A direct novel synthesis of highly uniform dispersed ruthenium nanoparticles over P6mm ordered mesoporous carbon by host-guest complexes

Article abstract of DOI:10.1039/c4ta00038b

We report a novel concept to prepare a highly ordered mesoporous carbon with a uniform dispersion of ruthenium nanoparticles of 1-2 nm size range using a nano-templating method, based on the combined utilization of a β-cyclodextrin host-guest complex and ruthenium trichloride as respective sources of carbon and metal. The composite material synthesized (Ru@MCA-2) through the polymerization and carbonization of these metallo-supramolecular assemblies possesses high surface area and high pore volume after the removal of the silica template and exhibits high catalytic activity in the hydrogenation of unsaturated fatty acid methyl esters. The reusability of this nanoreplicated catalyst is also demonstrated. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4ta00038b

Journal of
Materials Chemistry A
View Article Online
View Journal | View Issue
PAPER
A direct novel synthesis of highly uniform dispersed
ruthenium nanoparticles over P6mm ordered
mesoporous carbon by host–guest complexes†
Cite this: J. Mater. Chem. A, 2014, 2,
6641
N. Gokulakrishnan,abc G. Peru,abc S. Rio,abc J. F. Blach,abc B. Leger,abc D. Grosso,d
´
abc
E. Monflierabc and A. Ponchel
*
We report a novel concept to prepare a highly ordered mesoporous carbon with a uniform dispersion of
ruthenium nanoparticles of 1–2 nm size range using a nano-templating method, based on the combined
utilization of a b-cyclodextrin host–guest complex and ruthenium trichloride as respective sources of
carbon and metal. The composite material synthesized (Ru@MCA-2) through the polymerization and
carbonization of these metallo-supramolecular assemblies possesses high surface area and high pore
volume after the removal of the silica template and exhibits high catalytic activity in the hydrogenation of
unsaturated fatty acid methyl esters. The reusability of this nanoreplicated catalyst is also demonstrated.
Received 3rd January 2014
Accepted 14th February 2014
DOI: 10.1039/c4ta00038b
www.rsc.org/MaterialsA
into account the metal-support interactions.8,9 The preparation
of well-dispersed metallic nanoparticles on carbon materials
Introduction
has been described by different techniques, such as ion-
exchange or impregnation and the nanocatalysts generally show
higher catalytic activities than those obtained on conventional
supported systems.10–13 However, it is generally assumed that
the most important drawback of metal nanoparticles is their
tendency to aggregate, especially at high metal contents.
Since the rst synthesis of ordered mesoporous carbon using a
hard-templating approach in 1999, this material has attracted a
great deal of attention in a number of high-potential applica-
tions including adsorption, sensing, separation, biotechnology
and catalysis. Its uses are usually related to its attractive prop-
erties such as high surface area, large pore volume, well-tailored
pore size, chemical inertness and electrical conducting prop-
erty.1,2 From a practical viewpoint, it is generally accepted that
the nal properties of ordered mesoporous carbon can be
controlled in terms of porosity and hydrophobic/hydrophilic
balances, by tuning the textural characteristics of silica
templates and selecting appropriate carbon sources. For this
purpose, researchers have developed protocols by exploiting
various carbon sources ranging from aliphatic to aromatic
compounds, such as sucrose, furfuryl alcohol, acenaphthene,
anthracene or dihydroxynaphthalene.3–7
The templating approach based on the use of solid molds
offers opportunities for developing novel ordered nano-
structures containing metal nanoparticles with a high degree of
size control and dispersion.14 In this context, supramolecular
organic templates involving the cooperative assembly of mole-
cules via non-covalent interactions can be regarded as promising
candidates for the fabrication of tailored metal-incorporated
mesoporous carbons. Thus, b-cyclodextrins (cyclic oligosaccha-
rides constituted of seven D-glucopyranose units) are a class of
naturally occurring receptors, nding applications in a number
of different elds including analytical chemistry, sensing enan-
tiomeric separation, remediation and catalysis.15–19 Until now,
the contribution of cyclodextrins to the elaboration of nano-
structured materials and catalysts20–24 has been little explored
and surprisingly, the possibility of using host–guest complexes
with cyclodextrins for the synthesis of structured mesoporous
carbons has not yet been investigated. Thus, the pioneering
studies of Antonietti have shown that worm-like nanoporous
silica replicas could be prepared by using cyclodextrins20 and
cyclodextrin-based polypseudorotaxanes21 as porogen agents by
means of a one-pot sol gel process. Pore sizes close to those of
the diameters of cyclodextrins (1.5–2 nm) were obtained in the
nal silica materials. Concerning carbon materials, Han et al.
described the synthesis of porous carbon replicas using a sol–gel
nanocasting procedure involving the co-condensation of
Additionally, for the design of robust and high-efficiency
metal supported carbon catalysts, it is important not only to
control the size and shape of the nanoparticles, but also to take
aUniv Lille Nord de France, F-59000 Lille, France
b
´
´
UArtois, Unite de Catalyse et Chimie du Solide (UCCS), Faculte des Sciences Jean
Perrin, Rue Jean Souvraz, F-62307 Lens, France. E-mail: anne.ponchel@univ-artois.fr
cCNRS, UMR 8181, F-59650 Villeneuve d'Ascq, France
d
`
´
`
Laboratoire Chimie de la Matiere Condensee de Paris, UMR-7574, College de France,
11, place Marcelin Berthelot, 75231 Paris, Franceo
† Electronic supplementary information (ESI) available: The wide angle XRD
pattern of Ru@MCA-2 sample (Fig. S1). Raman spectra of the carbon materials
obtained from host–guest complexes (RAMEB
+ ADAC) and cyclodextrins
(RAMEB) (Fig. S2). Main characteristics of the activated carbon used as supports
for the preparation by impregnation of the control Ru/C catalysts (Table S1). See
DOI: 10.1039/c4ta00038b
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6641–6648 | 6641

View Article Online
Journal of Materials Chemistry A
Paper
methylated b-CD with tetramethylorthosilicate followed by 1-adamantane carboxylic was supplied by Aldrich Chemicals.
carbonization and silica removal.22 However, the carbon mate- Two different activated carbons used as control support cata-
rial resulted in a disordered microporous material with a pore lysts were selected from commercially available samples.
diameter of ca. 1.6 nm. More recently, the fabrication of meso- Nuchar®WV-B (denoted AC-WV) was supplied by MeadWest-
porous carbons derived from cyclodextrins with a narrower pore vaco Corporation, Covington, USA. It was produced from wood
size distribution from 3.2 to 4.2 nm was reported by adding a and activated with phosphoric acid. Norit®SA-2 denoted
surfactant (Triton®TX-15) to the silicon alkoxide source.23
(AC-NorA) was produced by Norit (Amersfoort, Netherlands)
In this paper, we introduce a novel concept to prepare from peat and activated steam. All other reactants were
uniformly dispersed metal nanoparticles embedded in the pore purchased from Aldrich Chemicals and Acros Organics in their
walls of an ordered mesoporous carbon using a nanocasting highest purity and used without further purication.
procedure based on the original use of a cyclodextrin host–guest
complex and a water-soluble metal salt as respective sources of Synthesis of SBA-15
carbon and metal. Our idea is to take advantage of the unique
topologies of cyclodextrin (rigid cyclic structure, large number
of hydroxyl groups and ability to form inclusion complexes with
The starting SBA-15 was prepared according to a procedure
reported in the literature.29 Briey, 4 g of P123 was dispersed in
30 g of water by stirring for 3 h. Then, 120 g of a 2 M HCl solution
molecules of appropriate size and shape) to ensure good
dispersion of the metal precursor within the cyclodextrin-based
environment in the rst step of synthesis, i.e. during the pore
lling, and subsequently facilitate the formation of isolated
nucleation sites and growth of size-controlled metal clusters
during the thermal structuring of the carbon phase. The
strategy has been examined through the nanoreplication of
SBA-15 using randomly methylated b-cyclodextrin/1-ada-
mantane carboxylic acid (RAMEB–ADAC) as the host–guest
system and ruthenium trichloride (RuCl3) as the metal
precursor. 1-Adamantane carboxylic acid has been selected for
its ability to form stable host–guest complexes with b-CDs.
Indeed, the adamantyl group is a highly symmetrical stable
structure, which perfectly ts inside the cavity of b-CDs, form-
ing 1 : 1 inclusion complexes with high association constants
(typically in the range of 104 to 105 Mꢀ1).25–28 In addition, the
presence of the carboxylic acid functionality in the guest
molecule is also used to interact favorably with metals. There-
fore it is expected that this host–guest concept will facilitate the
anchoring of the metal ion to the cyclodextrin (via the carboxylic
group attached to the adamantane skeleton, itself included in
the cyclodextrin cavity) in the rst steps of the synthesis and will
lead to a good distribution of the metal species in the nal
carbon replica. For this purpose, we will report that, under well-
dened conditions, the utilization of RAMEB–ADAC host–guest
assemblies and RuCl3 allows the synthesis of P6mm ordered
mesoporous carbon with uniformly dispersed ruthenium
nanoparticles. Furthermore, experiments in the hydrogenation
of vegetable oils (methyl oleate) will demonstrate the potential
of this material in the eld of heterogeneous catalysis.
ꢁ
was added to the mixture followed by 2 h stirring at 40 C and
subsequently, 9.0 g of tetraethylorthosilicate was then added to
the homogeneous solution. The mixture was then stirred at
40 ꢁC for 24 h and the resulting gel underwent hydrothermal
treatment by heating in an oven at 100 ꢁC for 48 h. The obtained
white precipitate was ltered, dried at 100 ꢁC and calcined in a
muffle furnace at 550 ꢁC in air. Under these experimental
conditions, the as-prepared silica has a pore size of ca. 8.0 nm
and a pore wall thickness of 2.9 nm. The pore wall thickness was
calculated according the formula, a0 ꢀ (wd/1.050), reported by
Kruk et al., where wd is the BJH pore diameter.30
Preparation of ordered mesoporous ruthenium–carbon
material from the host–guest complex and RuCl3
The ruthenium carbon replica has been prepared using a molar
composition RAMEB–ADAC–Ru of 5 : 3 : 1. The detailed proce-
dure is as follows: 16 mg of RuCl3 and 18 mg of ADAC were
added to 4 ml of water and stirred for 3 h at room temperature.
Then 0.6 g of RAMEB was added to the previous mixture and the
stirring was continued for further 2 h. Thereaer, 70 mg of
H2SO4 was added and stirred for 5 minutes, and the resulting
mixture was inltrated into 0.5 g of SBA-ꢁ15. Subsequently, the
mixture underwent polymerization at 60 C for 24 h. The solid
obtained was crushed into powder using a mortar and subjected
to inltration again with 10 mg of RuCl3, 12 mg of ADAC, 0.4 g
of RAMEB and 46 mg of H2SO4 under the same conditions to ll
the pores completely. Aer polymerization, the silica–ruthe-
nium–carbon composite was pyrolyzed at 900 ꢁC for 5 h under a
nitrogen ow of 150 ml minꢀ1 (heating rate of 3 ꢁC minꢀ1). The
resulting material was treated with 35% HF to eliminate the
silica, ltered by washing with an excess amount of water and
ethanol and dried at 100 ꢁC. The as-synthesized material is
designated as Ru@MCA-2.
Experimental
Chemicals
Tetraethylorthosilicate (TEOS), poly(ethylene glycol)-block-
poly(propylene glycol)-block-poly(ethylene glycol) (P123;
EO20PO70EO20; average molecular weight, 5800), hydrochloric
Preparation of control carbon materials from RAMEB and
RAMEB–ADAC
acid (HCl, 37 wt%) and sulfuric acid were purchased from A ruthenium-free control sample, denoted as MCA-3 has been
Aldrich. Randomly methylated b-cyclodextrin (RAMEB) with an prepared in a similar manner as that previously described,
average substitution of 1.8 methyl groups per cyclodextrin was except that no ruthenium salt was introduced during the
purchased from Wacker Chemie. Ruthenium chloride hydrate procedure. The synthesis of the MCA-3 carbon was carried out
(40–43% of Ru) was purchased from Acros Organics whereas by using a mixture of RAMEB to ADAC with a ratio of 5 : 3 as
6642 | J. Mater. Chem. A, 2014, 2, 6641–6648
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
follows: 16 mg of ADAC and 0.6 g of RAMEB were added
together in 4 ml of water and stirred for 2 h at room tempera-
ture. Aer addition of 70 mg of H2SO4, the resulting solution
was inltrated into 0.5 g of SBA-15. Subsequently, the mixture
Hydrogenation catalytic tests
General procedure. The hydrogenation of methyl oleate has
been chosen as a model reaction and performed as follows: the
stainless steel autoclave was charged with 25 mg of carbon–
ruthenium catalyst. 587 mg of methyl oleate (2.07 mmol, 300
equivalents per mol of Ru) was added into the autoclave and
hydrogen was introduced at a constant pressure until the desired
ꢁ
underwent polymerization at 60 C for 24 h. The solid is then
subjected to inltration and polymerization again with 11 mg of
ADAC, 0.4 g of RAMEB and 46 mg of H2SO4 under the same
conditions as those previously described. The carbonization
ꢁ
value is reached (1 to 3 MPa). The mixture was heated to 30 C
ꢁ
was carried out at 900 C for 5 h under nitrogen ow and the
and stirred at 1250 rpm. The reaction was monitored by the
volume of consumed H2 and by analysing aliquots of the reaction
mixture using a Shimadzu GC-17A gas chromatograph, equipped
with a methyl silicone capillary column (30 m ꢂ 0.32 mm) and a
ame ionization detector. Each catalytic run was performed in
duplicate and the reported results are the average between the
two runs.
Reusability. For the recycling procedure, aer complete
conversion of methyl oleate, the catalyst was recovered by
ltration and thoroughly washed with heptane and diethyl
ether until complete elimination of the reaction product. Aer
removal of the remaining diethyl ether under vacuum, the solid
was reduced under H2 at 300 ꢁC reloaded with methyl oleate and
dihydrogen and reused in hydrogenation as described above.
subsequent silica removal was performed by HF etching.
In addition, another ruthenium-free sample has been
prepared by replacing the mixture of RAMEB and ADAC with
cyclodextrin alone. The synthesis principle was based again on a
two consecutive inltration procedure with 0.6 and 0.4 g of
RAMEB, respectively followed by thermo-polymerization at
60 ꢁC. The carbonization was carried out at 900 ꢁC for 5 h under
nitrogen ow and subsequent silica removal was performed by
HF etching. The synthesized carbon material prepared from
RAMEB is designated as MCA-4.
Characterization methods
Small-Angle X-ray Scattering (GISAXS-Rigaku S-max 3000
equipped with a microfocus source l ¼ 0.154 nm and a 2D
Gabriel type detector placed at 1472 mm from the sample)
was used to assess the synthesized material structure. The
powder was loaded into a 2 mm in diameter perforation made
on a 2 mm thick aluminum plate before being aligned into
the beam. The analysis was performed in vacuum. The d100
spacing values were calculated by the formula d100 ¼ 2p/q
and unit cell constants by a0 ¼ 2d100/O3. Transmission elec-
tron microscopy (TEM) images were recorded on a TECNAI
electron microscope operating at an accelerating voltage of
200 kV. The mesoporous carbon was deposited on a carbon
coated copper grid. The carbon nanorod thickness was esti-
mated using Digital Micrograph Soware. Nitrogen adsorp-
tion isotherms were measured at ꢀ196 ꢁC on a Nova 2200
apparatus from Quantachrom Corporation, aer having
degassed the sample under vacuum at 250 ꢁC. The specic
areas were calculated from the Brunauer–Emmett–Teller
(BET) equation using P/P0 values in the 2.5 ꢂ 10ꢀ2 and 2.0 ꢂ
10ꢀ1 range and the pore size distributions were obtained
from the adsorption branch of the isotherm assuming that
the pores are cylindrical using the Barrett–Joyner–Halenda
(BJH) method. The total pore volumes were estimated at
P/P0 ¼ 0.98 assuming that all the pores were lled with
condensed nitrogen in the normal liquid state. The micro-
pore volume is determined using the t-plot method. The
percentage of micropore volume in the synthesized materials
is calculated as (micropore volume/total pore volume) ꢂ 100.
Raman analyses were carried out at room temperature with a
Dilor XY-800 confocal micro-Raman spectrometer coupled to
a microscope. The spectrum was recorded in back scattering
geometry, using the 514.5 nm line of an Ar–Kr laser focused
on the sample using a 50 magnication (spot of 2 lm; power
of 20 mW). Raman cartography was performed using 625
spectra (30 ꢂ 40 mm) at a resolution of 1 mm2.
Results and discussion
Scheme 1 depicts the synthesis of the ruthenium carbon replica,
Ru@MCA-2, prepared from a host–guest assembly acting as
metal carrier and carbon source. The synthesis procedure
involves the lling of the silica pores with a host–guest solution
formed between the methylated cyclodextrin and ruthenium
adamantane followed by a polymerization step with sulfuric
acid at 60 ꢁC. The resultant product is carbonized at 900 ꢁC in an
inert atmosphere and nally treated with HF to remove the
silica hard-template.
The structure and crystallinity of the Ru@MCA-2 material
has been elucidated by the analysis of the SAXS pattern, which
shows an intense low-angle diffraction peak and two weak
peaks, assigned to the (100), (110) and (200) reections of a 2D-
hexagonal P6mm structure, respectively (Fig. 1). These features
are indicative of the fact that the ruthenium–carbon replica
Scheme 1 Schematic illustration of the synthesis of Ru@MCA-2 by
hard templating from host–guest complexes.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6641–6648 | 6643

View Article Online
Journal of Materials Chemistry A
Paper
carbonization and silica removal. Thus, the use of free and
complexed cyclodextrins as carbon sources is expected to lead to
relatively well-ordered structures of carbon nanorods because
the cyclodextrin molecules would easily form a condensed
structure during the gelation step due to their ability to pack in
head-to-head arrangements (channel structure).24 These results
suggest that the cyclodextrins get polymerized by preserving the
guest molecules inside the cavities, which are later carbonized
at higher temperatures. Interestingly, the process of carbon
formation involving the use of supramolecular assemblies can
be successfully applied to the case of metal–carbon composites
via the use of a suitable metal precursor, as evidenced by the
Ru@MCA-2 sample synthesized by inclusion complexation of
RAMEB and 1-adamantane carboxylic acid in the presence of
ruthenium trichloride. In addition, a comparison of the unit
cell parameters for the carbon replicas shows that the values are
12% lower than that for the SBA-15 template, which indicates a
slight decrease in the volume of channels within the silica
framework and structural shrinkage of the materials aer
Fig. 1 SAXS pattern of the synthesized materials: (a) SBA-15, (b)
Ru@MCA-2, (c) MCA-3 and (d) MCA-4.
thermal treatment at 900 C.26
ꢁ
preserves an ordered structure by retaining the hexagonal
symmetry of the SBA-15 silica template. It is worth noting that
the SAXS patterns obtained from the control carbon materials,
i.e. MCA-3 (RAMEB + ADAC) and MCA-4 (RAMEB) exhibit
similar reection planes, corroborating again that the carbons
are negative replicas of the SBA-15 without any marginal
structural disorder. These results are consistent with the
stabilization of a CMK-3 type mesoporous carbon whose struc-
ture is composed of carbon nanorods arranged in a hexagonal
pattern rigidly interconnected between each other.6
The unit cell parameters for the synthesized silica and
carbon materials calculated from the d100 spacing values are
provided in Table 1. It can be observed that the unit cell
constants are very similar for the three carbon materials with a
value of 9.5–9.6 nm, indicating that the polymerization of
RAMEB with or without adamantyl guests trapped inside the
cavity of the b-cyclodextrin does not result in any appreciable
change in the ordered structure of the carbon replicas aer
As can be seen from Fig. 2, the N2 adsorption–desorption
isotherms for all samples are characteristic of a type IV isotherm
of mesoporous materials with a hysteresis loop featuring the
capillary condensation at intermediate P/P0 pressures: ꢃ0.55–
0.85 for SBA-15 vs. ꢃ0.4–0.85 for the synthesized carbon mate-
rials. It is apparent, from the pore size distributions (PSD) that
only one type of pores in the carbon replicas is distinguished,
i.e. the pores formed by removal of the silica wall and conned
between the hexagonally packed carbon rods. This conrms
that the resulting carbon replicas derived from RAMEB are
inverse replicas of the silica framework by maintaining a high
degree of pore-size uniformity with, in addition, a high surface
area and pore volume (Table 1). Thus, it can be noted that the
textural features of MCA-3 (RAMEB + ADAC) are almost the
same as those of MCA-4 (RAMEB), in terms of specic surface
area (ꢃ940–960 m2 gꢀ1), total pore volume (ꢃ1.0–1.1 cm3 gꢀ1
)
and pore diameter (4.2 nm). It is remarkable that the overall
contribution of micropores is quite low in all these materials
(<5%). This result could be related to the fact that RAMEB is an
oligosaccharide with a relatively low oxygen-to-carbon ratio of
0.6 (molecular formula: C55H95O35) and that its conversion into
carbon by polymerization and carbonization at 900 ꢁC does not
develop large amounts of micropores within the carbon nano-
rods. Indeed, it is generally accepted that CMK-3 type materials
produced from the common precursor sucrose (O/C ¼ 0.9)‡ lead
to the formation of additional micropores and disordered
carbon networks, responsible for the increase in the specic
surface areas (ꢃ1300–1500 m2 gꢀ1).1–3 Finally, the features
obtained in this work can be viewed as an indication that
relatively ordered carbon structures can be directly obtained
from cyclodextrin inclusion complexes because the guest ada-
mantane molecules trapped in the host cavity of RAMEB would
Table 1 Structural and textural properties of the mesoporous carbons
prepared from RAMEB and starting silica hard-template
VPd (cm3 gꢀ1)
a
b
c
g
d100
a0
SBET
DP
(nm)
Sample
SBA-15
Ru@MCA-2 8.2
MCA-3
MCA-4
(nm) (nm) (m2 gꢀ1
)
Vmeso Vmicro
%
e
f
9.3
10.8
9.5
9.6
9.5
771
945
959
945
1.016 0.005
0.899 0.041
1.006 0.034
1.103 0.027
0.49 7.9
4.36 3.8
3.37 4.2
2.37 4.2
8.3
8.2
a
d-spacing calculated on the basis of the following formulas nl ¼
b
2d100 sin q and d100 ¼ 2p/q. Unit cell constant calculated by the
c
equation a0 ¼ 2d100/O3. Specic surface area calculated from the
Brunauer–Emmett–Teller equation in the P/P0 range of 0.025–0.20.
d
Determined at P/P0 ¼ 0.98. Total pore volume estimated at P/P0
¼
e
0.98.
Micropore volume determined by the t-plot method.
f
‡ CMK-3 was prepared from sucrose (molecular formula: C12H22O11) by a two
consecutive inltration procedure, involving a thermo-polymerization reaction
at 100 and 160 ꢁC. The carbonization was carried out at 900 ꢁC for 5 h under
nitrogen ow and the silica removal was performed by HF etching.6
Percentage of micropore calculated as the ratio of micropore volume
g
to total pore volume ꢂ 100. Calculated by the Barrett–Joyner–
Halenda method from the adsorption branch of the isotherm
assuming a cylindrical shape.
6644 | J. Mater. Chem. A, 2014, 2, 6641–6648
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
Fig. 2 N2 adsorption–desorption isotherms of the synthesized materials and their corresponding BJH pore size distributions: (a) SBA-15, (b)
Ru@MCA-2, (c) MCA-3 and (d) MCA-4.
not signicantly affect the process of carbon nanorod forma- 6 nm (Fig. 3e), which is somewhat smaller than that of the pore
tion. However, in the case of Ru@MCA-2, a little alteration in diameter of the SBA-15 silica template due to the shrinkage
the porous structure is encountered and this can be attributed during the thermal treatment. In addition, it can be seen that,
to the incorporation of ruthenium nanoparticles within the for this composite material prepared by exploiting the
carbon matrix.
complexation interactions of b-cyclodextrin and the ada-
In addition, we observe that the pore sizes of the carbon mantane derivative, ruthenium particles with an average size of
replicas are larger by approximately 0.5–0.9 nm than that of the ca. 1.5 nm are uniformly dispersed over the carbon nanorods.
pore wall thickness of the starting SBA-15 (3.3 nm). The devia- As shown in Fig. 3h, the histogram displays a narrow-size
tion may be related to the aforementioned structural shrinkage distribution, with more than 95% of the nanoparticles between
during the carbonization process. The structural characteriza- 1 and 2 nm in size. No agglomeration of nanoparticles was
tion of the synthesized carbon materials prepared from host– observed. It can be further noted that Ru@MCA-2 has also been
guest assemblies has been further explored by TEM analysis analyzed by wide angle XRD to support the absence of large
(Fig. 3). Fig. 3a and b show the TEM images of the MCA-3 aggregates in the sample (Fig. S1, ESI†). Thus, no reection
sample (RAMEB + ADAC) viewed perpendicular and parallel to peaks of ruthenium in the form of metallic and/or oxide phase
the carbon rods, respectively. Thus, we observe that MCA-3 emerge from the XRD pattern, further indicating that ruthe-
forms a periodic arrangement of interconnected lled carbon nium forms very small particles within the carbon matrix,
nanorods, with a parallel alignment between mesoporous which cannot be detected by wide angle XRD analysis.31,32
channels on a long range order, i.e. more than 600 nm long
Taken together, these results suggest that the synthesis
(Fig. 3a) and a honeycomb structure (Fig. 3b). This observation strategy presented in this study enables to distribute homoge-
corroborates the fact that MCA-3 has a 2D-hexagonal structure neously the ruthenium ions within the cyclodextrin–adamantane
similar to that in a CMK-3 structure (P6mm). The structure and network (itself conned to the pores of the silica template) and
metal dispersion of the Ru@MCA-2 sample was also veried by maintain the ruthenium dispersion in the subsequent steps of
TEM measurements (Fig. 3c to g). Like MCA-3, the Ru@MCA-2 polymerization, carbonization and HF etching. In addition, it is
sample exhibits a uniform array of mesopores with a long range worth noting that the ruthenium content has been determined in
order. The average carbon rod thickness is estimated to be ca. two different batches of Ru@MCA-2 samples by elemental
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6641–6648 | 6645

View Article Online
Journal of Materials Chemistry A
Paper
Fig. 4 (a) Raman spectrum of the Ru@MCA-2 sample; (b) Raman
cartography calculated from the intensity ratio of IG to ID at 1595 and
1310 cmꢀ1, respectively (625 pixels).
commercial carbon materials5 (activated carbon, carbon black)
and synthesized ordered mesoporous carbon materials, such as
CMK-3,33 CMT-1,34 CMK-3-MPc35 and OMCs-SGW.36 A two-
dimensional Raman cartography (30 ꢂ 40 mm) at a resolution of
1 mm2 was further performed to establish the spatial distribu-
tion of the two bands by calculating the ratio of the peak heights
of 1595 to 1310 cmꢀ1. The result of the mapping recorded on
Ru@MCA-2 is outlined in Fig. 4b and allows highlighting the
Fig. 3 TEM images of the carbon materials prepared from cyclodex-
trin inclusion complexes at different magnifications: (a) MCA-3 at
100 nm; (b) MCA-3 at 20 nm; (c) and (d) Ru@MCA-2 at 100 nm; (e)–(g)
Ru@MCA-2 at 20 nm and (h) the corresponding particle size distri- homogeneity of this sample on a larger scale than that usually
bution of Ru@MCA-2 obtained from the measurement of ca. 200
particles.
studied with the standard Raman microscope measurement.
Thus, we observe that the IG/ID ratio is mostly in the range of
0.75 to 0.90 and that ca. 3% of the pixels are less than or equal to
the threshold value of 0.6.
To illustrate the interest of the strategy of synthesis, the
activity of Ru@MCA-2 was investigated in the hydrogenation of
analysis (SCA, CNRS) and the metal loading was 2.7 ꢄ 0.03 wt%.
Interestingly, this value is relatively close to that which would
be expected by considering the complete transformation of
both organic precursors into carbon without the loss of metal
(ꢃ2.1 wt%). These results are further evidence of the efficiency of
Table
2 Catalytic performances of methyl oleate hydrogenation
the synthesis method that allows easy control of the metal
loading and size of the metallic nanoparticles in the nal
structure.
catalyzed by Ru catalysts
The synthesized mesoporous carbon materials were also
investigated by Raman spectroscopy to study their graphitic
character (Fig. 4a and S2, ESI†). All Raman spectra reveal a set of
two bands at 1310 and 1595 cmꢀ1, assigned to defects (D band)
and graphitic structures (G band) in carbon materials, respec-
tively. The formation of some graphitic domains within the
carbon networks prepared from cyclodextrins and inclusion
complexes is evidenced by the presence of the narrow band
ascribed to the stretching vibration modes of C]C bonds of
typical graphite layers at 1595 cmꢀ1. The other band at 1300
cmꢀ1 is associated with the vibration of carbon atoms in plane
terminations of disordered graphite. The IG/ID ratio of
Ru@MCA-2, MCA-3 and MCA-4 is calculated to be 0.83, 0.84 and
0.87, respectively and these values indicate the relatively good
graphitic nature of the carbon framework compared to other
Sample
H2 (MPa)
Conv.a (%)
Timeb (h)
TOFc (hꢀ1
)
Ru@MCA-2
Ru@MCA-2
Ru@MCA-2
Ru/AC-WV
Ru/AC-NorA
1
2
3
3
3
5d
15d
74
—
18
1
3
—
—
17
400
100
—
20
4d
a
b
c
Conversion aer 0.5 h. 100% conversion. Turnover frequency
dened as the number of mol of substrate per mol of ruthenium per
d
h. Conversion aer 3 h. Reaction conditions: catalyst (2.7 wt% Ru),
25 mg; Ru, 6.9 mmol; methyl oleate, 2.08 mmol; substrate/Ru (mol
molꢀ1), 300; heptane, 10 g; H2, 1–3 MPa; stirring rate, 1250 rpm;
temperature, 30 ꢁC.
6646 | J. Mater. Chem. A, 2014, 2, 6641–6648
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry A
methyl oleate, chosen as a model molecule of unsaturated fatty
The catalytic results also suggest that Ru@MCA-2 has the
acid methyl esters (Table 2). Hydrogenation of fatty oils and advantage of combining a high surface area and uniform pore
fatty acid methyl esters is one of the most important processes size within the hexagonal arrangement of the carbon nanorods.
in edible oil and oleochemical industries in order to make the These textural features are important factors for the diffusion of
products more resistant against air autoxidation, thermal the bulky hydrophobic substrate, i.e. methyl oleate, into the
decomposition, and other reactions that affect the avor. porous network of the catalyst prepared by hard-templat-
Catalytic tests have been performed in heptane at 30 C under ing.3,37,38 In contrast, the extremely low activity obtained with
ꢁ
various pressures of H2 ranging from 1 to 3 MPa and with a ratio Ru/AC-NorA, can be related to the predominant microporous
of substrate to ruthenium of 300. The reaction kinetic is structure of the activated carbon that provides steric hindrance
monitored by gas chromatography (procedure in the Experi- to the long-chain fatty acid methyl ester molecules and prevents
mental section).
their accessibility onto metal active sites. Finally, the reusability
For comparison, the results obtained with two control of Ru@MCA-2 has been evaluated on methyl oleate at 3 MPa H2
ꢁ
ruthenium carbon-supported catalysts containing the same and 30 C (Fig. 5). We observe that the Ru@MCA-2 catalyst is
metal loading as that of Ru@MCA-2 (2.7 wt% Ru) have been stable and reusable under the reaction conditions, preserving
included. Briey, these control materials were prepared by wet its activity aer 4 consecutive runs.
impregnation of RuCl3 followed by calcination under air at
ꢁ
ꢁ
250 C and reduction at 300 C under H2 (20 vol% in N2). Two
commercial activated carbons were selected to support the
Conclusions
ruthenium active phase, i.e., AC-WV (1690 m2 gꢀ1) and AC-NorA In conclusion, we have described a one-pot method for the
(1207 m2 gꢀ1). The corresponding textural parameters and preparation of a P6mm ordered mesoporous carbon having
surface titration are given in Table S1 (ESI†). Compared to uniform ruthenium nanoparticles ranging from 1 to 2 nm. The
AC-NorA, it is shown that AC-WV exhibits a higher surface area specicity of the method was based on the ability of methylated
(1690 vs. 1207 m2 gꢀ1), a larger average pore size (3.1 vs. 2.4 nm) b-cyclodextrin to act as both a polymerizable carbon source and
and a lower percentage of microporosity (41 vs. 60%).
a host molecule for favoring the inclusion of a carboxylic acid
The results presented in Table 2 reveal that the activity of derivative and the dispersion of the ruthenium precursor
Ru@MCA-2 is highly dependent on the pressure of H2, the during the pore lling and gelation steps. The ruthenium
higher the pressure, the higher the activity. Thus, the highest dispersion could be maintained without any sign of aggregation
turnover frequency of the Ru@MCA-2 catalyst for the hydro- aer the thermal structuring of the carbon phase at 900 ꢁC and
genation of methyl oleate reaches ca. 400 hꢀ1 at 3 MPa and silica removal. Finally, the nanocasting strategy involving the
notably this value is 4 times higher than that of control direct use of host–guest complexes could contribute to provide a
Ru/AC-WV. As the conversion of methyl oleate was negligible new generation of robust catalysts with high metal dispersions
in the presence of Ru/AC-NorA aer 3 hours of reaction, we and stability while ensuring the formation of ordered meso-
do not extend the time to complete the reaction. From these porous carbons of high specic surface areas and pore volumes.
results, it can be assumed that the Ru@MCA-2 catalyst These features are known to be key factors to ensure accessi-
prepared by the cyclodextrin approach has a more exposed bility to the active sites and efficient mass transports in
metallic surface thanks to a high dispersion of small Ru heterogeneous catalytic processes.
nanoparticles whereas the carbon-supported ruthenium
catalysts prepared by wet impregnation can more easily
undergo aggregation during the thermal treatments and
Acknowledgements
reaction course.
Dr N. Gokulakrishnan is thankful for nancial support by the
Institut de Recherche en Environnement Industriel. The TEM
facility in Lille (France) is supported by the Conseil Regional du
Nord-Pas de Calais and the European Regional Development
`
Fund (ERDF). The SAXS facility in the College de France is
supported by the CBRSn, the UPMC and Cnano Ile de France.
The authors gratefully acknowledge the National Agency of
Research (CYCLOMAT no. ANR-11-NANO-005) for funding.
Notes and references
1 R. Ryoo, S. H. Joo and S. Jun, J. Phys. Chem. B, 1999, 103,
7743.
2 T. Y. Ma, L. Liu and Y. Z. Yuan, Chem. Soc. Rev., 2013, 42,
3977.
Fig. 5 Reusability of the catalytic system in the hydrogenation of
methyl oleate. Reaction conditions: catalyst Ru@MCA-2 (2.7 wt% Ru);
methyl oleate/Ru (mol molꢀ1), 300; heptane, 10 g; H2, 3 MPa; stirring
rate, 1250 rpm; temperature, 30 ꢁC.
´
3 N. Gokulakrishnan, B. Leger, C. Lancelot, S. Fourmentin,
D. Grosso, E. Monier and A. Ponchel, Carbon, 2011, 49,
1290.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. A, 2014, 2, 6641–6648 | 6647

View Article Online
Journal of Materials Chemistry A
Paper
4 J. Lee, J. Kim and T. Hyeon, Adv. Mater., 2006, 18, 2073.
5 T. W. Kim, I. S. Park and R. Ryoo, Angew. Chem., Int. Ed.,
2003, 42, 4375.
22 B. H. Han, W. Zhou and A. Sayari, J. Am. Chem. Soc., 2003,
125, 3444.
23 W. Shen, X. Yang, Q. Guo, Y. Liu, Y. Song, Z. Han, Q. Sun and
J. Cheng, Mater. Lett., 2006, 60, 3517.
6 S. Jun, S. H. Joo, R. Ryoo, M. Kruk, M. Jaroniec, Z. Liu,
T. Ohsuna and O. Terasaki, J. Am. Chem. Soc., 2000, 122, 24 H. C. Wang, B. L. Li, J. T. Li, P. Lin, X. B. Bian, J. Li, B. Zhang
10712. and Z. X. Wan, Appl. Surf. Sci., 2011, 257, 4325.
7 C. H. Kim, D. K. Lee and T. J. Pinnavaia, Langmuir, 2004, 20, 25 W. C. Cromwell, K. Bystrom and M. R. Eink, J. Phys. Chem.,
5157. 1985, 89, 326.
8 R. Narayanan and M. A. El-Sayed, Langmuir, 2005, 21, 2027. 26 R. I. Gelb and L. M. Schwartz, J. Inclusion Phenom. Mol.
9 R. White, R. Luque, V. L. Budarin, J. H. Clark and
D. J. Macquarrie, Chem. Soc. Rev., 2009, 38, 481.
Recognit. Chem., 1989, 7, 537.
27 R. Palepu and V. C. Reinsborough, Aust. J. Chem., 1990, 43,
10 S. Ikeda, S. Ishino, T. Harada, N. Okamoto, T. Sakata,
2119.
H. Mori, S. Kuwabata, T. Torimoto and M. Matsumura, 28 E. S. Kwak and F. A. Gomez, Chromatographia, 1996, 43, 659.
Angew. Chem., Int. Ed, 2006, 118, 7221.
11 R. Narayanan and M. A. El-Sayed, J. Catal., 2005, 234, 348.
12 J. M. Nadgeri, M. M. Telkar and C. V. Rode, Catal. Commun.,
2008, 9, 441.
13 R. Luque, J. H. Clark, K. Yoshida and P. L. Gai, Chem.
Commun., 2009, 5305.
29 M. Hartmann and A. Vinu, Langmuir, 2002, 18, 8010.
30 M. Kruk, M. Jaroniec, Y. Sakamoto, O. Terasaki, R. Ryoo and
C. H. Ko, J. Phys. Chem. B, 2000, 104, 292.
31 R. W. G. Wyckoff, in Crystal Structures, Interscience
Publishers, New York, 2nd edn, 1963, vol. 1, pp. 7–83.
´
32 J. Haines, J. M. Leger, O. Schulte and S. Hull, Acta
14 Y. Li, G. Lan, H. Wang, H. Tang, X. Yan and H. Liu, Catal.
Commun., 2012, 20, 29.
15 J. Szejtli, Chem. Rev., 1998, 98, 1743.
16 R. Breslow and S. D. Dong, Chem. Rev., 1998, 98, 1997.
17 F. Hapiot, S. Tilloy and E. Monier, Chem. Rev., 2006, 106,
767.
18 F. Hapiot, A. Ponchel, S. Tilloy and E. Monier, C. R. Chim.,
2011, 14, 149.
Crystallogr., Sect. A: Found. Crystallogr., 1997, 53, 880.
33 L. Shen, E. Uchaker, C. Yuan, P. Nie, M. Zhang, X. Zhang and
G. Cao, ACS Appl. Mater. Interfaces, 2012, 4, 2985.
34 Y. Lo, S. J. Huang, W. H. Chen, Y. R. Peng, C. T. Kuo and
S. B. Liu, Thin Solid Films, 2006, 498, 193.
35 K. T. Lee, X. Ji, M. Rault and L. F. Nazar, Angew. Chem., Int.
Ed., 2009, 48, 5661.
36 Z. Wu, Y. Yang, D. Gu, Y. Zhat, D. Feng, Q. Li, B. Tu,
P. A. Webley and D. Y. Zhao, Top. Catal., 2009, 52, 12.
19 L. X. Song, L. Bai, X. M. Xu, J. H. He and S. Z. Pan, Coord.
Chem. Rev., 2009, 253, 1276.
´
37 N. Kania, B. Leger, S. Fourmentin, E. Monier and
20 S. Polarz, B. Smarsly, L. Bronstein and M. Antonietti, Angew.
Chem., Int. Ed., 2001, 40, 4417.
A. Ponchel, Chem.–Eur. J., 2010, 16, 6138.
´
38 N. Kania, N. Gokulakrishnan, B. Leger, S. Fourmentin,
21 B. H. Han and M. Antonietti, Chem. Mater., 2002, 14, 3477.
E. Monier and A. Ponchel, J. Catal., 2011, 278, 208.
6648 | J. Mater. Chem. A, 2014, 2, 6641–6648
This journal is © The Royal Society of Chemistry 2014