Efficient ruthenium nanocatalysts in liquid-liquid biphasic hydrogenation catalysis: Towards a supramolecular control through a sulfonated diphosphine-cyclodextrin smart combination

Article abstract of DOI:10.1002/cctc.201300467

The combination between a sulfonated diphosphine (L) and a cyclodextrin (CD) allowed the preparation of very stable water-soluble ruthenium nanoparticles (RuNPs) that displayed pertinent catalytic performances in hydrogenation of unsaturated substrates with a supramolecular control effect of the cyclodextrin. For comparison purpose, the RuNPs were produced by hydrogenation of the organometallic [Ru(1,5-cyclooctadiene)(1,3,5- cyclooctatriene)] complex under mild conditions (3 bar H₂; room temperature) and in the presence of L or a L/CD mixture as stabilizer leading to Ru/L and Ru/L/CD systems, respectively. The so-obtained nanoparticles were fully characterized by complementary techniques. Interestingly, NMR investigations evidenced 1) the strong coordination of the sulfonated diphosphine ligand at the metallic surface and 2) in the presence of cyclodextrin, the formation of an inclusion complex between L and CD that modified the coordination mode of the diphosphine. The investigation of both RuNPs systems in biphasic hydrogenation of unsaturated substrates pointed out relevant differences in terms of reactivity, thus evidencing the influence of the supramolecular interaction at the metallic surface on the catalytic performances of the nanocatalysts. This work took advantage of the supramolecular properties of a cyclodextrin to modulate the surface reactivity of diphosphine-stabilized ruthenium nanoparticles and may open new opportunities in the field of nanocatalysis. Getting smarter every day: The use of a sulfonated diphosphine with randomly methylated-β-cyclodextrin leads to very stable water-soluble ruthenium nanoparticles that display pertinent catalytic performances in hydrogenation of unsaturated substrates with a supramolecular control effect of the cyclodextrin. Copyright

Full text of DOI:10.1002/cctc.201300467

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201300467
Efficient Ruthenium Nanocatalysts in Liquid–Liquid
Biphasic Hydrogenation Catalysis: Towards
a Supramolecular Control through a Sulfonated
Diphosphine–Cyclodextrin Smart Combination
[a, b]
[a, b]
[c, d]
[c, d]
Miguel Guerrero,
Audrey Denicourt-Nowicki,
Yannick Coppel,
Nguyet Tran Thanh Chau,
Alain Roucoux,
[c, d]
[e]
[e]
[f]
Eric Monflier, Hervꢀ Bricout, Pierre Lecante, and
[a, b]
Karine Philippot*
The combination between a sulfonated diphosphine (L) and
a cyclodextrin (CD) allowed the preparation of very stable
water-soluble ruthenium nanoparticles (RuNPs) that displayed
pertinent catalytic performances in hydrogenation of unsatu-
rated substrates with a supramolecular control effect of the cy-
clodextrin. For comparison purpose, the RuNPs were produced
by hydrogenation of the organometallic [Ru(1,5-cycloocta-
diene)(1,3,5-cyclooctatriene)] complex under mild conditions
nated diphosphine ligand at the metallic surface and 2) in the
presence of cyclodextrin, the formation of an inclusion com-
plex between L and CD that modified the coordination mode
of the diphosphine. The investigation of both RuNPs systems
in biphasic hydrogenation of unsaturated substrates pointed
out relevant differences in terms of reactivity, thus evidencing
the influence of the supramolecular interaction at the metallic
surface on the catalytic performances of the nanocatalysts.
This work took advantage of the supramolecular properties of
a cyclodextrin to modulate the surface reactivity of diphos-
phine-stabilized ruthenium nanoparticles and may open new
opportunities in the field of nanocatalysis.
(
3 bar H ; room temperature) and in the presence of L or a L/
2
CD mixture as stabilizer leading to Ru/L and Ru/L/CD systems,
respectively. The so-obtained nanoparticles were fully charac-
terized by complementary techniques. Interestingly, NMR in-
vestigations evidenced 1) the strong coordination of the sulfo-
Introduction
The nanoparticles and nanotechnology field is a fast-growing
research area that has already led to significant breakthroughs
with a wide variety of potential applications in biomedical, op-
[1]
tical and electronic areas as well as in catalysis. During the
past two decades, “nanocatalysis” has emerged as a modern
area at the frontier between homogeneous and heterogene-
[a] Dr. M. Guerrero, Dr. Y. Coppel, Dr. K. Philippot
CNRS
[2]
ous catalysis. This emergence results from the development
LCC (Laboratoire de Chimie de Coordination)
of efficient synthesis methods based on the use of tools from
molecular chemistry allowing to form better defined and also
2
05, Route de Narbonne, F-31077 Toulouse (France)
Fax: (+33)561553003
E-mail: karine.philippot@lcc-toulouse.fr
[3]
better characterized metal nanoparticles as catalysts.
[
b] Dr. M. Guerrero, Dr. Y. Coppel, Dr. K. Philippot
Among numerous investigations of metal nanoparticles in
catalysis, some of them are directly inspired from organometal-
lic chemistry, with the use of metal–organic precursors as the
source of metal atoms as well as of basic ligands (e.g., amines,
phosphorous derivatives) to protect the metal surface and con-
Universitꢀ de Toulouse
UPS, INPT; LCC
F-31077 Toulouse (France)
[
c] Dr. N. T. T. Chau, Prof. A. Roucoux, Dr. A. Denicourt-Nowicki
Ecole Normale Supꢀrieure de Chimie de Rennes
CNRS, UMR6226
[4]
trol the growth of the particles. Besides its role to obtain
stable nanoparticles of controlled size, the ligand can also be
chosen to transfer its physical-chemical properties to the nano-
particles systems, such as solubility or selectivity. In that con-
text, our group and others have used original ligands able to
tune the properties of metal nanoparticles of interest in organ-
Avenue du Gꢀnꢀral Leclerc, CS 50837, F-35708 Rennes cedex 7 (France)
[
d] Dr. N. T. T. Chau, Prof. A. Roucoux, Dr. A. Denicourt-Nowicki
Universitꢀ europꢀenne de Bretagne
[
e] Prof. E. Monflier, Dr. H. Bricout
UMR CNRS 8181
Universitꢀ d’Artois, Facultꢀ des Sciences Jean Perrin
Rue Jean Souvraz, SP 18 - F-62307 Lens Cedex (France)
[5]
ic catalysis. More recently, we have developed metal nano-
particles soluble in water (despite the sensitivity towards mois-
ture of the organometallic complexes used as metal sources)
by an adequate choice of water-soluble ligands such as 1,3,5-
[f] Dr. P. Lecante
CNRS
Centre d’Elaboration de Matꢀriaux et d’Etudes Structurales (CEMES)
2
9, rue Jeanne Marvig, BP 4347, 31055 Toulouse Cedex (France)
[6]
triaza-7-phosphaadamantane and alkyl sulfonated diphos-
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300467.
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3802

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Results and Discussion
[
7]
phines that were active in hydrogenation of unsaturated sub-
strates, in aqueous/organic biphasic catalysis conditions.
In the present contribution, the main objective is to increase
and modulate the catalytic performances of ruthenium nano-
catalysts by addition of a suitable molecular receptor. For this
purpose we used a cyclodextrin as coadditive in the synthesis
of sulfonated diphosphine-stabilized metal nanoparticles. This
The ruthenium nanoparticles (RuNPs) were synthesized by the
direct hydrogenation of the metal organic precursor [Ru(cod)-
(cot)] (cod=1,5-cyclooctadiene; cot=1,3,5-cyclooctatriene) in
THF under mild conditions (3 bar H ; room temperature) in the
2
presence of 1,4-bis[(di-m-sulfonatophenyl)phosphino]butane
(L) as sulfonated diphosphine ligand (L) or a combination of
this ligand with RAME-b-CD (Scheme 1).
[8]
idea derived from recent works in organometallic catalysis
[
9]
and nanocatalysis, in which cyclodextrins (CDs) have been
successfully used as mass-transfer promoters to improve a cata-
lytic reaction in aqueous/organic biphasic conditions. Spectro-
scopic studies have demonstrated that b-CD and randomly
methylated (RAME)-b-CD can interact with sulfonated phos-
phines by forming inclusion complexes, thus tuning the cata-
These RuNPs have been characterized in solution by liquid
NMR and dynamic light scattering (DLS), after isolation of the
particles and dispersion in water. Grids were prepared from
THF and aqueous solutions for TEM and high-resolution (HR)
TEM analysis. The purified nanoparticles as powders were also
characterized by solid-state NMR spectroscopy, elemental anal-
ysis, and wide-angle X-ray scattering (WAXS). The obtained re-
sults are hereafter presented as depending on the mode of
stabilization used for the synthesis of the particles.
[
10]
lytic performances of the metal centers. Our strategy then
relied on the combination of the advantages of both a sulfonat-
ed diphosphine as efficient stabilizer for metal nanoparticles in
aqueous solution and a cyclodextrin for its shuttle and supra-
molecular control effects in catalysis. The goal was to have
highly stable metal nanoparticles in water to characterize the
ligand L/CD association and to study its influence on the reac-
tivity during hydrogenation reaction of aromatic model sub-
strates, in terms of conversion and selectivity. Furthermore, this
work integrates well in the principles of green chemistry by an-
swering to a few aims as: high number of reactive sites at the
surface of metal nanoparticles to increase catalytic perfor-
mance, immobilization of the nanocatalysts in aqueous phase
as an environmentally friendly way to produce organic com-
pounds, use of an aqueous/organic biphasic process for recov-
ering and recycling of the catalyst, and use of a smart combi-
nation between a ligand and a phase transfer promoter to im-
prove selectivity through a supramolecular control.
Synthesis and characterization of the sulfonated-diphos-
phine-stabilized RuNPs
First of all, the RuNPs were prepared by using the sulfonated
diphosphine ligand L, in different [L]/[metal] molar ratio (0.1,
0.2, or 0.5, Scheme 1; method a). This sulfonated diphosphine,
containing four carbon atoms between the two phosphorus
atoms, was chosen according to the best catalytic results previ-
[7]
ously obtained with this ligand. The so-obtained RuNPs were
found highly stable in THF solution, as no precipitation of bulk
metal was observed at least after 3 months, thus demonstrat-
0
ing the efficiency of L in stabilizing the Ru nanoclusters. After-
wards, the nanoparticles were isolated as a black solid by pre-
cipitation from the THF colloidal solution through addition of
pentane and, remarkably, they were easily redissolved in water.
Aqueous colloidal solutions were used for characterization of
Herein we thus report: 1) the synthesis of water-soluble
ruthenium nanoparticles using either a sulfonated diphosphine
or its combination with the RAME-b-CD, the latter of which, as
far as we know, has been de-
scribed here for the first time;
2) their
characterization
by
liquid and solid-state NMR spec-
troscopy, elemental analysis,
TEM, high-resolution (HR) TEM,
wide-angle X-ray scattering
(WAXS), and dynamic light scat-
tering (DLS); and 3) the investi-
gation of these nanosystems in
hydrogenation reaction of poly-
functional aromatic molecules
(styrene, acetophenone, and m-
methylanisole) in substrate–
water biphasic conditions. This
study evidences the influence of
the cyclodextrin, if present, on
the catalytic performances of
the nanocatalysts compared to
those obtained with diphos-
phine-stabilized nanocatalysts.
3
Scheme 1. Synthesis of the RuNPs systems. R=H or CH , degree of substitution=1.8.
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3803

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
the particles in solution as well as for their investigation in cat-
alytic hydrogenation reactions during which they were again
found visually stable without observation of agglomerates. The
excellent water solubility of these RuNPs and their high stabili-
ty result from 1) the strong s-coordination of the phosphorous
atom to the metal surface (see below) and 2) coulombic repul-
sion of the charged sulfonated groups of the ligand molecules.
TEM analysis from colloidal solutions in THF and water re-
vealed the presence of small and well-dispersed particles dis-
playing a mean diameter between 1.2 and 1.5 nm, depending
[
7]
on the [L]/[metal] molar ratio (0.1, 0.2, and 0.5). These mean
diameters correspond to the metallic cores of the particles as
the ligand shells are not visible by TEM. Importantly, the trans-
fer of the RuNPs into water did not induce any relevant
change in dispersion and in their mean diameters as presented
in Figure 1 for the [L]/[Ru] ratio of 0.1. The hexagonal close
Figure 3. HRTEM analysis of RuNPs with [L]/[Ru] ratio=0.1.
ameter of 1.5 nm as determined by TEM (Figure 3) and distan-
ces of inter-reticular planes in agreement with face-centered
cubic (fcc) structure. However, they are single crystals display-
ing reticular planes extending over the entire particles. Well-
oriented nanoparticles allowed us to measure the lattice pa-
rameters through fast-Fourier transform (FFT) spectroscopy in-
dicating they were very close to the theoretical ones expected
for the hcp structure of bulk Ru. Moreover, energy-dispersive
X-ray analysis (EDX) performed during HRTEM studies revealed
a characteristic pattern of metallic Ru (Supporting Information,
Figure S1).
The RuNPs in aqueous solution were also analyzed by dy-
namic light scattering (DLS) giving interesting information on
the nanoparticles dispersion in water (Figure 4). There is
Figure 1. TEM analysis from colloidal solution in THF and water of RuNPs
with [L]/[Ru] ratio=0.1.
packed (hcp) structure of bulk Ru was determined by WAXS
(Figure 2), with coherence lengths in the range 1.0–1.5 nm, in
good agreement with the mean diameter measured by TEM.
HRTEM analysis of RuNPs with [L]/[Ru] ratio of 0.1 from aque-
ous solutions revealed well-crystallized RuNPs with a mean di-
Figure 4. DLS analysis of RuNPs with [L]/[Ru] ratio=a) 0.1, b) 0.2, and c) 0.5.
a good correlation in the mean diameters of the nanoparticles
obtained by DLS (1.5–1.8 nm) with those measured by TEM
(
1.2–1.5 nm). These results indicate that the NPs are isolated
and do not agglomerate in water.
The interaction of L at the NPs surface was first characterized
1
31
by H and P solution NMR performed on D O colloidal solu-
2
tions containing RuNPs stabilized by L in the [L]/[Ru]=0.1, 0.2,
1
and 0.5 ratio. First of all, the H NMR spectrum obtained for
[
L]/[Ru]=0.1 (Figure S2) contained numerous sharp signals of
very weak intensity, which can be attributed to the partial hy-
drogenation of the ligands owing to the reductive synthesis
Figure 2. WAXS of RuNPs with [L]/[Ru] ratio=0.1 (c), 0.2 (c), and
0.5 (c). r=rayon.
conditions (decomposition of the precursor under H ), as previ-
2
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3804

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
ously observed with nonsulfonated L or other phosphines con-
[
11]
31
taining phenyl groups. On the P spectrum registered for
L]/[Ru]=0.1 (Figure S3), there was only a low sharp peak at
[
d=40.2 ppm assigned to diphosphine oxide moieties free in
solution but no signal was visible for the free diphosphine nor
for the coordinated diphosphine. In solution NMR, the absence
of signals for ligands coordinated to nanoparticles is generally
attributed to several factors, including metal Knight shift
owing to close proximity with the metal, fast T2 relaxation re-
sulting from a diminution of the ligand tumbling at the nano-
particles surface, chemical exchange in the intermediate range
[12]
of the chemical shift time scale and surface anisotropy.
Solution NMR studies performed on aqueous colloidal solu-
tions containing higher amount of ligand (RuNPs prepared
with [L]/[Ru] ratio of 0.2 and 0.5) gave also rise to the absence
of signals for coordinated diphosphine as well as to the pres-
ence of signals for oxidized free diphosphine. However, in
these cases, a sharp signal at d=À15.2 ppm was observed on
31
Figure 5. P CPMAS solid-state NMR of a) L, b) RuNPs stabilized by L with
the [L]/[Ru] ratio=0.1, c) mixture of L+CD, and d) RuNPs stabilized by L and
CD with [Ru]/[L]/[CD] ratio=1.0:0.1:1.0. N: Spinning side bands.
Synthesis and characterization of the diphosphine L/RAME-
b-CDs stabilized RuNPs
3
1
the P NMR spectra with increased intensity if [L]/[Ru] in-
creased (Figure S3). This signal has the expected chemical shift
for the free ligand, which was confirmed by measuring the dif-
fusion coefficient found equal to the one of L alone (2.3Æ0.1ꢂ
Inspired by previous observations with molecular catalysts in
water, we attempted to modify the coordination properties of
the sulfonated diphosphine L by adding an extra agent,
namely a cyclodextrin (CD). Cyclodextrins are known for their
shuttle effect and for modifying coordination properties of sul-
fonated diphosphines into molecular complexes thus tuning
their catalytic performances.
À10
2
À1
10
m s ). These results suggest that there was no exchange
between attached L ligands to the RuNPs surface and free L li-
gands in solution or if there was one, it was very slow at the
NMR timescale. Recently, transfer NOE spectroscopy (trNOE)
has been reported as a powerful NMR measurement for the
characterization of grafted ligands at the nanoparticles surface
A second set of experiments was then performed with
a combination of 1,4-bis[(di-m-sulfonatophenyl)phosphino]bu-
tane (L) and RAME-b-CD (Scheme 1; method b). Several quanti-
ties of RAME-b-CD (0.2, 1.0, and 5.0 equiv.) were investigated.
Considering the ability of CDs to form supramolecular inclu-
sion entities with phosphine ligands of molecular complexes in
[
13]
in fast exchange with free ligands in solution. In the present
case, only weak positive NOE or zero quantum artifacts were
observed in the NOESY spectrum of a colloidal solution con-
taining RuNPs prepared with 0.5 equivalents of L, typical of
protons with high mobility. The absence of trNOE signals con-
firms that, if there was any exchange of L ligands coordinated
to the surface of RuNPs with L molecules free in solution, this
was relatively slow on the NMR timescales.
[10]
aqueous media, a similar behavior could be expected with
diphosphine-stabilized NPs. Our objective was then to study
the influence of the CD in the reaction medium both on the
stability of the sulfonated-diphosphine-stabilized RuNPs and
on their catalytic properties that could result from a modifica-
tion of the coordination of the diphosphine at the metal sur-
face. For that purpose, the CD was added from the beginning
of the RuNPs synthesis.
Finally, to obtain more information about the system, the
nanoparticles were analyzed by cross-polarization magic angle
31
spinning (CPMAS) solid-state NMR spectroscopy. The
P
CPMAS NMR spectrum of RuNPs with [L]/[Ru] ratio of 0.1 ex-
hibited a broad signal between 30 and 60 ppm (Figure 5b)
corresponding to the coordinated ligand through the two di-
phenylphosphino groups because no signal was detected at
approximately À16 ppm, as observed for the free-ligand
TEM and HRTEM analyses in water (Figure 6 and 7, respec-
tively) of the [Ru]/[L]/[CD] nanoparticles did not reveal any dif-
ference in comparison with those of [Ru]/[L]. The nanoparticles
were well-dispersed on the grids and displayed similar diame-
ter and diameter distributions. They displayed distances be-
tween inter-reticular planes in accordance with the hcp
structure.
31
P NMR (d=À15.8 ppm; Figure 5a).
From all these results, we can conclude that the use of
higher quantity than [L]/[Ru]=0.1 for the preparation of
RuNPs stabilized by 1,4-bis[(di-m-sulfonatophenyl)phosphino]-
butane led to excess ligand that did not exchange with the co-
ordinated molecules at the particle surface. Moreover, we pre-
viously observed in catalysis that a higher [L]/[Ru] ratio was
The DLS measurements performed on the three L/CD-stabi-
lized RuNPs samples ([Ru]/[L]/[CD]=1.0:0.1:0.2, [Ru]/[L]/[CD]=
1.0:0.1:1.0, and [Ru]/[L]/[CD]=1.0:0.1:5.0) gave rise to higher
mean diameters than those of L-stabilized RuNPs obtained by
previous measurements (Figure 8). However, the observed
mean diameters were dependent upon the quantity of CD
present during the particles synthesis, higher CD ratios leading
to higher mean diameters. The hydrodynamic radii of the
whole particles (that means the metallic cores with the ligand
[
7]
not beneficial in terms of activity. For these reasons, in the
continuation of this study, only the [L]/[Ru] ratio of 0.1 was
considered.
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3805

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
lution. It appears clearly that CD surrounded the RuNPs. Con-
sidering the probable proximity of L and CD at the particle sur-
face, a supramolecular interaction between L and CD could
[10]
take place as previously observed for molecular complexes.
To check the potential formation of an inclusion complex, be-
tween the aromatic rings of the diphosphine and the cavity of
the RAME-b-CD cyclodextrin that could modulate the coordina-
tion properties of L, deep-NMR experiments were performed,
in D O solution and in the solid state.
2
First of all, a mixture [L]/[CD]=1.0 (without metal) was stud-
1
ied in the concentration conditions of the RuNPs synthesis. H
31
and P NMR spectra indicate a clear interaction between L and
CD by resonance shifts if L and CD were mixed in D O (Fig-
2
1
ure S4 and S5). Only one set of H NMR resonances were ob-
served for CD and L suggesting the presence of a fast equilibri-
um between associated and dissociated states. The diffusion
coefficient of CD was slightly reduced from 1.7Æ0.1ꢂ
À10
2
À1
À10
2
À1
1
0
m s to 1.5Æ0.1ꢂ10 m s , and that of L exhibited
À10 2 À1
a more pronounced reduction from 2.3Æ0.1ꢂ10 m s to
À10
2
À1
1
.6Æ0.1ꢂ10 m s . Remarkably, intermolecular NOEs were
observed between some CD protons and the aromatic protons
of L (Figure S6). These NOEs could be related to the inclusion
of one aromatic ring of L inside the cyclodextrin cavity as al-
[
10]
ready described by some of us. We also observed that the
initially positive intramolecular NOEs of L (Figure S6 left)
became negative (Figure S6 right), thus indicating a decrease
of the average local mobility of the L protons, owing to the in-
teraction with CD. To summarize, all these data confirm the
presence of a weak interaction between L and CD and a fast
exchange (on the NMR timescale) between associated and dis-
sociated states. Interestingly, this interaction could also be ob-
served in the solid state, in which important modifications of
Figure 6. TEM analysis from colloidal solution water of RuNPs with [L]/[Ru]
ratio=0.1 and CD. a) [Ru]/[L]/[CDs]=1.0:0.1:0.2, b) [Ru]/[L]/[CDs]
ratio=1.0:0.1:1.0, and c) [Ru]/[L]/[CDs] ratio=1.0:0.1:5.0.
13
31
C and P CPMAS NMR L resonances were detected in the
31
presence of CD (Figure 5c). Notably the P CPMAS signal of L
alone at d=À15.8 ppm was split in two resonances at d=
À11.7 and d=À4.8 ppm for [L]/[CD]=1.0. From these two
phosphorus resonances that reveal the presence of two none-
quivalent phosphorus atoms in the diphosphine ligands, we
can assume the formation of an inclusion complex between
the diphosphine and the CD, through an interaction of phenyl
groups of L in the cavities of the CD molecules. However, the
formation of this inclusion complex is not affected by excess
CD.
Figure 7. HRTEM analysis for the [Ru]/[L]/[CDs] (ratio=1.0:0.1:0.2).
Finally, NMR experiments were also performed with the
RuNPs system prepared with L and CD in the ratio [Ru]/[L]/
[
CD]=1.0:0.1:1.0, both in D O and in the solid state, giving rise
2
to results very similar to the ones obtained for the mixture [L]/
31
[
CD]=1.0 (Figure 5d and 9). Notably in the P CPMAS NMR
spectrum, two major signals were observed at d=À11.7 and
d=À4.8 ppm, as already detected for [L]/[CD]=1.0. Further-
more, the broad signal between 30 and 60 ppm (Figure 5b)
previously observed for the RuNPs prepared with a [Ru]/[L]=
Figure 8. DLS analysis of RuNPs with [L]/[Ru] ratio=0.1 and CD a) 0.0, b) 0.2,
c) 1.0, and d) 5.0 equivalents.
0.1 ratio and attributed to coordinated diphosphines to the Ru
surface was also detected but with a very weak intensity. Con-
jointly, these results indicate that, in the presence of a large
amount of free CD relative to diphosphine ([Ru]/[L]/[CD]=
1.0:0.1:5.0), L interacts more strongly with the CD molecules
shells) were determined by DLS, thus giving direct information
on the influence of CD in the organization of the RuNPs in so-
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3806

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
cles are able to activate dihydrogen and that whatever the
quantity of CD used was, the number of hydrides per surface
Ru atom is not affected.
To investigate the influence of the RAME-b-CD on the cata-
lytic performances of our previously described sulfonated-di-
phosphine-stabilized nanomaterials, we chose to compare
them in the catalytic hydrogenation of functionalized aromatic
substrates, namely styrene, acetophenone, and m-methylani-
sole. The presence of CD should have an influence on the reac-
tivity in terms of selectivity and kinetic properties. First, owing
to the strong interaction between one phosphorous atom of
the ligand and CD giving rise to an inclusion complex as dem-
onstrated previously, the coordination of the ligand at the
metal surface was modified in the presence of CD which can
lead to different activity and selectivity. Secondly, the presence
of CD in excess should help the transfer of the aromatic sub-
strate towards the metallic surface through the hydrophobic
cavity that can host organic compounds and further allow an
increase the kinetic behavior of the hydrogenation reaction.
To study the influence on the reduction rate and chemose-
lectivity of diphosphine-stabilized nanoparticles chemically
modified by CD, similar catalytic experiments were performed
with both nanocatalysts Ru/L and Ru/L/CD. The hydrogenation
reactions were performed in pure biphasic liquid–liquid condi-
tions (substrate/water) with a [substrate]/[RuNPs] molar ratio
Figure 9. NOESY spectra of RuNPs stabilized by 0.5 equivalents of L (left)
and by L and CDs with [L]/[CD] ratio=5.0 (right). NOE cross-peaks between
the aromatic L and CD protons are shown in the dotted rectangle.
than with the RuNPs. Thus, the use of a mixture of CD and L
induced a strong diminution of the affinity of L for the RuNPs
surface and, in consequence, CD molecules were probably
close to the surface of RuNPs.
Additional NMR experiments performed after adding excess
CD to a colloidal solution containing preformed sulfonated di-
phosphine stabilized RuNPs led to similar results for [Ru]/[L]
nanoparticles (no change observed on the NMR spectra).
These results point out that the coordination of the diphos-
phine at the nanoparticles’ surface was very strong and was
not modified by the postaddition of CDs. In summary, the in-
clusion complex L/CD can be formed only if the diphosphine
and the CD are both present during the synthesis of the nano-
particles, thus influencing the coordination of the diphosphine
at their surface and consequently their properties.
of 100, at room temperature and under H pressure (1 and
2
10 bar). The selectivities were determined by GC analysis. The
results are summarized in Tables 1–3. It is noteworthy to men-
tion that, given the number of surface Ru atoms was much
lower than the total metal atoms present in the sample, the
values given in Tables 1, 2, and 3 are underestimated. For ex-
ample, if we consider the formation of full-shell clusters of
ruthenium, the RuNPs of 1.6 nm have an approximate disper-
sion (D; the fraction of exposed Ru) value of approximately
Application of the RuNPs in catalytic hydrogena-
tion reactions
[a]
Table 1. Styrene hydrogenation under hydrogen pressure catalyzed by RuNPs.
Owing to the synthesis conditions of our RuNPs (di-
hydrogen atmosphere), the presence of hydrides at
their surface was expected. The titration of surface
hydrides could be simply performed by investigating
the particles as catalysts in the hydrogenation model
reaction of norbornene with no extra hydrogen
added. The measurement of the amount of alkane
formed by GC analysis allowed determining the suit-
able quantity of hydrogen atoms for reducing the
alkene, and further to calculate the H/surface Ru
atomic ratio considering the nanoparticles mean di-
ameter. The presence of hydrides at the surface of
sulfonated-diphosphine-capped metal nanoparticles
was previously confirmed with 1.6 hydrogen atoms
[
b]
[c]
[d]
Entry Nanocatalyst
H
2
pressure
t
Product selectivity [%] TON TOF
À1
[
bar]
[h] ST
EB
EC
[h
]
1
2
3
4
5
6
7
8
Ru/L
1
1
1
1
10
10
10
10
40
40
40
40
2
0
0
0
0
0
0
0
0
0
0
0
9
0
0
0
77
100
100
100
91
100
100
100
23
71
178
591
2700
67
2 (3)
5 (11)
15 (24)
68 (108)
33 (57)
81 (142)
Ru/L/0.2CD
Ru/L/1.0CD
Ru/L/5.0CD
Ru/L
Ru/L/0.2CD
Ru/L/1.0CD
Ru/L/5.0CD
2
2
2
162
604
303 (473)
2448 1224 (2160)
À3
[
a] Reaction conditions: RuNPs (10 mg), styrene (3.9ꢂ10 mol), H
temperature, water (10 mL). [b] EB=ethylbenzene; EC=ethylcyclohexane; ST=sty-
rene, determined by GC analyses. [c] Initial TON is expressed as
2
(1 or 10 bar), room
[
7]
per surface Ru atom for Ru/L. This value was herein
determined for aqueous colloidal solutions of [Ru]/
[
(
L]/[CD] nanoparticles, giving rise to similar values
1.3, 1.3, and 1.2 for 0.2, 1.0, and 5.0 equiv. of CD, re-
metalÀ1
metalÀ1 À1
molsubstrate converted mol
molsubstrate converted mol
(molsubstrate converted =100%). [d] Initial TOF is expressed as
h
and in brackets TOF is corrected for expression as
À1 À1
molsubstrate converted molsurface Ru atoms
h .
spectively). These results indicate that the nanoparti-
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3807

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
catalyst offered only 91% of totally reduced product, thus
showing that a high amount of CD (5.0 equiv.) potentially
tunes the kinetic behaviors and selectivities of the hydrogena-
tion reaction. If the dihydrogen pressure was increased to
Table 2. Acetophenone hydrogenation under hydrogen pressure
catalyzed by RuNPs.
[
a]
10 bar, comparable results were observed but the reaction was
accelerated, with completion in only 2 h against 40 h under at-
mospheric pressure at room temperature. However, the Ru/L/
5
.0CD nanosystem presented a more pronounced difference
[
b]
[c]
[d]
Entry Nanocatalyst
H
2
pressure
t
[h]
Products [%] TON TOF
À1
with 77% of selectivity into ethylbenzene (EB) and 23% of EC
[
bar]
A
B
C
[h
]
compared with the results observed under 1 bar H . It appears
2
1
2
3
4
5
6
7
8
Ru/L
1
1
1
1
20
20
20
20
2
2
2
2
0
84
86
82
91
0
16
11
12
71
151
414
567
71
4 (6)
8 (13)
21 (34)
28 (45)
36 (57)
89 (142)
that 5.0 equivalents of CD limit the total hydrogenation of sty-
rene and that this effect is increased at higher pressure. This
phenomenon could be explained by the formation of a com-
petitive inclusion complex owing to the presence of a large
excess of CD, which could wrap more efficiently the aromatic
ring of styrene, thus limiting its hydrogenation in the same re-
action time. According to the TOF values based on the amount
of Ru introduced, CD-rich nanocatalysts finally appeared more
active, thus evidencing the shuttle effect of the CD, which im-
proves the approach of the C=C double bond of styrene to-
wards the metal surface.
Ru/L/0.2CD
Ru/L/1.0CD
Ru/L/5.0CD
Ru/L
Ru/L/0.2CD 10
Ru/L/1.0CD 10
Ru/L/5.0CD 10
3
6
9
0
0
0
10
100
100
34
0
178
14 52
14 60
514 257 (411)
26 1890 944 (1510)
À3
[
(
a] Reaction conditions: RuNPs (10 mg), acetophenone (3.9ꢂ10 mol), H
1 or 10 bar), room temperature, water (10 mL). [b] Determined by GC
2
analysis; A: 1-cyclohexylethanone, B: 1-phenylethanol, C: 1-cyclohexyle-
thanol, determined by GC analysis. [c] Initial TON is expressed as
À1
molsubstrate converted molmetal
.
h
[d] Initial
TOF
is
expressed
as
metalÀ1 À1
molsubstrate converted mol
and in brackets TOF is corrected for expres-
À1 À1
The second set of catalytic tests was performed with aceto-
phenone, in similar reaction conditions (room tem-
sion as molsubstrate converted molsurface Ru atoms
h .
perature, 1 or 10 bar H ). Under 1 bar H (Table 2;
2
2
entries 1–4), turnover number (TON) and TOF values
tended to increase, showing a positive effect of the
CD, compared to the values for the Ru/L nanocata-
lyst. In terms of selectivities, the 1-phenylethanol (B,
[a]
Table 3. Hydrogenation of 1-methoxy-3-methylbenzene (m-methylanisole).
82–86%) was the major product but 1-cyclohexyle-
[
b]
[c]
[d]
thanone (A, 3%–9%) and 1-cyclohexylethanol (C,
11–16%) were also produced, which indicates that
the hydrogenation of the aromatic cycle could take
place.
2
Entry Nanocatalyst H pressure t Conversion Diastereomeric TON TOF
[
b]
À1
[
bar]
[h] [%]
excess [%]
[h
]
1
2
3
4
5
6
7
8
Ru/L
1
1
1
1
40 18
40 67
40 45
40 35
51
60
66
62
51
62
72
100
13
120
266
945
57
<1
3 (5)
7 (11)
24 (38)
28 (45)
89 (142)
Ru/L/0.2 CD
Ru/L/1.0CD
Ru/L/5.0CD
Ru/L
Ru/L/0.2CD 10
Ru/L/1.0CD 10
Ru/L/5.0CD 10
However, no 1-cyclohexylethanol was detected
with the Ru/L/5.0CD system. The increase in reac-
10
2
80
tion pressure from 1 to 10 bar H provided strong
2
2 100
2 100
2 100
178
591 295 (472)
2700 1350 (2160)
acceleration of the kinetic properties for all catalytic
systems. As already observed under 1 bar H , TON
2
À3
and TOF values increased with higher quantity of
CD. Concerning the selectivity, the total hydrogenat-
ed product (1-cyclohexylethanol; C) was obtained
with 100% selectivity with Ru/L and Ru/L/0.2CD
nanocatalysts after 2 h of reaction, whereas the sys-
tems more rich in CD (Ru/L/1.0CD and Ru/L/5.0CD)
presented only 34 and 26% of C after the same
time, indicating an influence of the CD on the course of the re-
action at a suitable amount of CD (CDꢀ1.0). 1-cyclohexyletha-
none (A) was also detected for Ru/L/CD with CDꢀ1.0, in
[
2
a] Reaction conditions: RuNPs (10 mg), methylanisole (3.9ꢂ10 mol), H (1 or 10 bar),
room temperature, water (10 mL). [b] Determined by GC analysis. [c] Initial TON is ex-
À1
pressed
as
molsubstrate converted molmetal
.
[d] Initial
TOF is
expressed
as
metalÀ1 À1
molsubstrate converted mol
h
and in brackets TOF is corrected for expression as
À1 À1
molsubstrate converted molsurface Ru atoms
h .
0
.62, indicating that approximately two-thirds of the total
amount of Ru atoms are on the surface (which are active for
the hydrogenation). Thus, we used this value to estimate turn-
over frequency (TOF) corrected by the fraction of surface metal
atoms, given in brackets in Tables 1, 2, and 3.
higher content than that observed under 1 bar H (14%
2
against 3–9%) for the catalytic systems Ru/L/0.2CD and Ru/L/
1.0CD. Finally, according to the TOF values and as already seen
for styrene, CD-rich nanocatalysts appear more active, thus evi-
dencing the shuttle effect of the CD.
The RuNPs systems (Ru/L; Ru/L/0.2CD; Ru/L/1.0CD; and Ru/
L/5.0CD) were first evaluated in the hydrogenation of styrene
(
Table 1). Whatever the catalytic system, after 40 h under 1 bar
H , both the exo-cyclic double bond and the aromatic ring
In summary, it appears that the presence of the cyclodextrin
during the synthesis of the nanoparticles has an influence on
the catalytic performances of the sulfonated-diphosphine-sta-
bilized RuNPs in the hydrogenation of styrene and acetophe-
2
were reduced giving rise to 100% conversion with a selectivity
into ethylcyclohexane (EC) of 90–100%. Although all the other
systems led to the complete formation of EC, the Ru/L/5.0CD
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3808

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
none, both in terms of activity and selectivity. First of all, the
increasing TOF values with increasing CD content show that
nanocatalysts more rich in CDs are more active, thus evidenc-
ing the shuttle effect of the CD. Nevertheless, we observed
some differences depending on the quantity of CD. At a rela-
tively low quantity of CD (0.2 equiv.), the effect was low. This
can be explained by the fact that the CD was mainly involved
in the formation of an inclusion complex with the diphosphine
ligand at the metal surface, thus leading to stable RuNPs that
were active but with no boosting effect. At higher quantity of
CD, there was free CD in the reaction medium which improved
the catalytic system, and this was even more pronounced with
CD=5.0 compared to the situation with CD=1.0. Concerning
the selectivity, we also noticed some differences if CD was
present at a ratio CDꢀ1.0. In the case of styrene, a high quan-
tity of CD (5.0 equiv.) limited the total hydrogenation of sty-
rene, probably owing to the formation of an inclusion complex
between the aromatic ring and the excess CD thus avoiding its
hydrogenation in the same reaction time. In the case of aceto-
phenone, the product selectivity tended also to vary with the
CD content: CDꢀ1 contents led to higher quantities of partial-
ly hydrogenated products (1-phenylethanol and 1-cyclohexyle-
thanone) and this was even more pronounced with CD=
for the Ru/L/CD nanocatalysts whatever the quantity of CD
was, all leading to 100% conversion in 2 h. TON and TOF
values were also increased if higher quantities of CD were
present in the reaction medium. Diastereomeric excesses be-
tween 50 and 100% were noticed, in favor of the kinetic cis
product, evidencing that the quantity of CD had an influence
on the selectivity of the reaction at higher hydrogen pressure,
and the more CD-rich nanocatalysts (Ru/L/5.0CD) were the
more selective.
All these results highlight the interest of combining a sulfo-
nated diphosphine ligand with a cyclodextrin to tune the cata-
lytic performances of RuNPs, as the formation of strong inclu-
sion complexes between the ligand and/or the substrate
within the cage of the cyclodextrin may lead to different selec-
tivities and excess CD improve the activity through a mass-
transfer promoter effect.
Conclusions
Sulfonated diphosphines were very efficient ligands for the sta-
bilization of ruthenium nanoparticles (RuNPs) synthesized
through the organometallic approach, giving rise to well-con-
trolled nanoclusters in the diameter range 1.2–1.5 nm and dis-
playing very low diameter dispersity. The water-solubility of
the sulfonated diphosphine allowed obtaining very stable
aqueous colloidal solutions (up to several months) by a simple
transfer of the isolated particles into water. Deep-NMR studies
conducted in the solid state as well as on the aqueous colloi-
dal solutions evidenced the strong interaction of the sulfonat-
ed diphosphine ligand with the RuNPs surface. If a randomly
methylated b-cyclodextrin was used as coadditive in the syn-
thesis of the sulfonated-diphosphine-stabilized RuNPs, the for-
mation of supramolecular inclusion complexes between the
sulfonated diphosphine and cyclodextrin molecules was ob-
served. The existence of an interaction between ligand and cy-
clodextrin was clearly evidenced by NMR spectroscopy (both
in solution and in solid state), which appeared even stronger
at higher cyclodextrin content, thus disrupting severely the co-
ordination properties of the ligand towards the metal surface.
However, it was observed that this interaction could take place
only if the cyclodextrin is present during the synthesis of the
particles.
5.0 equivalents. Interestingly, representative TEM images from
colloidal solution in water of RuNPs after catalysis showed that
the nanoparticles were not agglomerated (Figure S7).
The investigation in catalysis was pursued by studying the
hydrogenation of a substrate of higher interest, namely 1-me-
thoxy-3-methylbenzene (m-methylanisole). This disubstituted
aromatic substrate was chosen to evaluate the influence of the
CD on the reaction stereoselectivity because two diastereo-
meric products (cis/trans) can be formed. The catalytic reac-
tions were performed at room temperature and under 1 or
1
0 bar H (Table 3). The reaction presented slowly moderate
2
conversions (18–67%) after 40 h, with the Ru/L nanocatalyst
the less active. The highest conversion was observed for the
nanocatalyst Ru/L/0.2CD with 67% of totally hydrogenated
product after 40 h. As previously observed for styrene and ace-
tophenone, higher quantities of CD led to lower conversions
(45 and 35 for Ru/L/1.0CD and Ru/L/5.0CD, respectively) but
higher TONs and TOFs were observed. Finally, as usually ob-
served in the hydrogenation of aromatic derivatives with pure
heterogeneous catalysts or aqueous colloidal suspensions, the
formation of the thermodynamically less favorable cis-diaste-
reomer was promoted with diastereomeric excesses up to
The catalytic properties of the sulfonated diphosphine-stabi-
lized RuNPs and sulfonated-diphosphine–cyclodextrin-stabi-
lized RuNPs were compared in the hydrogenation reaction of
unsaturated model substrates (styrene, acetophenone, and m-
methylanisole) in biphasic liquid–liquid conditions. All the
RuNPs displayed pertinent catalytic performances. Based on
the TOF values, the relevant differences in terms of activity and
selectivity highlighted the tuning of the catalytic performances
of the nanocatalysts in the presence of cyclodextrin. The cyclo-
dextrin acted as a phase-transfer promoter by increasing the
activity of the reaction but also affected the selectivity. As evi-
denced by NMR studies, the influence on selectivity may result
from the formation of an inclusion complex between the CD
and the diphosphine ligand at the RuNPs surface, which modi-
fied their catalytic properties.
100%. Nevertheless, in all cases, we could observe a decrease
in the diastereomeric excesses (ꢁ60%) after longer reaction
times (40 h), indicating also the formation of the trans diaste-
[
14]
reoisomers through a probable roll-over mechanism. What-
ever the system Ru/L/CD used, no significant difference was
observed.
To increase the kinetic properties and the conversion rate in
short times, the same experiments were also performed under
1
0 bar H (Table 3). Undoubtedly, the hydrogenation of the aro-
2
matic cycle was accelerated, with better conversions (80–
00%) after 2 h. The Ru/L nanocatalyst was also the less active
conversion 80%) but no difference in activity was observed
1
(
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3809

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Taking advantage of the supramolecular properties of a cy-
clodextrin to modulate the surface reactivity of diphosphine-
stabilized RuNPs, this original work may open up new opportu-
nities in the field of nanocatalysis.
ment. IR were run on a Perkin–Elmer FT spectrophotometer, series
À1
2
1
000 cm as KBr pellets or polyethylene films in the range 4000–
50 cm . Data collection for WAXS was performed at the CEMES
À1
CNRS (Toulouse) on small amounts of powder. All samples were
sealed in 1 mm diameter Lindemann glass capillaries. The measure-
ments of the X-ray intensity scattered by the samples irradiated
with graphite monochromatized MoKa (0.071069 nm) radiation
were performed by using a dedicated two-axis diffractometer. Mea-
surement time was 15 h for each sample. Scattering data were cor-
rected for polarization and absorption effects, then normalized to
one Ru atom and Fourier transformed to obtain the radial distribu-
tion functions. To make comparisons with the crystalline structure
in real space, a model was generated from bulk Ru parameters.
The classic Debye’s function was then used to compute intensity
values subsequently Fourier transformed in the same conditions as
the experimental ones.
Experimental Section
Reagents and general procedures
All operations concerning nanoparticles syntheses were performed
in Schlenck or Fischer–Porter glassware or in a glove box under
argon atmosphere. The organometallic complex used as precursor,
(
(
1,5-cyclooctadiene)(1,3,5-cyclooctatriene)ruthenium(0)
[Ru(cod)(cot)]) was purchased from Nanomeps-Toulouse. Sulfonat-
complex
ed diphosphine (1,4-bis[(di-m-sulfonatophenyl)phosphino]butane
[7]
(
L) was synthesized by following a published procedure and
RAME-b-CD (Cavasol W7M) was purchased from Wacker Chemie
GmbH in its pharmaceutical grade and was used as received.
RAME-b-CD is a partially methylated b-cyclodextrin and its degree
Synthesis of ruthenium nanoparticles
[15]
of substitution is equal to 1.8 per glucopyranose unit.
The RuNPs system ([Ru]/[L]/[CD]=1.0:0.1:1.0) has been chosen to
describe the synthesis of the RuNPs, but all the syntheses per-
formed were done in this way. In a typical reaction, Ru/0.1L/1.0CD,
Solvents were dried and distilled before use: THF over sodium ben-
zophenone and pentane over calcium hydride. All reagents and
solvents were degassed before use by means of three freeze–
pump–thaw cycles. Water was distilled twice by conventional
method before use to prepare nanoparticle suspensions. Styrene
and acetophenone used as substrates in catalysis were purchased
from Acros Organics or Sigma-Alfa Aesar and used without further
purification.
[Ru(cod)(cot)] (150 mg, 0.476 mmol) were introduced in a Fischer–
Porter bottle and left in vacuum during 0.5 h and cooled to 193 K.
THF (150 mL), containing a mixture of L (41 mg, 0.048 mmol, [L]/
[Ru]=0.1) and CD (640 mg, 0.048 mmol, [CD]/[Ru]=1.0) were then
added. The Fischer–Porter bottle was heated to 298 K and then
pressurized with H (3 bar). After 18 h, a homogenous brown colloi-
2
dal solution was obtained. The volume of the solution was reduced
to approximately 10 mL by solvent evaporation before its transfer
onto a solution of deoxygenated pentane (100 mL). A brown pre-
cipitate formed, which was filtered and dried in vacuum, giving
rise to the nanoparticles as a dark brown powder. The molar ratio
of [L]/[Ru] was fixed to 0.1, taking account our previous results in
Characterization techniques
Samples for TEM/HRTEM analyses were prepared by slow evapora-
tion of a drop of crude colloidal solution deposited onto holey
carbon-covered copper grids under argon (in a glove box) for THF
solutions and under air for aqueous solution. TEM and HRTEM anal-
yses were performed at the Service Commun de Microscopie Elec-
tronique de l’Universitꢀ Paul Sabatier (UPS–TEMSCAN). TEM images
were obtained by using a JEOL1011 electron microscope operating
at 100 kV with resolution point of 4.5 ꢃ. HRTEM observations were
performed with a JEOL JEM 2010 electron microscope working at
[7]
catalysis, and the molar ratio of [CD]/[Ru] was varied from 0.2 to
.0. In all cases, these ruthenium colloids were found to be stable
5
with time under argon atmosphere without precipitation after sev-
eral months. Inductively coupled plasma MS analysis for the differ-
ent RuNPs: Ru/L (55.5 wt% Ru); Ru/L/0.2CD (22.1 wt% Ru); Ru/L/
1
.0CD (6.7 wt% Ru); Ru/L/5.0CD (1.5 wt% Ru).
2
00 kV with a resolution point of 2.5 ꢃ. The diameter distributions
Quantification of hydrides at the surface of ruthenium nano-
particles
were determined through manual analysis of enlarged micrographs
with Imagetool software to obtain a statistical diameter distribu-
tion and a mean diameter (counting a minimum of 150 particles).
FFT treatments were performed with Digital Micrograph Version
The quantification of hydrogen atoms adsorbed onto the surface
of Ru nanoparticles by gas chromatography (GC) analyses was per-
formed on aqueous colloidal solutions by following a previously
1
.80.70.
1
13
1
[7]
HNMR, C { H} NMR, HSQC, COSY, and NOESY experiments were
recorded on a Bruker Avance 500 spectrometer equipped with
a 5 mm triple resonance inverse Z-gradient probe. All samples
described procedure. On each fresh colloidal solution, 5.0 equiva-
lents of olefin (2-norbornene), previously filtered through alumina,
were added. Samples were taken from the solutions for GC analy-
ses and estimation of the norbornene conversion into norbornane.
The measurement of the amount of alkane formed by GC analysis
allows determining the necessary quantity of hydrogen atoms for
reducing the alkenes, and further to the H/surface Ru atomic ratio
considering the nanoparticles mean diameters.
were prepared in D O. The 2D NOESY measurements were per-
2
formed with a mixing time of 100 ms. All diffusion measurements
were made by using the stimulated echo pulse sequence. The dif-
fusion dimension was processed with single-exponential analysis
involving least-squares fitting (Topspin software). Solid-state NMR
experiments were recorded on a Bruker Avance 400 spectrometer
31
Gas chromatography was performed by using an HP 5890 Series II
Gas Chromatograph with a SGE BP1 nonpolar 100% dimethyl poly-
siloxane capillary column of 50 mꢂ0.32 mmꢂ0.25 mm. The method
used for the quantification of hydrides consisted of 15 min at 408C
equipped with a 4 mm probe. P CPMAS spectra were recorded
1
31
with a recycle delay of 5 s and a contact time of 2 ms. H and
chemical shifts are given relative to TMS and to an external 85%
H PO sample, respectively.
P
3
4
À1
and a ramp of 88Cmin until 2508C. Conversions were deter-
Elemental analyses (C, H, and N) were performed by the staff of
Chemical Analyses Service of the LCC on a Eurovector 3011 instru-
mined as following: the peak area of the norbornane divided by
the sum of the peak area of norbornene and norbornane.
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3810

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Catalytic hydrogenation reactions
[2] a) Nanoparticles and Catalysis (Ed.: D. Astruc), Wiley-VCH, Weinheim,
2
008; b) Z. X. Li, W. Xue, B. T. Guan, F. B. Shi, Z. J. Shi, H. Jiang, C. H. Yan,
Gas chromatography
Nanoscale 2013, 5, 1213–1220.
3] Nanomaterials in Catalysis (Eds.: P. Serp, K. Philippot), Wiley-VCH, Wein-
heim, 2013.
[
For all hydrogenation reactions, the conversion and the selectivity
were determined by GC. A Carlo Erba GC 6000 with a flame ioniza-
tion detector equipped with a Factor Four column (30 m, 0.25 mm
inner diameter) was used for styrene hydrogenation analyses. Pa-
[4] P. Lara, K. Philippot, B. Chaudret, ChemCatChem 2013, 5, 28–45.
[5] a) C. Burda, X. Chen, R. Narayanan, M. A. El-Sayed, Chem. Rev. 2005, 105,
1025–1102; b) A. Gual, C. Godard, K. Philippot, B. Chaudret, A. Deni-
court-Nowicki, A. Roucoux, S. Castillꢄn, C. Claver, ChemSusChem 2009,
2, 769–779; c) M. Zahmakıran, M. Tristany, K. Philippot, K. Fajerwerg, S.
ꢅzkar, B. Chaudret, Chem. Commun. 2010, 46, 2938–2940; d) E. Guyon-
net Bilꢀ, E. Cortelazzo-Polisini, A. Denicourt-Nowicki, R. Sassine, F.
Launay, A. Roucoux, ChemSusChem 2012, 5, 91–101; e) E. Guyonnet
Bilꢀ, A. Denicourt-Nowicki, R. Sassine, F. Launay, A. Roucoux, Dalton
Trans. 2011, 40, 6524–6531; f) D. Gonzalez-Galvez, P. Lara, O. Rivada-
Wheekaghan, S. Conejero, B. Chaudret, K. Philippot, P. N. M. W. Van
Leeuwen, Catal. Sci. Technol. 2013, 3, 99–105; g) D. Peral, F. Gꢄmez-Vil-
larraga, J. Garcꢆa-Antꢄn, X. Sala, J. Pons, J. Carles Bayꢄn, J. Ros, M. Guer-
rero, L. Vendier, P. Lecante, K. Philippot, Catal. Sci. Technol. 2013, 3,
rameters were as follows: initial temperature, 408C; initial time,
À1
1
3
0 min; ramp, 108C min ; final temperature, 808C; final time,
0 min; injector temperature, 2208C; detector temperature, 2508C.
For acetophenone and m-methylanisole hydrogenation reactions,
the reaction products were analyzed by using a Fisons Instrument
GC 9000 series with a flame ionization detector equipped with
a chiral Varian Chiralsil-Dex CB capillary column (30 m, 0.25 mm
inner diameter). Parameters were as follows: isotherm program
with oven temperature, 908C (m-methylanisole) or 1308C (aceto-
phenone); carrier gas pressure, 50 kPa.
4
75–489.
6] P.-J. Debouttiꢇre, V. Martinez, K. Philippot, B. Chaudret, Dalton Trans.
009, 10172–10174.
[
[
[
2
Atmospheric hydrogenation reactions
7] M. Guerrero, A. Roucoux, A. Denicourt-Nowicki, H. Bricout, E. Monflier, V.
Colliꢇre, K. Fajerwerg, K. Philippot, Catal. Today 2012, 183, 34–41.
8] a) F. Hapiot, A. Ponchel, S. Tilloy, E. Monflier, C. R. Chim. 2011, 14, 149–
A 25 mL round-bottom flask, charged with 10 mL of the colloidal
0
suspension of Ru NPs (10 mg) stabilized with the sulfonated di-
166; b) H. Bricout, F. Hapiot, A. Ponchel, S. Tilloy, E. Monflier, Curr. Org.
phosphine or with a mixture of sulfonated diphosphine–cyclodex-
trin and the appropriate amount of substrate (styrene, acetophe-
none, m-methylanisole) at RT was connected to a gas burette
Chem. 2010, 14, 1296–1307; c) F. Hapiot, L. Leclercq, N. Azaroual, S.
Fourmentin, S. Tilloy, E. Monflier, Curr. Org. Synth. 2008, 5, 162–172.
9] a) A. Nowicki, Y. Zhang, B. Lꢀger, J. P. Rolland, H. Bricout, E. Monflier, A.
Roucoux, Chem. Commun. 2006, 296–298; b) A. Denicourt-Nowicki, A.
Ponchel, E. Monflier, A. Roucoux, Dalton Trans. 2007, 5714–5719; c) C.
Hubert, A. Denicourt-Nowicki, A. Roucoux, D. Landy, B. Leger, G.
Crowyn, E. Monflier, Chem. Commun. 2009, 1228–1230; d) R. Herbois, S.
Noꢈl, B. Lꢀger, L. Bai, A. Roucoux, E. Monflier, A. Ponchel, Chem.
Commun. 2012, 48, 3451–3453; e) N. T. T. Chau, S. Handjani, J.-P.
Guegan, M. Guerrero, E. Monflier, K. Philippot, A. Denicourt-Nowicki, A.
Roucoux, ChemCatChem 2013, 5, 1497–1503.
[
(
500 mL) and a flask to balance pressure. Then, the system was
filled with hydrogen (PH2 =1 bar) and the mixture was magnetically
stirred at 1500 rpm. Samples were removed from time to time (2 h,
2
0 h, and 40 h) to monitor the reaction by GC in previously men-
tioned conditions.
High-pressure hydrogenation reactions
[
[
10] a) M. Ferreira, H. Bricout, A. Sayede, A. Ponchel, S. Fourmentin, S. Tilloy,
E. Monflier, Adv. Synth. Catal. 2008, 350, 609–618; b) L. Caron, H. Bric-
out, S. Tilloy, D. Landy, S. Fourmentin, E. Monflier, Adv. Synth. Catal.
2004, 346, 1449–1456; c) C. Binkowski, J. Cabou, H. Bricout, F. Hapiot, E.
Monflier, J. Mol. Catal. A-Chem. 2004, 215, 23–32; d) M. Canipelle, L.
Caron, C. Christine, S. Tilloy, E. Monflier, Carbohydr. Res. 2002, 337, 281–
The stainless steel autoclave was charged with the aqueous sus-
pension of the sulfonated diphosphine-stabilized or the sulfonated
diphosphine–cyclodextrin-stabilized Ru NPs and a magnetic stirrer
at RT. The investigated substrate was added into the autoclave and
was degassed three times. Finally, the hydrogen gas was admitted
0
287.
to the system at a constant pressure (10 bar H ). The mixture was
2
11] a) J. Garcꢆa-Antꢄn, M. Rosa Axet, S. Jansat, K. Philippot, B. Chaudret, T.
stirred vigorously at RT. Samples were removed from time to time
Pery, G. Buntkowsky, H.-H. Limbach, Angew. Chem. 2008, 120, 2104–
(
2 h, 20 h, and 40 h) to monitor the reaction by GC in previously
2108; Angew. Chem. Int. Ed. 2008, 47, 2074–2078; b) D. Gonzꢉlez-
mentioned conditions.
Gꢉlvez, P. Nolis, K. Philippot, B. Chaudret, P. W. N. M. van Leeuwen, ACS
Catal. 2012, 2, 317–321.
[
[
12] a) E. Ramꢆrez, L. Eradꢇs, K. Philippot, P. Lecante, B. Chaudret, Adv. Funct.
Mater. 2007, 17, 2219–2228; b) E. Ramꢆrez, S. Jansat, K. Philippot, P. Le-
cante, M. Gꢄmez, A. M. Masdeu-Bultꢄ, B. Chaudret, J. Organomet. Chem.
Acknowledgements
2
004, 689, 4601–4610.
The authors thank V. Colliꢁre and L. Datas (LCC–CNRS and UPS–
TEMSCAN) for TEM/HRTEM facilities. The authors are grateful to
CNRS and the Agence Nationale de la Recherche (ANR-09-BLAN-
13] B. Fritzinger, I. Moreels, P. Lommens, R. Koole, Z. Hens, J. C. Martins, J.
Am. Chem. Soc. 2009, 131, 3024–3032.
[14] A. Kalantar Neyestanaki, P. Maki-Arvela, H. Backman, H. Karhu, T. Salmi,
0
194) for the financial support of the SUPRANANO program.
J. Vayrynen, D. Y. Murzin, J. Catal. 2003, 218, 267–279.
[
15] a) M. Ferreira, F. X. Legrand, C. Machut, H. Bricout, S. Tilloy, E. Monflier,
Dalton Trans. 2012, 41, 8643–8647; b) F. X. Legrand, M. Sauthier, C. Fla-
haut, J. Hachani, C. Elfakir, S. Fourmentin, S. Tilloy, E. Monflier, J. Mol.
Catal. A-Chem. 2009, 303, 72–77.
Keywords: cyclodextrins
ruthenium · supramolecular chemistry
· nanoparticles · phosphanes ·
nd
[
1] Nanoparticles: From Theory to Application, 2 ed. (Ed.: G. Schmid),
Received: June 17, 2013
Wiley-VCH, Weinheim, 2012.
Published online on September 20, 2013
ꢁ
2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 3802 – 3811 3811