On the influence of diphosphine ligands on the chemical order in small RuPt nanoparticles: Combined structural and surface reactivity studies

Article abstract of DOI:10.1039/c2dt31646c

Diphenylphosphinobutane (dppb) stabilized bimetallic RuPt nanoparticles were prepared by co-decomposition of [Ru(COD)(COT)] [(1,5-cyclooctadiene)(1,3,5- cyclooctatriene)ruthenium] and [Pt(CH₃)₂(COD)] [dimethyl(1,5-cyclooctadiene) platinum(ii)] organometallic precursors under mild conditions (room temperature, 3 bar of dihydrogen) and in the presence of dppb. The determination of the nanoparticles' chemical composition was made possible thanks to a combination of several characterization techniques (HREM, STEM-HAADF, WAXS, EXAFS, IR, NMR) associated with surface reactivity studies based on simple catalytic reactions. The obtained nanoparticles display a ruthenium rich core and a disordered shell containing both ruthenium and platinum. The results were compared with those obtained on nanoparticles of similar size and composition but not containing ligands. The complexity observed in the present structure of these nanoparticles arises from the high chemical affinity of the diphosphine ligand used as a stabilizer for both metals.

Full text of DOI:10.1039/c2dt31646c

Dalton
Transactions
An international journal of inorganic chemistry
www.rsc.org/dalton
Volume 42 | Number 2 | 14 January 2013 | Pages 301–592
ISSN 1477-9226
COVER ARTICLE
Philippot, Chaudret et al.
On the influence of diphosphine ligands on the chemical order in small
RuPt nanoparticles: combined structural and surface reactivity studies

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
On the influence of diphosphine ligands on the
chemical order in small RuPt nanoparticles: combined
Cite this: Dalton Trans., 2013, 42, 372
structural and surface reactivity studies†
a,b
a,b
c
c
Patricia Lara, Tuğçe Ayvalı, Marie-José Casanove, Pierre Lecante,
d
e
a,b
e
Alvaro Mayoral, Pier-Francesco Fazzini, Karine Philippot* and Bruno Chaudret*
Diphenylphosphinobutane (dppb) stabilized bimetallic RuPt nanoparticles were prepared by co-
decomposition of [Ru(COD)(COT)] [(1,5-cyclooctadiene)(1,3,5-cyclooctatriene)ruthenium] and [Pt-
(
3 2
CH ) (COD)] [dimethyl(1,5-cyclooctadiene) platinum(II)] organometallic precursors under mild conditions
(
room temperature, 3 bar of dihydrogen) and in the presence of dppb. The determination of the nano-
particles’ chemical composition was made possible thanks to a combination of several characterization
techniques (HREM, STEM-HAADF, WAXS, EXAFS, IR, NMR) associated with surface reactivity studies based
on simple catalytic reactions. The obtained nanoparticles display a ruthenium rich core and a disordered
shell containing both ruthenium and platinum. The results were compared with those obtained on nano-
particles of similar size and composition but not containing ligands. The complexity observed in the
present structure of these nanoparticles arises from the high chemical affinity of the diphosphine ligand
used as a stabilizer for both metals.
Received 23rd July 2012,
Accepted 14th September 2012
DOI: 10.1039/c2dt31646c
www.rsc.org/dalton
to find efficient ways of synthesis but also to have as complete
as possible characterization tools to define precisely their
Introduction
7
Metal nanoparticles are experiencing considerable interest due chemical composition.
to their catalytic potential which has long been used in hetero-
We have developed for nearly 20 years the synthesis of
geneous catalysis but which also appears to be promising in nanoparticles through an organometallic approach. In this
8
1
solution. It is expected that a complex but selective chemistry way, we can obtain small and well-controlled nanoparticles
could arise from the use of nanoparticles displaying a well con- whose surface is only covered by hydrides resulting from reac-
2
trolled chemical order and accommodating ligands, in par- tion conditions (decomposition of metal precursors under
3
ticular asymmetric ones.
Among others, bimetallic dihydrogen) and ligands that are used as stabilizers. The
4
nanoparticles are attractive materials due to the synergy choice of the ligand is crucial as it can influence the character-
expected to result from the presence of the two metals. For istics of the particles, not only their surface properties but also
5
large nanoparticles (>5 nm), the control of the chemical order their structure (size, shape, chemical composition). While the
within one particle is well documented but this is more role of the ligand was clearly identified for numerous mono-
6
difficult for small particles. Since this control is crucial for metallic systems, only a few examples concern bimetallic nano-
9
bimetallic nanoparticles, especially in catalysis, it is necessary particles stabilized by ligands.
Herein we describe the synthesis of diphenylphosphinobu-
tane-stabilized bimetallic RuPt nanoparticles (RuPt/dppb)
a
CNRS, LCC (Laboratoire de Chimie de Coordination), 205, Route de Narbonne,
F-31077 Toulouse, France
starting from [Ru(COD)(COT)] [(1,5-cyclooctadiene)(1,3,5-
cyclooctatriene)ruthenium] and [Pt(CH (COD)] [dimethyl(1,5-
3 2
)
b
Université de Toulouse, UPS, INPT, LCC, F-31077 Toulouse, France
cyclooctadiene) platinum(II)] organometallic precursors in a
metal ratio 1/1. These nanoparticles are synthesized by
decomposition of the metal complexes in THF solution and
c
CNRS, Centre d’Elaboration de Matériaux et d’Etudes Structurales (CEMES), 29, rue
Jeanne Marvig, BP 4347 31055 Toulouse Cedex, France
d
LMA, Laboratorio de Microscopias Avanzadas, Campus Río Ebro C/ Mariano
Esquillor s/n, 50018 Zaragoza, Spain
under mild conditions (3 bar H ; r.t.). This work follows two
previous studies from the group concerning RuPt nanosystems
prepared under similar reaction conditions (THF; 3 bar H ;
2
r.t.; Ru/Pt = 1/1). In the first one, the formation of a RuPt alloy
2
e
LPCNO, Laboratoire de Physique et Chimie de Nano-Objets, 135, Avenue de
Rangueil, F-31077 Toulouse, France. E-mail: chaudret@insa-toulouse.fr;
Fax: +33561559697; Tel: +33561559655
†
Dedicated to Prof. David Cole-Hamilton for his achievements in organometallic
chemistry and homogeneous catalysis and for over 35 years of friendship.
with an fcc structure was observed using [Ru(COD)(COT)] and
3
72 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
[
Pt(dba) ] as metal sources and the well-known polymer poly- FEI Tecnai F20 instrument operated at 200 kV and fitted with a
2
1
0
vinylpyrrolidone (PVP) as a stabilizer. In the second study, field emission gun and a CEOS spherical aberration corrector
core–shell RuPt nanoparticles were prepared in PVP, only by of the objective lens. The point resolution of this instrument is
changing the platinum precursor [Pt(dba)
2
]
for [Pt- 0.12 nm. HREM images were recorded on a 2k CCD camera at
1
1
(
CH ) (COD)], this latter being decomposed at a slower rate.
a sampling density of 0.017 nm per pixel. STEM/HAADF obser-
3
2
Thus, the chemical segregation to get core–shell RuPt resulted vations were carried out at the LMA laboratory in Zaragoza,
from kinetics (decomposition rate) and thermodynamic (pre- using a Cs corrected Titan Microscope equipped with an XFEG
ferred location in the particle) parameters of the two metal source.
precursors as well as from the sterical properties of the
Wide-angle X-ray scattering (WAXS) was performed at
polymer, knowing that PVP is inert towards the metals as little CEMES-CNRS. Samples were sealed in 1 mm diameter Linde-
or no chemical interaction takes place. In the present work, we mann glass capillaries. The samples were irradiated with
kept the same metal sources, [Ru(COD)(COT)] and [Pt- graphite-monochromatized molybdenum K
α
(0.071069) radi-
(
CH (COD)], but instead of a neutral sterical stabilizing ation and the X-ray intensity scattered measurements were per-
3 2
)
agent, we used a coordinating ligand, namely diphenylphos- formed using a dedicated two-axis diffractometer. Radial
phinobutane. Our objective was to study the influence of the distribution functions (RDF) were obtained after Fourier trans-
electronic properties of this ligand on the synthesis, size and formation of the reduced intensity functions.
chemical composition of the final nanoparticles. For this
EXAFS measurements were performed at HASYLAB in
purpose, a combination of characterization techniques (TEM, Hamburg on beamline C in transmission mode at room temp-
HREM, STEM-HAADF, WAXS, EXAFS, IR, MAS-NMR) and reac- erature. Both Pt LIII and Ru K absorption edges could be
tivity studies (CO adsorption and styrene hydrogenation) was measured for all samples. Data analysis was performed using
used, in order to obtain complementary structural and spectro- the Athena software. Least-squares refinement of structural
scopic approaches for elucidating the structure of the particles. parameters could also be performed using the Artemis
Our results show how a simple ligand such as a diphosphine software.
1
13
can add a significant degree of complexity to a system like
Solid state NMR (MAS-NMR) with and without H– C cross
RuPt nanoparticles. The ligand does not only decorate the par- polarization (CP) were performed at the LCC on a Bruker
ticles but affects also the chemical composition of the Avance 400WB instrument equipped with a 4 mm probe with
particles.
the sample rotation frequency being set at 12 kHz unless
otherwise indicated. Measurements were carried out in a
4
mm ZrO rotor.
2
Gas chromatography was performed using an HP 5890
Series II Gas Chromatograph with an SGE BP1 non-polar 100%
dimethyl polysiloxane capillary column (50 m × 0.32 mm ×
Experimental
General and materials
All operations were carried out using standard Schlenk tubes,
Fischer-Porter bottle techniques or in a glove-box under an
argon atmosphere.
0
.25 μm). The method used for the quantification of hydrides
−
1
consists of 15 min at 40 °C and a ramp of 8 °C min until
250 °C. The method used for styrene hydrogenation exper-
−
1
Solvents were purified just before their use as follows: THF
iments consists of 3 min at 80 °C and a ramp of 8 °C min
until 210 °C.
(Sigma-Aldrich) by distillation under argon over sodium–
benzophenone, pentane (SDS) through filtration on a column
in a purification system (Braun). [Ru(COD)(COT)] was pur-
FTIR spectra were recorded on a Perkin-Elmer GX2000 spec-
trometer in the range 4000–400 cm− . All the samples were
prepared as KBr pellets under an argon atmosphere.
ICP analyses were performed by Antellis Toulouse and
elemental analysis at the LCC.
1
chased from Nanomeps Toulouse, [Pt(CH ) (COD)] from Strem
3
2
1
3
13
2
Chemicals, CO and H from Air Liquid, CO ( C, 99.14%)
from Eurisotop, bis(diphenylphosphino)butane (dppb) from
Sigma Aldrich; all these reactives were used without
purification.
Synthesis of RuPt/dppb nanoparticles
Colloid 1: RuPt/dppb (8.3 mM) NPs. The organometallic
complexes [Ru(COD)(COT)] (189 mg, 0.601 mmol) and [Pt-
Characterization techniques
RuPt/dppb NPs were characterized by TEM, HREM and STEM- (CH ) (COD)] (200 mg, 0.601 mmol) were introduced in a
3
2
HAADF after deposition of a drop of the crude THF colloidal Fischer-Porter reactor and dissolved in 40 mL of THF pre-
solution on a covered holey copper grid. TEM analyses were viously degassed by three freeze–pump cycles. The resulting
performed at the “Service Commun de Microscopie Electroni- yellow solution was cooled at −60 °C and a THF solution
que de l’Université Paul Sabatier” (TEMSCAN-UPS) by using a (32 ml) containing dppb (127.92 mg, 0.300 mmol) was intro-
JEOL JEM 1011 CX-T electron microscope operating at 100 kV duced into the reactor. The Fischer-Porter bottle was pressur-
with a point resolution of 4.5 Å. The approximation of the par- ized with 3 bar of H and the solution was left to reach slowly
2
ticles mean size was made through a manual analysis of the room temperature. The homogeneous solution, which
enlarged micrographs by measuring a number of particles on turned black after 20 min of reaction, was kept under stirring
a given grid. HREM analyses were performed at CEMES on a overnight at room temperature. After this period of time,
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 372–382 | 373

View Article Online
Paper
Dalton Transactions
excess of H
2
was eliminated and the volume of solvent was solutions through an alumina column. Assuming a size of
reduced to 10 mL under vacuum. 40 mL of pentane were then 1.8 nm for Colloid 1 which corresponds to ca. 45 to 50% atoms
added to the colloidal suspension which was cooled down to on the surface, ca. 1.4 hydrides per surface atom were found.
−30 °C to precipitate the particles. After filtration under argon
with a cannula, the black solid powder was washed twice with Surface reactivity with CO
pentane (2 × 40 mL) and filtrated again before drying under
vacuum.
Elemental analysis: C, 17.2; H, 1.6%
Ru content (ICP): 25.9%
Pt content (ICP): 41%
Colloid 2: RuPt/dppb (2.5 mM) NPs. The organometallic
complexes [Ru(COD)(COT)] (142 mg, 0.451 mmol) and [Pt-
The coordination of carbon monoxide on the surface of RuPt/
dppb nanoparticles was performed in the solid state as
follows. A purified sample of nanoparticles was introduced in
a Fischer-Porter bottle. The reactor was pressurized with 3 bar
of H for at least 4 h to eliminate oxygen from the nanoparti-
2
cles’ surface. The H
2
gas was then evacuated under vacuum for
1
0
5 min, and the Fischer-Porter bottle further pressurized with
.2, 0.6 or 1 bar of CO for different reaction times. Then, the
(CH
3
)
2
(COD)] (150 mg, 0.451 mmol) were introduced into a
13
Fischer-Porter reactor and dissolved in 100 mL of THF pre- 13
viously degassed by three freeze–pump cycles. The resulting
yellow solution was cooled at −60 °C and a THF solution
CO gas was evacuated under vacuum for 20 min and solid
13
state C NMR and FTIR spectra (KBr pellets) were recorded.
(
80 ml) containing dppb (96.2 mg, 0.226 mmol) was intro-
duced into the reactor. The Fischer-Porter bottle was pressur-
ized with 3 bar of H and the solution was left to reach slowly
Surface reactivity with styrene
In a typical experiment, 0.03 mmol of total metal RuPt/dppb
nanoparticles were introduced into a Fisher-Porter reactor
under an argon atmosphere. Then, 5 ml of THF and 1 ml of
styrene, which were degassed before use, were added and the
2
the room temperature. The homogeneous solution, which
turned black after 20 min of reaction, was kept under stirring
overnight at room temperature. After this period of time,
excess of H2 was eliminated and the volume of solvent was
reduced to 10 mL under vacuum. 40 mL of pentane were then
added to the colloidal suspension which was cooled down to
2
reaction mixture was pressurized to 3 bar H . The reaction
mixture was maintained at 25 °C using a thermostated water
bath. Aliquots were taken from the reaction medium at
different reaction times and analyzed by GC to evaluate the
reactivity of the nanoparticles.
−
30 °C to precipitate the particles. After filtration under argon
with a cannula, the black solid powder was washed twice with
pentane (2 × 40 mL) and filtrated again before drying under
vacuum.
Elemental analysis: C, 25.8%; H, 3.1%
Ru content (ICP): 21.8%
Pt content (ICP): 34.8%
Results and discussion
RuPt nanoparticles were prepared by co-decomposition of the
organometallic
CH (COD)] in a 1 : 1 metal ratio. The reaction was performed
in THF at room temperature (r.t.) under H pressure (3 bar)
precursors
[Ru(COD)(COT)]
and
[Pt-
(
3 2
)
Quantification of hydrides
2
A solution of freshly prepared RuPt/dppb nanoparticles was and in the presence of 0.25 equiv. of a diphosphine as the sta-
submitted to 5 cycles of 1 min vacuum/1 min argon in order to bilizing ligand, namely bisdiphenylphosphinobutane (dppb)
eliminate any hydrogen dissolved in the THF. Then, 2-norbor- (Scheme 1). Two different metal concentrations were used as
nene was added (5 equiv.) and its conversion into 2-norbor- this parameter can modify the size and homogeneity of the
nane was monitored by GC from previously filtered reaction particles.
Scheme 1 Synthesis of RuPt/dppb NPs.
3
74 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 1 TEM images with the corresponding size histograms of Colloid 1 (a) and Colloid 2 (b).
The first RuPt/dppb sample (Colloid 1) was prepared by nanoparticles is not affected by metal concentration. ICP
using a concentration of 8.3 mM of each metal (Ru and Pt) in analysis revealed the presence of both metals with 21.8% wt
THF, and in the presence of 0.25 molar equiv. of dppb com- and 34.8% wt for Ru and Pt respectively. These values corres-
pared to each metal. After 10 min under 3 bar H2 at room pond to a relative metal ratio of 1.2 : 1, in good agreement with
temperature, the color of the reaction mixture changed from the quantity of metals introduced for the synthesis of the
intense yellow to dark brown. The reaction mixture was kept particles.
stirring for 14 h to ensure the total decomposition of the pre-
With the aim of fully understanding the crystalline struc-
cursors. After this period of time, the reaction was stopped, ture and the chemical order in these RuPt/dppb nanoparticles,
the dihydrogen pressure was eliminated and TEM and HREM Colloids 1 and 2 were characterized by HREM and WAXS
grids were prepared for analysis. Then, the nanoparticles were (wide-angle X-ray scattering). In addition, Colloid 1 was ana-
purified by precipitation with cold pentane, giving rise to a lyzed by STEM-HAADF (scanning transmission electron
black powder that was used for solid-state characterization and microscopy using a high angle annular dark field detector)
reactivity studies.
and Colloid 2 by EXAFS (extended X-ray absorption fine struc-
TEM analyses performed on Colloid 1, showed the for- ture). Finally, surface reactivity studies towards CO and styrene
mation of nanoparticles displaying a mean size of ca. 1.8(±0.2) were performed with Colloid 1.
nm. They are homogeneous in shape and size with a narrow
HREM images of both colloids show the presence of par-
size distribution (Fig. 1a). ICP analysis revealed the presence of ticles displaying hexagonal-close packed (hcp) structure. As the
both metals with 25.9% wt and 41.0% wt for Ru and Pt pixel size in the HREM images is 0.017 nm, it is not possible
respectively. These values correspond to a relative metal ratio to precisely determine the distance between the atomic planes.
of 1.2 : 1, in agreement with the quantity of metals introduced Typically (111) planes in Pt are spaced by 0.2265 nm and
for the synthesis of the particles.
(10 ˉ1 0) planes in Ru are spaced by 0.2342 nm. The measured
The second RuPt/dppb sample (Colloid 2) was prepared in a distances, around 0.23–0.24 nm, are consistent with both para-
similar way but under more diluted conditions introducing meters. For this reason, we rather analyzed the displayed pat-
2
.5 mM of each metal, Ru and Pt. Under these reaction con- terns since hcp and fcc patterns strongly differ. Fig. 2 presents
ditions, a homogeneous population of NPs having a mean size a general view of Colloid 1 and two enlarged HREM images of
of ca. 1.8(±0.2) nm was also obtained as evidenced by TEM particles displaying respectively a neat hcp pattern and a
(Fig. 1b). This result indicates that the size of the RuPt heavily faulted fcc pattern (referred to as fcc-like). In order to
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 372–382 | 375

View Article Online
Paper
Dalton Transactions
enhance the contrast of these particles, which suffered from
an important background due to the carbon film, some
Fourier filtering was performed on the presented images.
An HREM image of Colloid 2 is presented in Fig. 3. The par-
ticle indicated by a circle appears in the inset together with the
Fourier transform of the image. Clearly this particle displays
an hcp pattern fully consistent with a Ru structure. However,
some disorder is regularly observed far from the particle core.
Such disorder may be due either to pure surface effects
(
ligands, incomplete shell etc.) or to compositional
inhomogeneity.
STEM-HAADF was used to investigate the presence of a
local chemical order. In STEM-HAADF the image contrast is
approximately proportional to the number of atoms in a
2
column with each atom contributing proportionally to Z . This
is the reason why this technique is usually referred to as Z-con-
trast imaging. It has already been shown that RuPt/PVP nano-
particles can be easily identified with this technique as
Fig. 2 HREM images of Colloid 1 (general view). The HREM image on the left
evidences the hcp pattern of a particle observed along a [0001] zone axis. The
HREM image on the right is closer to an fcc pattern observed along a [110]
zone axis. Note that this particle presents numerous stacking faults.
1
1
displaying a core–shell structure.
In the present case,
however, the particles appear neither as hollow nor as brighter
in the middle (Fig. 4). In the first case (hollow contrast) this
would be an indication of the presence of a Ru core. In the
second case (brighter core) two possibilities should be
considered:
1
) The core is brighter since the heavier element (Pt) is con-
centrated in the middle.
) The composition is uniform and the contrast variation is
2
only arising from the particle shape (which is thicker in the
middle). The inhomogeneous contrast variation observed by
Z-contrast clearly shows that no compositional homogeneity
can be observed and that the particles present, locally at least,
some kind of chemical disorder.
Because of the small size of the particles, X-ray diffraction
measurements in the reciprocal space only provide broad
peaks which do not allow a clear indexation on one of the two
patterns expected for these nanoparticles, i.e. hcp like for Ru
rich alloys or fcc like for Pt rich ones. However, observation in
the real space after Fourier transform of the WAXS spectrum
provides more information (Fig. 5). The coherence length,
which is a direct measurement of the size of crystalline
Fig. 3 HREM image of Colloid 2. The circled particle, presented in the inset, is
observed along a [0001] zone axis of an hcp structure. The FFT of this particle is domains, however sensitive to disorder, is close to 1.9 nm for
also presented.
both colloids. This value is in good agreement with the mean
Fig. 4 STEM-HAADF images of Colloid 1.
3
76 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
model including 6 Ru atoms at 0.261 nm and 6 atoms at
0.266 nm, however taking into account an amplitude reduction
factor of 0.4 likely related to both size and disorder effects.
Since data at the Pt edge clearly indicate the presence of Ru–Pt
interactions, the fact that such contributions are not clearly
observed at the Ru edge is likely caused by a very different
static disorder between the Ru–Ru and Ru–Pt distances. EXAFS
is very sensitive to such effects. This difference actually points
to a model where a large part of the Ru atoms are assembled
in a well ordered Ru core, and the rest in a much more disor-
dered Pt–Ru alloy.
Surface reactivity studies were carried out to probe the
surface state of the RuPt/dppb nanoparticles. First of all, due
to the reaction conditions used for the synthesis of the par-
ticles, i.e. dihydrogen pressure, the presence and amount of
hydrides present at the surface of the particles was deter-
mined. This was performed by applying a simple method pre-
viously described in the group for monometallic Ru NPs
stabilized by a polymer (PVP) or by ligands (amine or dipho-
sphine) and bimetallic RuPt NPs stabilized by PVP. This
method consists in reacting freshly prepared particles with a
simple olefin, namely 2-norbornene, without adding extra
dihydrogen and to follow the hydrogenation of this olefin into
Fig. 5 WAXS analysis and simulations of bimetallic RuPt/dppb nanoparticles
2-norbornane by gas chromatography (GC). Thus, the only
(
from top to bottom: fcc simulation, Colloid 1, Colloid 2, hcp simulation).
source of hydrogen atoms is the ones present on the surface of
the particles. The number of surface hydrides can be calcu-
size determined by TEM/HREM analyses (ca. 1.8 nm). Indexa- lated from the olefin conversion values and the mean size of
tion of the radial distribution function (RDF) is ambiguous: it the particles. This method was applied for Colloid 1, giving
cannot be considered clearly either as fcc or as hcp. By com- rise to 1.4 hydrides per total surface metal atom. This result
parison with perfect patterns, close observation of the has to be compared on the one hand with that obtained for
1
1
0
.6–1.1 nm range, very sensitive to structural organization, core–shell RuPt NPs stabilized in PVP (1.7 H/(Ru + Pt)s), and
indicates that both colloids are closer to the hcp system. on the other hand to that measured for Ru/PVP and Ru/dppb
1
2
Because of this deviation to perfect structures, it is not possi- nanoparticles (1.3 H/Ru and 1.1 H/Ru respectively). Thus,
s
s
ble to make adjustments on a model and the average bond and as previously observed for monometallic RuNPs, a slightly
length can only be estimated from the location of the first lower number is obtained for dppb-stabilized RuPt NPs in
peak. Exactly the same value is observed for both colloids, comparison with PVP-stabilized ones (1.4 against 1.7 H/
0
.272 nm, nearly perfectly intermediate between values (Ru + Pt)s).
1
3
reported for Ru–Ru (0.265 nm) and Pt–Pt (0.277 nm).
In a second step, the adsorption of C labeled carbon mon-
EXAFS measurements performed on Colloid 2 at both Pt oxide at the surface of the RuPt/dppb nanoparticles was inves-
LIII and Ru K absorption edges reveal strong differences in tigated by infrared (FTIR) and NMR (MAS NMR with or
local environment: compared to the Pt foil measured under without CP) spectroscopies in the solid state. This reaction has
the same conditions for reference, the modulus of the Fourier been performed since CO is widely known as an infrared and
transform (Fig. 6, left) presents a peak splitting caused by a NMR sensitive probe molecule for identifying coordination
destructive interference which is related in this case to a mixed sites on surfaces and nanoparticles.11,13–16 Thus, we have
environment of Pt atoms. This effect is caused by differences shown by IR and NMR investigations that CO molecules are
in bonding distances, but mostly in the phase shift of the dynamic on the surface of ruthenium nanoparticles only steri-
backscattered wave which is very different for Pt or Ru neigh- cally stabilized by PVP, whereas the presence of extra ligands
bors. It is indeed a precious tool to evidence, and model, such prevents exchange between bridging and terminal CO
1
6
mixed environments. In our case, the phenomenon is however groups. In our previous attempts concerning RuPt nanoparti-
poorly defined and it was not possible to obtain a stable and cles, among other techniques, IR and NMR techniques were
accurate fit of structural parameters. To obtain this pattern, it also very useful, helping us to identify for two RuPt/PVP
was necessary to define an average local environment of Pt systems prepared under different conditions: (1) the formation
atoms including a majority of Pt atoms at 0.272 nm and a min- of alloy nanoparticles of defined compositions resulting from
ority of Ru atoms at 0.268 nm. Quite strikingly, the same func- the relative concentration of both metal precursors in the
1
0
tion at the Ru K edge is nearly identical to the one obtained initial solution or (2) a core–shell structure resulting from
for bulk Ru and good agreement could be obtained with a the difference in the kinetics of decomposition of the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 372–382 | 377

View Article Online
Paper
Dalton Transactions
Fig. 6 Experimental (continuous line) and simulated (dashed line) EXAFS functions (bottom) and related Fourier transforms (top) at the Pt LIII (left) and Ru K (right)
absorption edge of monometallic NPs and Colloid 2 (from top to bottom in a single frame).
1
1
13
The presence of a 13CO absorption band at ca. 1947 cm−1 is
precursors. To study the adsorption of CO on the surface of
the particles, various experiments have been performed in a detected after a short reaction time at low carbon monoxide
Fischer-Porter reactor by using purified solid RuPt/dppb nano- pressure (0.2 bar; 10 min; green). The presence of this band is
particles and changing the gas pressure and the reaction time confirmed at higher pressure and reaction time (0.6 bar; 2 h;
1
3
(
0.2–1 bar of CO; 10 min; 3 days). The IR spectra registered turquoise), with a slight shift towards higher frequencies (ca.
1
3
−1
for Colloid 1 after exposure under a CO atmosphere are pre- 1954 cm ). These results indicate that the adsorption of CO is
sented in Fig. 7 together with the spectrum obtained for puri- efficient on the surface of the particles and that it evolves
fied nanoparticles as a reference. The purified Colloid 1 shows under air, but the coordination mode cannot be defined. Com-
−
1
absorption bands in the range 3070–2850 cm
and ca. paring these results with those of the core–shell RuPt system
884 cm− corresponding to C–H and CvC absorption bands for which linear CO at ca. 1995 cm was mainly detected and
1
−1
1
of the ligand molecules, respectively. This evidences the pres- attributed to CO coordinated on Pt atoms, it appears that the
ence of the dppb ligand at the surface of the particles after the surface of the particles does not contain only Pt atoms.
purification treatment. The coordination of the dppb ligand to However, a comparison with pure Ru/dppb NPs, for which the
3
1
1
13
the surface of the particles was also confirmed by P{ H} MAS reaction with CO (2 bar; 14 h) gave rise to a CO band at
NMR, with observation of a broad signal at 4.78 ppm and 1950 cm− , is in agreement with the presence of Ru atoms at
sharper signals at 37.63, 31.71 and 25.57 ppm (all of them the surface of the RuPt nanoparticles.
1
1
3
have spinning side bands) that could correspond to dppb and
The RuPt/dppb/ CO nanoparticles were also characterized
partially hydrogenated dppb (due to the synthesis conditions) by MAS-NMR, and compared to purified RuPt/dppb nanoparti-
1
3
coordinated to the metal surface or to dppb in different chemi- cles. The C-MAS NMR spectrum of the purified nanoparticles
cal environments (coordinated to Ru or Pt) (see Fig. 8).
shows signals in the range 10–40 ppm and ca. 110 and
3
78 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 7 FTIR spectra registered for Colloid 1.
31
1
Fig. 8
P{ H} MAS NMR of Colloid 1.
1
35 ppm attributed to the alkyl and phenyl carbons of the respectively at 205.17 and 199.75 ppm which could result from
dppb ligands respectively, evidencing the coordination of the the presence of terminal CO groups on two different metals or,
dppb ligand at the surface of the particles (Fig. 9a). When the more likely, of respectively multicarbonyl and terminal CO
1
3
particles were reacted with 0.2 bar of CO during 10 min, we groups. This small modification with time can be related to
could observe the growth of two sets of signals near 200 ppm the slight displacement observed for the CO frequency in IR
and 230 ppm corresponding respectively to terminal and brid- respectively to 1954 cm− after 2 h and 1959 cm after 3 days.
1
−1
ging CO ligands (Fig. 9b). The presence of spinning side bands
These IR and NMR experiments after CO adsorption clearly
demonstrates the absence of mobility of CO on these nano- do not correspond to a reaction with monometallic nanoparti-
particles. It is interesting to compare these results with the cles and favor a bimetallic structure with a complex surface
adsorption of CO on Ru/dppb nanoparticles. Thus in the case where some Ru is present.
of monometallic nanoparticles, we observe first a bridging
In a last step, we studied the reactivity of the RuPt/dppb
coordination mode for CO and then both. In the present case, particles (Colloid 1) towards the hydrogenation of styrene
terminal CO groups are first observed which is in agreement under dihydrogen pressure. This reaction is typical of ruthe-
with the presence of two different metals at the surface.
nium nanoparticles in which case it is easily poisoned by
1
3
17
When the nanoparticles were reacted with 0.6 bar of CO CO. It could therefore serve as a further characterization of
during 2 h (Fig. 9c), or 1 bar during 3 days (Fig. 9d), a more the surface. In a typical experiment, 0.03 mmol total metal
intense signal grows near 200 ppm. It consists of two signals RuPt/dppb NPs were introduced into a Fisher-Porter bottle
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 372–382 | 379

View Article Online
Paper
Dalton Transactions
Fig. 11 Evolution of the styrene hydrogenation reaction with time using
1
3
0
.03 mmol total metal RuPt/dppb/ CO NPs exposed to 0.2 bar of CO for
0 min at 25 °C under 3 bar of H
1
2
.
1
3
C-MAS NMR of RuPt/dppb NPs (Colloid 1): from bottom to top, (a) 2.1 h−1) than with pure Ru/dppb NPs (TOF = 1.5 h−1). The
18
Fig. 9
13
1
3
1
3
before CO adsorption, (b) after 0.2 bar CO, 10 min, (c) after 0.6 bar CO,
observed increase in the catalytic activity is attributed to a
synergic effect of Ru–Pt atoms on the surface of the nanoparti-
cles. Such a synergic effect of Ru–Pt nanoparticles was pre-
viously reported in the literature inter alia for the
hydrogenation of aromatic hydrocarbons, toluene hydrogena-
tion and selective hydrogenation of phenylacetylene to sty-
rene. Consequently, these results are also in agreement with
the presence of both metals at the surface of the RuPt
nanoparticles.
13
2
h, (d) after 1 bar CO, 3 days.
5
c
5
b
5
a
The effect of CO adsorption on the catalytic properties of
RuPt/dppb NPs in the styrene hydrogenation reaction can be
very interesting since it is expected that Pt and Ru atoms will
behave differently (Fig. 11). When using RuPt/dppb NPs with a
surface partially covered by CO (0.2 bar, 10 min), we observed
that the reaction rate decreases drastically. As in previous
work, we showed that replacing CO by H
RuNPs is difficult at room temperature,
the reaction results from a poisoning of the surface active sites
due to the presence of CO. However, despite the surface poi-
soning by CO, full styrene conversion to ethylcyclohexane
2
on the surface of
the slower rate of
Fig. 10 Evolution of the styrene hydrogenation reaction with time using
.03 mmol total metal RuPt/dppb NPs at 25 °C under 3 bar of H (styrene/
RuPtsurface = 582).
16,17
0
2
under an argon atmosphere with 5 ml of THF and 1 ml of could still be achieved which is not the case under similar con-
1
7
styrene (ratio styrene/M = 291). Then, the reaction mixture was ditions with pure Ru/dppb nanoparticles. This result indi-
pressurized to 3 bar H and stirred at 25 °C using a tempera- cates therefore that the surface of the RuPt nanoparticles is
2
ture controlled water bath. Aliquots were taken at different less deactivated by CO than that of monometallic Ru nano-
reaction times to estimate the reactivity of the particles particles and is again in strong agreement with the presence of
1
9
through GC analysis.
a bimetallic surface.
In summary, the data resulting from the styrene hydrogen-
The hydrogenation graph of styrene with RuPt/dppb NPs is
shown in Fig. 10 in which two regimes of consumption of ation experiments confirm the presence of both Ru and Pt
hydrogen can be observed. In the first one, the vinyl group of metals at the surface of the RuPt nanoparticles. Even if a
styrene is reduced forming ethylbenzene. Then, in a second decrease in the reaction rate is observed due to a blocking of
step, ethylbenzene is converted into the complete saturated the active sites of RuPt/dppb NPs by CO adsorption, the nano-
form of styrene that is ethylcyclohexane. These two regimes particles remain active for the total hydrogenation of styrene.
diverge sufficiently to allow the isolation of the first hydrogen-
ation product, ethylbenzene, with 93% selectivity.
When the catalytic activity of RuPt/dppb NPs was compared
Conclusion
with the performance of Ru/dppb NPs in styrene hydrogen-
1
7
ation, we observed that the total conversion of styrene to In this paper, we have attempted the full characterization of
1
8
ethylcyclohexane is faster with RuPt/dppb NPs (TOF
=
bimetallic RuPt nanoparticles stabilized by a diphosphine
3
80 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
ligand. This work follows previous studies concerning other
RuPtNPs of similar size and composition, resulting from the
same organometallic method and using the same or similar
precursors but stabilized by a polymer (PVP) instead of a
ligand. The surprise lies in the strong difference induced by
the presence of the diphosphine ligand. Thus, instead of an
ordered alloy or of a core–shell structure when using the same
precursors, we obtain here a very disordered structure where a
core of ruthenium is evidenced by HREM and EXAFS and a
shell of Pt–Ru by WAXS and spectroscopic methods as well as
by its catalytic activity. The origin of this structure could be the
reaction of the diphosphine with [Ru(COD)(COT)] at the early
stage of the reaction producing a bis(diphosphine) complex as
has been observed 30 years ago with bis(diphenylphosphino)-
References
1 G. A. Somorjai, H. Frei and J. Y. Park, J. Am. Chem. Soc.,
2009, 131, 16589–16605.
2 F. Humblot, D. Didillon, F. Lepeltier, J. P. Candy, J. Corker,
O. Clause, F. Bayard and J. M. Basset, J. Am. Chem. Soc.,
1998, 120, 137–146.
3 (a) H. U. Blaser and M. Studer, Acc. Chem. Res., 2007, 40,
1348–1356; (b) T. Mallat, E. Orglmeister and A. Baiker,
Chem. Rev., 2007, 107, 4863–4890.
4 (a) T. A. Yamamoto, S. Kageyama, S. Seino, H. Nitami,
T. Nakagama, R. Horioka, Y. Honda, K. Ueno and
H. Daimon, Appl. Catal. Gen., 2011, 396, 68–75;
(b) C.-H. Chen, L. S. Sarma, D. Y. Wang, F.-J. Lai, C.-C.
A. Andra, S.-H. Chang, D. G. Liu, C.-C. Chen, J.-F. Lee and
B.-J. Hwang, ChemCatChem., 2010, 2, 159–166;
(c) T.-Y. Chen, T.-L- Lin, T.-J. M. Luo, Y. Choi and J.-F. Lee,
ChemPhysChem., 2010, 11, 2383–2392; (d) A. Murthy and
A. Manthiram, Electrochem. Commun., 2011, 13, 310–313;
(e) Y.-C. Wei, C.-W. Liu, W.-J. Chang and K.-W. Wang,
J. Alloys Compd., 2011, 509, 535–541; (f) M. Liu, J. Zhang,
J. Liu and W. W. Yu, J. Catal., 2011, 278, 1–7; (g) T.-Y. Jean,
K.-S. Lee, S.-J. Yoo, Y.-H. Cho, S.-H. Kang and Y.-E. Sung,
Langmuir, 2010, 26, 9123–9129.
5 (a) C. Li, Z. Shao, M. Pang, C. T. Williams, X. Zhang and
C. Liang, Ind. Eng. Chem. Res., 2012, 51, 4934–4941;
(b) D. A. Dehm, X. Zhang and J. M. Buriak, Inorg. Chem.,
2010, 49, 2706–2714; (c) T. Arakawa, H. Seki, M. Ohshima,
H. Kurokawa and H. Miura, Bull. Chem. Soc. Jpn., 2009, 82,
627–629.
2
0
methane (dppm). This complex would be very difficult to
decompose in comparison with [Ru(COD)(COT)] and [Pt-
(
CH ) (COD)]. Therefore, in the reaction mixture, we would
3 2
have first decomposition of unreacted [Ru(COD)(COT)],
decomposition of [Pt(CH (COD)] catalyzed by the presence of
3 2
)
the Ru nuclei and finally decomposition of the diphosphine
ruthenium complex. This would lead for these small sizes to a
very disordered onion structure, compatible with all obser-
vations. The main interest of this study is therefore the demon-
stration of the power of molecular chemistry to engineer the
synthesis of complex nano-objects, here the complexity lying
in the chemical order of the nanoparticles. Another interest is
the demonstration that these nanoparticles can now hydrogen-
ate arenes even after CO poisoning, thus pointing to the inter-
est of preparing bimetallic species.
In conclusion, we report in this paper a way of modulating
the chemical order in nanoparticles using ligands able to react
with the molecular precursors. Given the variety of ligands and
molecular complexes available, this method opens the door to
a fine tailoring of the structure and surface of complex nano-
objects.
6 (a) D. Ciuculescu, C. Amiens, M. Respaud, A. Falqui,
P. Lecante, R. E. Benfield, L. Jiang, K. Fauth and B. Chaudret,
Chem. Mater., 2007, 19, 4624–4626; (b) N. Atamena,
D. Ciuculescu, G. Alcaraz, A. Smekhova, F. Wilhelm,
A. Rogalev, B. Chaudret, P. Lecante, R. E. Benfield and
C. Amiens, Chem. Commun., 2010, 2453–2455.
7
S. Sun, C. B. Murray, D. Weller, L. Folks and A. Moser,
Science, 2000, 287, 1989–1192.
8
Organometallic Derived-I: Metals, Colloids and Nanoparti-
cles, Philippot K and Chaudret B, in Comprehensive Orga-
nometallic Chemistry III, Crabtree RH & Mingos MP (Eds-
in-Chief), Elsevier, Volume 12—Applications III: Functional
Materials, Environmental and Biological Applications,
Dermot O’Hare (Volume Ed.), 2007, Chapter 12–03, 71–99.
Acknowledgements
The authors thank V. Collière and L. Datas at UPS-TEMSCAN
and Y. Coppel at LCC for electron microscopy and NMR facili-
ties respectively. Financial support from the French national
METSA network of CNRS and CEA for HREM experiments at
CEMES is acknowledged. The EXAFS experiments were carried
out at the light source DORIS III at DESY, a member of the
Helmholtz Association (HGF). The research leading to these
results has received funding from the European Community’s
Seventh Framework Programme (FP7/2007–2013) under grant
agreement n° 226716. CNRS, ANR (Siderus project ANR-08-
BLAN-0010-03), INTERREG SUDOE (TRAIN2 project), the Euro-
pean Union for ERC Advanced Grant (NANOSONWINGS 2009-
9 (a) C. Desvaux, F. Dumestre, C. Amiens, M. Respaud,
P. Lecante, E. Snoeck, P. Fejes, P. Renaud and B. Chaudret,
J. Mater.Chem., 2009, 19, 3268–3275; (b) O. Margeat,
D. Ciuculescu, P. Lecante, M. Respaud, C. Amiens and
B. Chaudret, Small, 2007, 3, 451–458; (c) M. C. Fromen,
P. Lecante, M.-J. Casanove, P. B. Guillemaud, D. Zitoun,
C. Amiens, B. Chaudret, M. Respaud and R. E. Benfield,
Physical Review B: Condens. Matter, 2004, 69, 235416–
235425.
2
2
46763) and FP7-NMP2-Large program grant (Synflow 2010-
46461) are also thanked for financial support. P. Lara Muñoz 10 C. Pan, F. Dassenoy, M.-J. Casanove, K. Philippot,
is grateful to the Spanish Ministerio de Educación for a
research contract.
C. Amiens, P. Lecante, A. Mosset and B. Chaudret, J. Phys.
Chem. B, 1999, 103, 10098–10101.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 372–382 | 381

View Article Online
Paper
Dalton Transactions
1
1
1 P. Lara, M.-J. Casanove, P. Lecante, P.-F. Fazzini, K. Philippot 16 F. Novio, K. Philippot and B. Chaudret, Catal. Lett., 2010,
and B. Chaudret, J. Mater. Chem., 2012, 22, 3578–3584. 140, 1–7.
2 J. García-Antón, M. R. Axet, S. Jansat, K. Philippot, 17 F. Novio, D. Monahan, Y. Coppel, G. Antorrena,
B. Chaudret, T. Pery, G. Buntkowsky and H. H. Limbach,
K. Philippot and B. Chaudret, J. Am. Chem. Soc., unpub-
Angew. Chem., Int. Ed., 2008, 47, 2074–2078.
lished results.
13 S. E. Shore, J. P. Ansermet, C. P. Slichter and J. H. Sinfelt, 18 TOF = [(moles of ethylcyclohexane generated)/(moles of
Phys. Rev. Lett., 1987, 58, 953–956. RuPtsurface atoms × time)].
14 J. P. Ansermet, C. P. Slichter and J. H. Sinfelt, J. Chem. 19 B. Hammer, O. H. Nielsen and J. K. Nørskov, Catal. Lett.,
Phys., 1988, 88, 5963–5971. 1997, 46, 31–35.
15 L. R. Becerra, C. A. Klug, C. P. Slichter and J. H. Sunfelt, 20 B. Chaudret, G. Commenges and R. Poilblanc, J. Chem.
J. Phys. Chem., 1993, 97, 12014–12019.
Soc., Dalton Trans., 1984, 1635–1639.
3
82 | Dalton Trans., 2013, 42, 372–382
This journal is © The Royal Society of Chemistry 2013