leaching-reprecipitation process required for an efficient cat-
alysis, but at the same time it must preserve the crystalline
form of the alloy to limit the aggregation phenomenon.
With these requirements in mind, we reasoned, that gold
could be, if not ideal, a good candidate for the validation of
the proof-of-concept of our strategy since it is well known
ure 1d). Their broad line-width could be explained by the
presence of small particles (~5 nm) as well as the local
chemical composition heterogeneity of the Pd Au
nano-
x
1Àx
particles. We underline that the presence of PdO was not
observed on the diffractogram. Chemical mapping of the as-
prepared samples were acquired thanks to the scanning
TEM technique coupled with energy dispersive X-ray spec-
troscopy analysis (STEM-EDX; Figure 1e and f). It shows
homogeneous Pd–Au nanostructures and no development of
Pd-rich or Au-rich nanoparticles, excluding the common and
unwanted segregation phenomenon. The mean chemical
composition of the whole zone represented on Figure 1
could be measured; Au: 55at.% and Pd: 45at.%.
[13]
that gold forms stable alloy with palladium.
Pd and Au
atoms are miscible over a wide range of composition. The
equilibrium phase diagram predicts that a face-centred cubic
(
fcc) solid solution of Au and Pd would be formed over the
entire compositional range, with no specific ordering
[14]
(
except for Au Pd and AuPd ).
3 3
With this concept in mind, we first worked on the devel-
opment of a direct way to prepare the bimetallic Pd–Au/C
catalyst, as well as the monometallic Pd/C and Au/C cata-
lysts for comparative evaluations. We recently reported that
palladium nanoparticles could be efficiently deposited on
Although X-ray photoelectron spectroscopy (XPS) analy-
0
0
sis identified Au and Pd as the major species present in the
as-prepared material, some ionic palladium (19at.%) was
also detected (Figure S3 in the Supporting Information).
The absence of signal accounting for PdO in the XRD pat-
charcoal from a methanolic solution of Pd ACHTNUGTERNNN(UG OAc) under an
2
[15]
atmosphere of H (1 atm). We set up for an extension of
tern presented on Figure 1, suggests that Pd Au particles
2
x 1Àx
this strategy for the preparation of bimetallic Pd–Au/C cata-
lysts. To our great pleasure, we discovered that the reduction
are likely covered by a thin oxide layer. The layer is too
thin to be detected by XRD as well as HRTEM. All these
analyses definitively confirmed the formation of the expect-
ed Pd–Au alloy. Similarly, 2.5, 5 and 10% (wt.%) Pd/C and
5% Au/C catalysts were also prepared following the same
experimental procedure for their catalytic evaluation (see
the Supporting Information). The common features in both
monometallic systems were: 1) the existence of homogene-
ously distributed nanosized particles onto the carbon sup-
port, and 2) their well-defined crystalline fcc structure. How-
ever, the particle size of the Pd/C samples ranges from 2 to
of a methanolic mixture of Pd
ACHTUNGENTRNUNG( OAc) , KAuCl and charcoal
2 4
under H at 258C led to a charcoal-supported Pd–Au alloy
2
(
Scheme 1). Importantly, the mild and easily reproducible
Scheme 1. Preparation of Pd–Au/C catalysts.
3
0 nm (Figure 1a) causing broader XRD peaks than those
observed for the Au/C samples (Figure 1b). Contrary to Aun
samples, XPS analysis also revealed that about 1:3 of the
total Pd species were under their +II oxidation state. This
result suggests the formation of PdÀO bonds at the nanopar-
conditions allowed the quantitative deposition of the metal
onto the charcoal leaving the solvent almost free of metal
residues after a single filtration through a nylon membrane
(
0.45 mm), as determined by ICP-MS analysis. This feature,
ticle surface since palladium oxidises more readily than gold
(Figure 2). The extent of oxidation is higher in the monome-
tallic sample with respect to the bimetallic catalyst.
frequently overlooked by chemists preparing heterogeneous
catalysts, remains essential for developing sustainable chem-
istry since metallic wastes are not produced with our proc-
ess. Three different Pd–Au/C catalysts were prepared, ac-
cording to the metal loading on the support (i.e., 2.5, 5 and
The freshly prepared 5% Pd/C and 5% Pd–Au/C catalysts
were also analysed by electrochemistry to evaluate their cat-
alytic efficiency. For this purpose, catalysts were deposited
on a glassy carbon surface in a Nafion diffusion layer (see
the Supporting Information). In a first step, electrodes were
10 wt.% per metal). We selected charcoal as support for its
robustness toward a variety of conditions and for its low
cost.
À1
examined in acidic aqueous medium (H SO 0.1 molL ) by
2
4
With the bimetallic catalyst in hand we extensively ana-
lysed its chemical composition through a full set of standard
characterisations. Transmission electron microscopy (TEM)
investigations evidenced the formation of nanoparticles
mainly distributed onto the support. A bimodal distribution,
containing small particles in the range of 10–15 nm along
with very rare aggregated particles of about 100 nm (stand-
ard deviation ~10%) was observed (Figure S4 in Supporting
Information). The X-ray diffraction (XRD) pattern of the
bimetallic solid sample exhibits the expected reflexion peaks
of an fcc phase. The average cell parameter of 4.005 ꢄ con-
firms a bimetallic structure considering the value of the
monometallic phase (aAu =4.05 ꢄ and aPd =3.96 ꢄ; Fig-
using cyclic voltammetry. The electrodes containing 5% Pd/
C and 5% Pd–Au/C catalysts were then compared with an
electrode only covered with pure carbon. Cyclic voltammo-
grams for each electrodes (Figure S5 in the Supporting In-
formation) showed a classical metallic behaviour with redox
peaks associated to the oxidation/reduction of palladium as
well as the reduction of water beyond À0.3 V (vs. SCE) and
the H oxidation around À0.25 V. Interestingly, the hydro-
2
gen adsorption/desorption give two redox systems, exhibit-
ing two different peaks, likely attributed to two different H2
binding sites. More importantly, the over-potential of water
reduction and H oxidation on Pd is not affected by the
2
presence of Au, but the charge required for the anodic oxi-
&
2
&
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!