, J.; Yang, H. J. Am. Chem.
Soc. 2012, 134, 11880−11883.
(13) Han, B.; Carlton, C. E.; Kongkanand, A.; Kukreja, R. S.;
Theobald, B. R.; Gan, L.; O’Malley, R.; Strasser, P.; Wagner, F. T.;
Shao-Horn, Y. Energy Environ. Sci. 2015, 8, 258−266.
(14) Zhang, L.; Xia, Y. Adv. Mater. 2014, 26, 2600−2606.
(15) Rathmell, A. R.; Bergin, S. M.; Hua, Y.; Li, Z.; Wiley, B. J. Adv.
Mater. 2010, 22, 3558−3563.
(16) Zhang, C.; Hwang, S. Y.; Trout, A.; Peng, Z. J. Am. Chem. Soc.
2014, 136, 7805−7808.
(17) Niu, G.; Ruditskiy, A.; Vara, M.; Xia, Y. Chem. Soc. Rev. 2015,
44, 5806−5820.
(18) Song, H.; Tice, J. D.; Ismagilov, R. F. Angew. Chem., Int. Ed.
2003, 42, 768−772.
(19) Trivedi, V.; Doshi, A.; Kurup, G. K.; Ereifej, E.; Vandevord, P. J.;
Basu, A. S. Lab Chip 2010, 10, 2433−2442.
(20) Nightingale, A. M.; Bannock, J. H.; Krishnadasan, S. H.;
O’Mahony, F. T. F.; Haque, S. A.; Sloan, J.; Drury, C.; McIntyre, R.;
deMello, J. C. J. Mater. Chem. A 2013, 1, 4067−4076.
(21) Zhang, L.; Niu, G.; Lu, N.; Wang, J.; Tong, L.; Wang, L.; Kim,
M. J.; Xia, Y. Nano Lett. 2014, 14, 6626−6631.
(22) Song, H.; Chen, D. L.; Ismagilov, R. F. Angew. Chem., Int. Ed.
2006, 45, 7336−7356.
(23) LaMer, V. K.; Dinegar, R. H. J. Am. Chem. Soc. 1950, 72, 4847−
4854.
(24) Ahrenstorf, K.; Heller, H.; Kornowski, A.; Broekaert, J. A. C.;
Weller, H. Adv. Funct. Mater. 2008, 18, 3850−3856.
ASSOCIATED CONTENT
S Supporting Information
*
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.nanolett.6b01340.
The experimental details are provided, together with
photographs of the flow system, TEM images, ATRFTIR spectra, XPS, CV curves, CuUPD, ORR polarization
curves of the catalyst. (PDF)
3856
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857
Letter
Nano Letters
(25) Saita, S.; Maenosono, S. Chem. Mater. 2005, 17, 6624−6634.
(26) Mourdikoudis, S.; Liz-Marzán, L. M. Chem. Mater. 2013, 25,
1465−1476.
(27) Barr, T. L. J. Phys. Chem. 1978, 82, 1801−1810.
(28) Shao, M.; Odell, J. H.; Choi, S. I.; Xia, Y. Electrochem. Commun.
2013, 31, 46−48.
(29) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Appl.
Catal., B 2005, 56, 9−35.
(30) Shao, M.; Peles, A.; Shoemaker, K. Nano Lett. 2011, 11, 3714−
3719.
3857
DOI: 10.1021/acs.nanolett.6b01340
Nano Lett. 2016, 16, 3850−3857