Enhanced CO tolerance for hydrogen activation in Au-Pt dendritic heteroaggregate nanostructures

Article abstract of DOI:10.1021/ja056924+

Au-Pt heteroaggregate nanostructures were prepared by sequential reduction methods. The structures have ～11 nm Au cores with Pt "tendrils" attached to the Au surface. The heteroaggregates are active H₂ oxidation catalysts and show high activity

Full text of DOI:10.1021/ja056924+

Published on Web 01/21/2006
Enhanced CO Tolerance for Hydrogen Activation in Au-Pt Dendritic
Heteroaggregate Nanostructures
Shenghu Zhou, Kevin McIlwrath, Greg Jackson, and Bryan Eichhorn*
Departments of Chemistry and Biochemistry, and Mechanical Engineering, UniVersity of Maryland,
College Park, Maryland 20742, and Hitachi Instruments, Hitachi High Technologies America, 5100 Franklin DriVe,
Pleasanton, California 94588
Received October 10, 2005; E-mail: eichhorn@umd.edu
We report here the synthesis, characterization, and unusual
catalytic properties of Au-Pt dendritic heteroaggregate nanostruc-
tures for oxidation reactions in CO/H2 mixtures. PtAu bimetallic
nanoparticles (NPs) have been extensively studied for a variety of
catalytic applications. PtAu bimetallics as well as Au@Pt and
Pt@Au core-shell NPs have all been prepared and characterized.1-4
Reactivity studies show that the catalytic behavior in this bimetallic
system is dependent upon the NP architecture and is significantly
different from that of the individual metal NPs. In particular, a
recent study of the AuPt bimetallic NPs prepared through den-
drimer-based synthesis showed that CO oxidation was greatly
enhanced relative to the pure Pt and Au NP systems.2
Figure 1. TEM images of (a) Au NP core particles, (b) Au-Pt dendritic
aggregates (particle enlargement in inset), and (c) EELS line scan across
the Au-Pt particle showing atomic composition (Au-red, Pt-green) as a
function of beam displacement across the particle. Particle center occurs at
∼18 nm.
The use of CO-contaminated H2 fuels in low-temperature fuel
cells will be facilitated by anode catalysts that oxidize CO while
maintaining high activity for H2 electrooxidation at temperatures
under 100 °C. While CO oxidation can be significantly enhanced
in various monometallic5-8 and bimetallic nanostructures,2 these
systems have often been designed strictly for preferential oxidation
of CO at low temperatures with reduced activity for H2 oxidation.
For example, supported Au-based nanocatalysts show remarkable
CO oxidation behavior at very low temperatures,5-8 but their
ineffectiveness as H2 oxidation electrocatalysts does not make them
attractive candidates for fuel cell anode applications. In contrast,
excellent H2 activators, such as Pt NPs, are poisoned by CO below
100 °C. While some alloy NPs, such as the PtRu systems,9 have
shown promise as CO-tolerant H2 electrocatalysts, the presence of
CO above a few 100 ppm still causes a significant drop in the H2
oxidation kinetic rates below 100 °C. The lack of a mechanistic
understanding of the Pt-Ru system limits the ability for developers
to optimize this system. With their very high selectivity for CO
oxidation and high H2 thermal oxidation activity below 100 °C in
mixed CO/H2 fuel streams, the Au-Pt dendritic heteroaggregate
nanostructures reported here show potential to overcome these
limitations. Comparing this system to Pt-Ru systems suggests that
the Pt-Au NPs may be an excellent candidate for a low-temperature
anode electrocatalyst.
Monometallic Pt (5 nm) and Au (11 nm) NPs were prepared by
employing minor modifications of published procedures.10 The Au-
Pt heteroaggregate particles were prepared using modified standard
procedures for sequential core-shell particle growth. The Au core
was prepared by reducing HAuCl4 in a decahydronaphthalene/
oleylamine mixture (∼6:1, v:v) at 110 °C for 12 h. The resulting
purple colloidal solution was cooled to room temperature, charged
with Pt(acac)2 (1:1, Pt:Au), and refluxed for an additional 2 h. The
Pt-Au particles were precipitated from the resulting dark colloidal
mixture (ethanol and centrifugation), washed (acetone), and re-
suspended in toluene. All supported catalysts were prepared by
adding γ-alumina (80-120 m2 g-1, 3 µm) to the corresponding
colloidal toluene suspensions and subsequently removing the toluene
solvent in vacuo. All catalyst loadings were prepared to contain
1.0 wt % Pt. The Au control catalyst was 1.0 wt % Au.
The Au-Pt heteroaggregate NPs have been characterized by
XRD, EDX, TEM, STEM phase mapping, and EELS single particle
analysis. The TEM images (Figure 1 and S-1) show that the Au
NP precursor and the Au cores of the heteroaggregate structures
are identical in size (6-18 nm range; 11 nm, av.). The TEM image
of the bimetallic (Figure 1b) shows that the Pt metal grows from
the Au seed as dendritic tendrils (∼7 nm diameter) to form a
heteroaggregate structure that is architecturally distinct from the
bimetallic and core-shell AuPt systems previously reported.1-4 The
heteroaggregate structure is consistent with the XRD patterns of
the “as prepared” NPs, which show non-alloyed Au peaks with
shoulders arising from the Pt tendrils. Annealing the sample at 400
°C for 3 h crystallizes the Pt components to give well-defined Pt
and Au diffraction patterns (Figure S-2). STEM phase mapping
(Figure S-3) is also consistent with dendritic Pt tendrils extending
from the Au core, but the overlap in the X-ray lines complicates
the interpretation. Definitive evidence of the Au core/Pt tendril
structure is seen in the EELS line scans across individual hetero-
aggregate particles (Figures 1c and S-3). The data show that the
core is essentially pure Au, whereas the Pt density is concentrated
at the periphery of the particles. EDX analysis of the bulk sample
is consistent with a 1:1 mixture of Pt and Au. The data give a
consistent picture of a pure Au core with Pt tendrils attached to
the Au particle surface. This unique dendritic heterostructure is
similar to the dendritic manganese oxide nanostructures11 and related
nano-multipods,12 except that the bimetallic heterostructure de-
scribed here has a core that is compositionally distinct from the
tendrils. This behavior contrasts typical core-shell seeded growth
mechanisms but is consistent with the PtAu phase diagram and
related heterodimer particle syntheses.13 The aggregate structure
and the lack of any monometallic particles in the colloidal sus-
pension strongly suggests that Pt growth is templated on the Au
surface. The dendritic heteroaggregate structure of the current Au-
Pt system gives rise to anomalous catalytic properties not observed
in any of the other monometallic or bimetallic architectures.
9
1780
J. AM. CHEM. SOC. 2006, 128, 1780-1781
10.1021/ja056924+ CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Figure 2. Temperature-programmed oxidation reactions. (a) Comparison of Pt and Au-Pt heteroaggregate catalysts for H2 oxidation in H2/CO/O2 fuel
streams (50:0.2:0.5) with Ar balance. CO and O2 are normalized to inlet composition. H2O composition is normalized with respect to H2O generated if the
limiting reactant O2 were completely converted to H2O. When all O2 and CO are consumed, 80% of the O2 is used to form H2O and the other 20% is used
for conversion of CO to CO2. (b) H2 oxidation with the Au-Pt heteroaggregate catalyst in 1000 ppm CO (H2/CO/O2 fuel streams, 50:0.1:0.5) showing the
84 °C H2 lightoff. The 90% maximum for normalized H2O composition is again based on O2, and the remaining 10% is used for CO oxidation.
Temperature-programmed reactor experiments (TPR) for oxida-
tion of CO-contaminated H2 fuels were conducted to estimate the
potential of the Au-Pt heteroaggregates for CO-tolerant H2
electrooxidation catalysts. The alumina-supported catalysts were
exposed to streams of H2/CO/O2 flowing at 21 cm/s, and the exhaust
was continuously monitored by mass spectrometry (Thermo Prima
δB). Oxygen-deficient fuel streams were employed to evaluate the
discrimination between CO versus H2 oxidation. The initial
experiments were conducted with 2000 ppm CO concentrations and
a hydrogen-to-CO ratio of 250:1. The results (Figure 2a) show that
the activity of the monometallic Pt catalyst is significantly impeded
by the CO impurity, which raises the H2 lightoff temperature to
175 °C. In pure H2, the same Pt catalyst has a 50 °C lightoff and
achieves 100% O2 conversion by 60 °C. In contrast, the Au-Pt
heteroaggregate catalyst is significantly more active in the presence
of CO, showing a 105 °C lightoff under identical conditions and
Pt loadings. Note that less than 1% of the H2 is consumed and the
observed 80% yield of water is based on the limiting reagent, O2.
The remaining 20% of the O2 is used for the oxidation of CO to
CO2. Control experiments were also conducted in which Au NPs
and monometallic Au/Pt NP mixtures were co-deposited from
colloidal suspensions and evaluated under the same conditions and
loadings (Figure S-4). The Au catalysts had little activity for CO
or H2 oxidation under 200 °C. It is well-known5-8 that Au NPs
larger than 5 nm are not active for CO oxidation, and the lack of
reactivity from the 11 nm NPs described here is not surprising.
Moreover, the monometallic Au/Pt mixture displayed reactivity very
similar to that of pure Pt (Figure S-4). The remarkable enhancement
in CO tolerance observed for the Au-Pt heteroaggregate catalyst
is due to its bimetallic nature and its heteroaggregate architecture
and is not just a function of its elemental composition.
H2 lightoff in both experiments and reaches completion in the 1000
ppm CO case. This observation is consistent with CO poisoning
models in which strong CO binding to Pt surfaces blocks sites for
O2 and H2 activation. Since CO does not bind strongly to large Au
NPs, we speculate that the Au core most likely activates O2 and
facilitates14 CO oxidation at the Au-Pt interface (the triple-phase
boundary). By oxidizing the CO at the interface, the Pt tendrils are
cleansed of CO contaminates and resume their primary function
as H2 oxidation catalysts. The detailed mechanism for enhanced
CO tolerance of the Au-Pt heteroaggregate NPs is currently under
investigation in order to explore the potential use of these catalyst
for CO-tolerant anodes in low-temperature fuel cell applications.
Acknowledgment. This material is based upon work supported
by the National Science Foundation under Grant No. 0401850, and
the U.S. Department of Energy’s Oak Ridge National Lab under
the Advanced Reciprocating Engine Systems Program. We thank
Dr. Ray Tweston of Gatan for assistance with the EELS data.
Supporting Information Available: Additional TEM data, particle
phase maps, EELS data, control reactivity studies, and experimental
details (5 pages, print/PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Zhang, B.; Li, J. F.; Zhong, Q. L.; Ren, B.; Tian, Z. Q.; Zou, S. Z.
Langmuir 2005, 21, 7449-7455.
(2) Lang, H.; Maldonado, S.; Stevenson, K. J.; Chandler, B. D. J. Am. Chem.
Soc. 2004, 126, 12949-12956.
(3) Wu, M. L.; Chen, D. H.; Huang, T. C. Chem. Mater. 2001, 13, 599-606.
(4) Henglein, A. J. Phys. Chem. B 2000, 104, 2201-2203.
(5) Bulushev, D. A.; Yuranov, I.; Suvorova, E. I.; Buffat, P. A.; Kiwi-Minsker,
L. J. Catal. 2004, 224, 8-17.
(6) Daniells, S. T.; Overweg, A. R.; Makkee, M.; Moulijn, J. A. J. Catal.
2005, 230, 52-65.
(7) Kung, H. H.; Kung, M. C.; Costello, C. K. J. Catal. 2003, 216, 425-
432.
To further probe the relationship between CO contamination and
H2 lightoff, we evaluated the AuPt heteroaggregate catalyst for H2
oxidation in a 1000 ppm CO fuel stream. The results (Figure 2b)
show that H2 lightoff drops to 84 °C and 100% oxygen conversion
by 92 °C. Again, the observed 90% yield of water is based on O2
feed concentrations, and less than 1% of the H2 is consumed.
Additional tests in highly O2-deficient feeds (H2/CO/O2 - 50:0.45:
0.2) show that CO is selectively oxidized (see Figure S-4).
In the Pt and Au-Pt heteroaggregate TPR experiments, CO
oxidation and H2 lightoff are clearly linked. CO oxidation precedes
(8) Chen, M. S.; Goodman, D. W. Science 2004, 306, 252-255.
(9) Avgouropoulos, G.; Ioannides, T. Appl. Catal. B 2005, 56, 77-86.
(10) Hiramatsu, H.; Osterloh, F. E.Chem. Mater. 2004, 16, 2509-2511.
(11) Yuan, J.; Li, W. N.; Gomez, S.; Suib, S. J. Am. Chem. Soc. 2005, 127,
14184-14185.
(12) Zitoun, D.; Pinna, N.; Frolet, N.; Belin, C. J. Am. Chem. Soc. 2005, 127,
15034-15035.
(13) Gu, H.; Yang, Z.; Gao, J.; Chang, C. K.; Xu, B. J. Am. Chem. Soc. 2005,
127, 34-35.
(14) Deng, X.; Min, B. K.; Guloy, A.; Friend, C. M. J. Am. Chem. Soc. 2005,
127, 9267-9270.
JA056924+
9
J. AM. CHEM. SOC. VOL. 128, NO. 6, 2006 1781