Organosilica-metallic sandwich materials as precursors for palladium and platinum nanoparticle synthesis

Article abstract of DOI:10.1039/c5ra15871k

A promising strategy to design highly active catalysts is to use multicomponent catalysts including active metals (such as platinoids) and organosilica as a substrate. In this article, we develop stable OrganoSilica Sandwiches (OSS) consisting of sandwiched metallic layers between organosilica layers. The formation of such a sandwich-like organosilica@platinoid@organosilica material was obtained from bis-silica-alkylammonium carbamate dimers, acting as a template for the immobilization of platinoid salt complexes or metallic nanoparticles in the interlayer space after a reduction treatment. The layered templates and the sandwich structures were investigated by small angle X-ray scattering, coupled thermogravimetric analysis and mass spectrometry, and Raman spectroscopy. It was demonstrated that the meso-structure of the initial template material is not affected by the chemical steps involved in the synthesis of the material and does not depend on the incorporated metal.

Full text of DOI:10.1039/c5ra15871k

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Organosilica-metallic sandwich materials as
precursors for palladium and platinum nanoparticle
Cite this: RSC Adv., 2015, 5, 77619
synthesis†
*
Romain Besnard, Julien Cambedouzou, Guilhem Arrachart, Xavier F. Le Goff
*
and Stephane Pellet-Rostaing
´
A promising strategy to design highly active catalysts is to use multicomponent catalysts including active
metals (such as platinoids) and organosilica as a substrate. In this article, we develop stable OrganoSilica
Sandwiches (OSS) consisting of sandwiched metallic layers between organosilica layers. The formation of
such
a sandwich-like organosilica@platinoid@organosilica material was obtained from bis-silica-
alkylammonium carbamate dimers, acting as a template for the immobilization of platinoid salt
complexes or metallic nanoparticles in the interlayer space after a reduction treatment. The layered
templates and the sandwich structures were investigated by small angle X-ray scattering, coupled
thermogravimetric analysis and mass spectrometry, and Raman spectroscopy. It was demonstrated that
the meso-structure of the initial template material is not affected by the chemical steps involved in the
synthesis of the material and does not depend on the incorporated metal.
Received 7th August 2015
Accepted 8th September 2015
DOI: 10.1039/c5ra15871k
www.rsc.org/advances
molecules indeed offer most of the requirements sought for the
formation of hybrid organic–inorganic materials, namely a
1. Introduction
condensable part, and terminal chelating sites with metal
adsorption abilities. To get optimal properties, the organically
functionalized silica should present a high density of functional
groups with a satisfying homogeneity and accessibility. These
requirements are usually difficult to obtain with conventional
approaches, which consist in post-graing a functional alkox-
ysilane on pre-synthesized mesoporous silica, due to the
blocking effect at the entrance of the pores.35 The functionali-
zation by co-condensation is not ideal either recommended
because the obtained functional density is weaker and the
mesostructure is less dened.
To avoid such drawbacks, an opportunity is to use a bisily-
lated compound for the preparation of periodic mesoporous
organosilica (PMOs), in which organic bridges are incorporated
in the three-dimensional network structure of the silica
matrix.36–38 One limitation of this approach may be linked to the
weak accessibility of the organic fragments incorporated into
the walls, which can be a concern in the eld of solid phase
extraction. On the other hand, the use of amphiphilic organo-
silane molecules could be an alternative route in the prepara-
tion of a structured functional silica material presenting a high
organic function density, as already proposed in a few
studies.39–41
Mesostructured hybrid silicas exhibit singular properties due to
the combination of organic–inorganic properties and quantum
connement effects.1 Tailoring these materials for specic
applications, particularly metal adsorption2,3 and catalysis4,5
requires both compositional and structural control. In partic-
ular, materials presenting graed organic functions able to
selectively complex ions are of high interest in the eld of solid/
liquid extraction, liquid waste processing and metal immobili-
zation.6,7 Among noble metals, platinoids are well known for
their catalytic properties depending on their oxidation state and
their crystallographic structure.8–14 Noble metal NPs are usually
prepared by so templating approaches, using surfactants and
organic solvents.15–20 In these approaches, the shape and size of
nanoparticles is driven by the morphology of the organic
micelles containing the metallic salt. The use of porous mate-
rials as hard templates has been also described.21–34 The asset of
such a method is the better control of the size and the shape of
the particles, in particular for the preparation of nanorods.33
The use of organosilane precursors is well suited to the
preparation of hybrid silica with complexation abilities. Such
´
Institut de Chimie Separative de Marcoule, UMR 5257 CEA/CNRS/UM/ENSCM,
`
BP17171, F-30207 Bagnols-sur-Ceze, France. E-mail: stephane.pellet-rostaing@cea.fr
The organization is specically driven by the supramolecular
assembly process from the organosilane precursors, long-chain
alkyl-siloxane with or without self-assembling cross-linker are
of particular interest for the design of lamellar bilayer nano-
structure.42–44 These systems have recently been reviewed.45
† Electronic supplementary information (ESI) available: Fig. S1: in situ SAXS
experiment of M1 and M1-Pt(IV) materials at 25 ꢀC and 120 ꢀC; Fig. S2: SAXS
patterns of M1-Pd(0) and M1-Pt(0) in the range of 25 nmꢁ1 to 31 nmꢁ1 and the
corresponding ts of structure peak; Fig. S3: TEM images of M1-Pt(0) material.
See DOI: 10.1039/c5ra15871k
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 77619–77628 | 77619

View Article Online
RSC Advances
Paper
Thanks to the possibilities offered by this approach, amino- deposited on 400-mesh carbon-coated copper grids. Aer
undecyl-triethoxysilane, which can be considered as an ethanol evaporation, the grid was submitted to analysis.
amphiphilic organosilane, has been used for the preparation of
SEM images were obtained from a FEI QUANTA 200 ESEM
mesostructured materials starting from their self-assembly into FEG operating at 15 kV equipped with a Everhart-Thornley
reverse cylindrical micelles or lamellar arrangements.46,47 detector. X-EDS analyses were performed using a Bruker AXS
Furthermore, the self-assembly of the amino-undecyl- X-Flash 5010 detector coupled with SEM to quantitatively
triethoxysilane in pentane or in water with CO2 as assembly determine the elementary atomic percentages.
agent led to the formation of bis-silylated-alkylammonium
carbamate dimers arranged in a lamellar structure.46,48
TGA analyses were performed on a Setaram Setsys Evolution
18 instrument under air ow (20 mL minꢁ1) with a heating rate
The removal of CO2 from ammonium carbamate groups in of 2 ꢀC minꢁ1. The coupled mass spectrometer was a Hiden
such materials can easily offered available amino group by analytical QGA instrument.
heating at moderate temperature.46,48 The amino group is one of
Elemental analyses were performed by the Service Central
the most attractive surface functionality owing to its use as d'Analyse (CNRS, Vernaison, France).
chelating site for various metal ions. Such affinities for metal
Raman spectra were recorded on a Horiba Jobin Yvon Lab-
anions are particularly increased in acidic aqueous conditions RAM Aramis confocal Raman microscope, using an excitation
due to the formation of ammonium-metallic salts.
wavelength of l ¼ 532 nm with a laser spot size of ꢂ1 mm, and
In the present study, we used a similar methodology to with an objective of ꢃ50 long working distance. The incident
incorporate palladium and platinum chloride species at the laser power was kept lower than 0.8 mW in order to avoid any
interface between two bilayers obtained aer condensation of heating or structural sample damage.
silylated amines. The lamellar hybrid material acts as a
template for the immobilization in the interlayer space of the
platinoid salt, resulting in a stable sandwich-layered structure
of organosilica@platinoid@organosilica also called Organo-
Silica Sandwich (OSS). The formation of such a sandwich-like
material was investigated by Small angle X-ray scattering
(SAXS) and microscopy; the exact composition was determined
from complementary analyses (elemental analysis, X-EDS,
Raman spectroscopy and thermal measurements).
Platinoid salts located inside the impregnated material have
been further reduced by thermal treatments without damaging
the initial lamellar structure, offering a new approach for the
preparation of platinoids nanoparticles based on hybrid silica
templating. The combination of microscopy and X-ray scat-
tering techniques allowed us to follow the formation of the
crystalline nanoparticles and to highlight unexpected effects on
2.2. Synthesis of amino-undecyl-triethoxysilane
The organic precursor was synthesized in three steps in anhy-
drous conditions and adapted from the synthesis described in
the literature.50,51 Bromoundecene (Alpha Aesar) was rst mixed
with triethoxysilane and Karstedt catalyst both purchased from
ABCR to obtain bromoundecyl-triethoxysilane. Aer purica-
tion, the organosilane bromide was then condensed with
sodium azide (Sigma Aldrich) in acetonitrile to give the expected
azidoundecyl-triethoxysilane. Finally, the reduction of the azido
group with Palladium (10% Pd/C) from Sigma Aldrich in
anhydrous ethanol under hydrogen atmosphere led to amino-
undecyl-triethoxysilane.
2.3. Material syntheses
the shape, the size and the distribution of nanoparticles The procedure for preparing the initial material containing a
according to the metal immobilized in the material.
carbamate ammonium salt was adapted from the procedure
described by Alauzun et al.46 It was prepared as followed:
3 mmol of amino-undecyl-triethoxysilane (the amphiphilic
precursor) were introduced in a 50 mL ask under pure CO2 gas
affording a gel. Aer 30 min, 20 mL of pentane were poured in
2. Experimental section
2.1. Technical details
SAXS experiments were performed on a set-up operating in the ask and 81 mL of a hydrochloric acid solution was added at
transmission geometry. A Mo anode associated to a Fox2D pH 1.5. The suspension was stirred for 3 days at 30 ꢀC. The
multi-shell mirror (XENOCS) delivers a collimated beam of resulting material (noted M1) was obtained aer ltration. The
wavelength 0.710 A. Two sets of scatterless slits49 delimitate the solid was washed with ethanol and ether before drying under
˚
beam to a squared section of side length 0.8 mm. A MAR345 vacuum for 1 hour.
imaging plate detector allows simultaneously recording scat-
The impregnations of platinoid salts in the materials were
tering vectors q ranging from 0.25 nmꢁ1 up to 25 nmꢁ1. Samples performed in aqueous solution. 204 mg of platinic(IV) chloride
were placed in glass capillaries of 2 mm diameter. Absolute hexahydrate (H2PtCl6$6H2O purchased from Aldrich) were dis-
intensities were obtained by measuring a calibration sample of solved in 100 mL of water. The colour of the solution became
high density polyethylene (Goodfellow) for which the absolute dark-yellow. 70 mg of Palladium(II) chloride (PdCl2 purchased
scattering was already determined. All SAXS proles are plotted from Aldrich) were dissolved in 100 mL of a hydrochloric acid
in log–log scale in order to highlight the dependency of the solution (pH ¼ 1), resulting in a coloured brownish-red solu-
intensity versus a power law of the scattering vector q.
tion. Then, 100 mg of M1 material were added in each solution.
ꢀ
TEM was carried out at 200 kV on a JEOL 2200 FS micro- The resulting mixtures were heated at 70 C for 24 h to release
scope. The samples obtained aer heat treatment were ground the CO2. Then the resulting solids were ltered and washed with
into a powder and dispersed in ethanol. Samples were ethanol and ether. Aer drying under vacuum for 1 hour, the
77620 | RSC Adv., 2015, 5, 77619–77628
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
materials incorporating the platinum salt (noted M1-Pt(IV)) and
the palladium salt (noted M1-Pd(II)) were obtained as yellowish
and brownish red powders, respectively.
The reduction of impregnated materials was performed
through an in situ TGA experiment under 5% H2 in Argon ow
(20 mL minꢁ1) from 20 ꢀC to 140 ꢀC. The resulting materials
M1-Pt(0) and M1-Pd(0) became respectively dark grey and black.
3. Result and discussion
3.1. Characterization of impregnated materials
The incorporation of platinoid salts has been performed thanks
to the release of CO2 from the pre-formed silica-
alkylammonium carbamate M1 material. In order to evaluate
the effects of platinoid impregnation in the material, SEM
images have been performed before and aer impregnation.
The morphology and size of the materials were preserved aer
the impregnation of the platinoid salt solution in the M1
material according to the SEM images (Fig. 1). The homogeneity
of the impregnation was appreciated thanks to the X-EDS
analysis. The nal composition of the materials will be dis-
cussed later.
The SAXS prole of the initial M1 material (Fig. 2a) can be
described by a lamellar meso-structure, where bilayers of bis-
aminoundecylsilica are bridged by ammonium carbamate
group (Fig. 3), and as previously described.42 The lattice
parameter of this lamellar structure can be deduced from the
position of the Bragg peaks relative to this lamellar phase
(below 10 nmꢁ1) and is estimated at around 3.4 nm. The weak
intensity of the rst peak indicates that the electronic contrast
between the silica matrix and the ammonium carbamate group
is rather low (Fig. 3).
Fig. 2 SAXS profiles of (a) M1, (b) M1-Pd(II) and (c) M1-Pt(IV). The
squares and the black rings in the inset correspond respectively to the
supplemental structure peaks at wide angles of (b) M1-Pd(II) and (c)
M1-Pt(IV).
The largest feature observable at around 15 nmꢁ1 (typical
distance of 0.42 nm in the real space) can be related to the
typical mean lateral distance between two precursors.47
Fig. 3 Schematic representation of the lamellar structure with carbon
dioxide or metallic salt at the interface between two organosilica
bilayers. The value of the electronic density r of each part of the
lamellar structure is also represented. Inset: schematic representation
2ꢁ
of the crystalline monolayer phase of PdCl2 (composed of PdCl4
units) at the interface between two organosilica bilayers in the
M1-Pd(II) material.
Fig. 1 SEM images of the M1, M1-Pd(II) and M1-Pt(IV) materials.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 77619–77628 | 77621

View Article Online
RSC Advances
Paper
16.7 nmꢁ1, 23.7 nmꢁ1, and 26.2 nmꢁ1. As mentioned in the
works of Krishnaswamy et al. in 2006, dealing with a so crys-
talline
surfactant–Pt(II)
complex
involving
cetyl-
trimethylammonium bromide (CTAB) as surfactant and
chloroplatinate as metallic salt,15 the formation of a precipitate in
a primitive orthorhombic lattice of space group P2221 occurred.
The lattice parameters of this Pt(II)–CTAB complex are a ¼
0.753 nm, b ¼ 0.797 nm, c ¼ 3.956 nm. The c direction is related
to the extended alkyl chains of CTAB molecules, while the a and
b direction correspond to distances between Pt atoms. Taking
into account the similarity of this system with ours, and
provided that the chemical properties of Pd(II) and Pt(II) are
close, it appears that such an orthorhombic structure can be
considered in order to describe the M1-Pd(II) material. In our
case, the material involves silylated-undecylammonium layers
instead of CTAB layers. The c lattice parameter should be taken
equal to 3.14 nm as determined from the Bragg peak positions
indexed in the lamellar structure described above. If the same
lattice parameters are considered for the a and b direction in
this orthorhombic structure, the Bragg reections on the planes
(010), (011), (100), and (101) should appear between 7.9 and
8.6 nmꢁ1. The broad Bragg reection on the planes (004) of the
lamellar structure does not allow checking the presence of these
peaks. However, such a structure should lead to higher order
Bragg reections at 8.85, 9.25, 9.91 and 10.27 nmꢁ1, corre-
sponding respectively to the reections on the planes (012),
(102), (013), and (103) (Fig. 4a).
Fig. 4 Calculation of (a) the orthorhombic structure and (b) the
structure tetragonal structure compared to the M1-Pd(II) SAXS profile.
The red indexes exhibit Bragg reflections whose indexes h and k are of
different parity in the case of the orthorhombic structure.
Such peaks do not appear on the SAXS prole. It could be due
to the absence of long range correlation in the c direction.
Nevertheless, the Bragg reections on the planes (120) and
(210), expected respectively at 17.84 and 18.45 nmꢁ1 are also
missing, involving that the hypothetic structure corresponding
to a primitive orthorhombic structure is not likely. A face
The SAXS patterns of both M1-Pd(II) and M1-Pt(IV) powders
present a lamellar structure with nearly the same lattice
parameter (Fig. 2b and c). Therefore, no damage of the lamellar
structure is observed due to the incorporation of platinoid salts.
Furthermore, the relative intensity of the rst peak of the
lamellar structure (located at around 2 nmꢁ1) increases aer
impregnation, especially for the M1-Pt(IV) material.
This suggests that the electronic contrast between the silica
matrix and the interface between two bilayers increases.48
Therefore, the structural hypothesis of stable sandwich-
layered structures consisting in the repetition of stacked sand-
wiches involving a metal layer between two organosilica layers is
likely.
ꢀ
In situ SAXS experiments from room temperature to 120 C
have been carried out on the M1 and the M1-Pt(IV) materials
(Fig. S1†). From 70 ꢀC, the lamellar structure of M1 material is
damaged by CO2 removal as recently described in a previous
study.48 On the contrary, the lamellar structure of M1-Pt(IV) is
maintained even aer a similar thermal treatment. This
observation is a supplemental indication of the location of
platinoid salts inside the lamellar network, and of the contri-
bution of these salts to the cohesion of the lamellar structure.
The presence of structure peaks at wide angles was also noticed
(localized by circles and squares on Fig. 2b and c).
Their positions correspond to those of a platinoid salt, and the
low intensity and the wide shape of the peaks are typical of
crystalline nanodomains. M1-Pd(II) was rst considered, the SAXS
diagram displays four broad peaks located at 11.9 nmꢁ1
,
Fig. 5 TEM images of M1, M1Pd(II) and M1Pt(IV) materials.
77622 | RSC Adv., 2015, 5, 77619–77628
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
The structural modications between the initial and the
impregnated materials were also observed from TEM micro-
graphs (Fig. 5).
From the TEM images, a layered structure is observed for the
M1 material, whereas the M1-Pd(II) and M1-Pt(IV) materials
show nanodomains that could be platinoid chloride nano-
crystallites.
Owing to the differences between the Pt(II)–CTAB complex
structure depicted in the literature15 and that of the M1-Pd(II)
material, the composition of both complexes is probably
different. In order to evaluate the insertion ratio and conrm
our structural hypothesis, thermal analyses coupled with mass
spectroscopy of the initial M1 material and the impregnated
materials were carried out (Fig. 6).
The results for the M1 material show a CO2 and water release
of about 10.5 wt% between 70 ꢀC and 120 ꢀC. This departure is
almost nonexistent in the case of the M1-Pt(IV) (<1%) material,
reecting the full substitution of the CO2 by the platinum salt
during the impregnation. In addition, the increase of the
residual weight at the end of the analysis from 24.8 wt% to
46 wt% respectively for the M1 and the M1-Pt(IV) materials
conrms the impregnation of the metallic species into the
material.
However, considering the possible reduction of the platinum
salt even under the air ow, the residual mass at the end of the
thermal treatment does not allow the precise determination of
the materials composition.
The accurate composition of M1 and M1-Pt(IV) was deter-
mined by elemental analysis (Table 1).
While the results for the M1 material give atomic ratios in
agreement with that of the undecylamine chain, some incon-
sistency was observed for M1-Pt(IV). Indeed, the silicon/carbon
and silicon/nitrogen atom ratios are respectively 6 and 0.58.
The lack of carbon and nitrogen atoms is probably due to a
partial reduction process that occurred during the thermal
treatment.
In order to estimate more precisely the amount of platinoid
salts into the materials, the elemental analysis results were
cross-checked with X-EDS analyses of the impregnated mate-
rials M1-Pd(II) and M1-Pt(IV) (Table 2).
Fig. 6 Thermal analysis (TGA-MS) of (a) initial M1 material and (b) M1-
Pt(IV) material.
centred C (FC) orthorhombic structure would be in line with the
extinction of the Bragg reections on planes for which h and k
indexes are of different parities. In the latter case, the angle
between the a and b base vectors would be 86.8ꢀ, which is very
close to that of a tetragonal structure of lattice parameter a ¼
0.53 nm and c ¼ 3.14 nm. In addition, the Bragg reections on
the planes (020) and (021) are not observable, while those of
(200) and (201) are visible.
This information leads us to clearly favour the tetragonal
lattice of space group P4 with regards to the orthorhombic
structure. Fig. 4b shows that the tetragonal structure gives a
better agreement with the experimental data than the orthog-
onal structure.
According to the X-EDS analyses of M1-Pd(II), the composi-
tion was estimated to (OxSi–(CH2)11–NH2)2.6$PdCl2.2. The Pd/Cl
atomic ratio (measured at 2.18) suggests that the palladium has
an oxidation-state of 2.
Concerning the structure of the M1-Pt(IV) material, no rele-
vant references in the literature were found. Furthermore, its
SAXS pattern displays only three broad peaks at 9.2, 10.9 and
18.7 nmꢁ1, making the indexation very delicate. We therefore
could not give more detail on the structure of the platinoid salt
layer in this material. For both materials, the wide shape and
low intensity of the structure peaks could be due to a restrained
crystallinity, suggesting that the crystallites are probably very
small.
Table 1 Elemental analysis of M1 and M1-Pt(IV) materials
M1
M1-Pt(IV)
wt%
wt%
n/Si
1.00
11.40
0.95
—
n/Si
Si
C
N
Cl
Pt
0.11
0.55
0.05
—
0.10
0.27
0.03
0.17
0.21
1.00
6.00
0.58
1.32
0.29
—
—
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 77619–77628 | 77623

View Article Online
RSC Advances
Table X-EDS analysis of the M1-Pd(II) and M1-Pt(IV) materials.
Acquisitions were performed on polish pellets on 49 points for
M1-Pd(II) and 5 points for M1-Pt(IV)
Paper
2
Moreover, no additional vibration band were observed in
comparison to the original spectra of platinoid salts.
In particular, there is no direct bonding between the metal
and nitrogen atoms since the corresponding modes would
M1-Pd(II)
Cl/Pd
M1-Pt(IV)
Cl/Pt
appear around 500–550 cmꢁ1
.
From these results, we get
56,57
Si/Pd
Si/Pt
information on the structure of the platinoid salt inside the
lamellar matrix which is in square planar conformation for the
palladium salt and in octahedral conformation for the platinum
salt.
Atomic mean ratio
Standard deviation
2.18
0.17
2.56
0.12
4.35
0.27
2.80
0.08
Finally, to complete the structural study of M1-Pd(II) and give
a schematic representation of the structure, the position of the
different chemical species inside the lamellar structure had to
be determined. Usually, the shortest distance between two Pd
Concerning the M1-Pt(IV) material, both elemental and
X-EDS analyses proved that the Pt/Cl atomic ratio is included
between 4 and 4.5, suggesting that the platinum has an oxida-
tion state of 4 and involving a composition (OxSi–(CH2)11–
˚
inside a classic PdCl2 structure is close to 3.4 A. Here, this
˚
distance is of 5.3 A, suggesting the insertion of an amine
between two Pd atoms. Indeed, the addition of twice the
palladium ionic radius (62 pm) and twice the amine radius
NH2)2.8$PtCl4.4
.
In order to conrm these results, Raman spectroscopy has
been performed. Three main modes are allowed in the Raman
excitation for regular octahedral MX6 or planar square MX4
species, namely the symmetric and antisymmetric stretching
modes of M–Cl bounds and the bending mode of Cl–M–Cl
fragments.52,53
˚
(200 pm) gives a distance of 5.24 A, which is in good agreement
with the lattice parameter of the tetragonal structure high-
lighted by the SAXS experiment.
Owing to the M1-Pt(IV) octahedral conformation, the two
supplemental chloride out of the plane compared to the
M1-Pd(II) could cause the deformation from the tetragonal
structure into another structure.
Fig. 7a and b display the Raman spectra of M1-Pd(II) and
M1-Pt(IV) respectively.
Our structural study thus highlights the unexpected forma-
tion of a crystalline monolayer phase at the interface between
two organosilica bilayers, leading to an original hybrid
For comparison, the platinoid salt precursors are also
shown. The a-form of PdCl2 can be viewed as collinear poly-
meric chains in which each Pd atom alternates with two Cl
atoms. The Raman spectrum of PdCl2 (Fig. 7a) exhibits an
intense band at 274 cmꢁ1, corresponding to the symmetric
stretching frequency of Pd–Cl bounds.54 The antisymmetric
stretching mode is observable in the foot of the latter band at
290 cmꢁ1. The symmetric bending mode can also be revealed at
144 cmꢁ1, its intensity being lower due to the restrained
number of terminal Pd–Cl stretches in this “innite” chains
conguration. The spectrum of M1-Pd(II) does not show the
intense band at 274 cmꢁ1 anymore, but two intense bands at
288 and 306 cmꢁ1 that are assigned to the symmetric and
antisymmetric stretching frequency of Pd–Cl bounds.52,55
The appearance of the antisymmetric mode could involve
that PdClx units are not in the “innite” chains conguration in
the material. However, the bending mode usually observed
between 160 and 200 cmꢁ1 is not present in this spectrum,
signifying that the interface between the two bilayers prevents
this vibration mode to appear.
Comparatively, the symmetric and antisymmetric stretching
frequency of Pt–Cl bounds show up in the spectra of the
M1-Pt(IV) sample and that of the H2PtCl6$nH2O53 respectively at
315 and 340 cmꢁ1 (Fig. 7b).
A supplemental broad band at 150–180 cmꢁ1 was also
observed for the M1-Pt(IV) sample. The bending mode is thus
allowed through the in-plane and out of plane deformation of
the [PtCl6]2ꢁ unit. Nevertheless, the weaker intensity of the
latter band for M1-Pt(IV) compared to that of H2PtCl6$nH2O50
suggests that the bending mode would be permitted in M1-Pt(IV)
by the two supplemental chlorine atoms which would not
interact with the interfaces of the material.
Fig. 7 Raman spectra of (a) M1-Pd(II) and (b) M1-Pt(IV) and the corre-
sponding metallic salts (respectively PdCl2 and H2PtCl6$H2O).
77624 | RSC Adv., 2015, 5, 77619–77628
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
mesostructure: a trilayered material composed of an incorpo-
The weight losses during this range of temperature have
rated crystalline monolayer phase of palladium chloride at the been evaluated to 8% (Fig. 8a), in the same range than the
interface between two organosilica bilayers. Such a structure calculation, and no supplemental loss was observed aer.
can be described by an ordered packing of organo- In the case of M1-Pt(IV), the removal of four chlorides atoms
silica@platinoid@organosilica sandwiches, which we called was estimated to 14.6% which is in agreement with the exper-
“OrganoSilica Sandwiches” (OSS) arrangement (Fig. 3).
iment with a weight loss evaluated to 15.2% from 120 and
150 C (Fig. 8b).
ꢀ
Therefore, we can consider that the reduction is effective for
both materials and the thermal reduction is done with very
different behaviours as function of the metallic salt involved.
The resulting materials M1-Pd(0) and M1-Pt(0) became
respectively black and dark grey.
The thermal effect on the sandwich-layered structure aer
reduction has been evaluated by SAXS experiment at room
temperature for both impregnated materials (Fig. 9).
Firstly, in order to evaluate the thermal effect on the
sandwich-layered structure, we focused on the region of scat-
tering vectors q ranging from 1 nmꢁ1 to 10 nmꢁ1 of the SAXS
proles. The SAXS pattern of M1-Pd(0) displays structure peaks
located at 1.95, 3.9, 5.8 and 7.7 nmꢁ1 which correspond to the
slightly contracted initial lamellar structure of the M1-Pd(II)
material.
3.2. Thermal reduction effects
The stable sandwich-layered structures described in the para-
graph above were subjected to a reduction treatment. The
reduction of impregnated materials was performed through an
in situ TGA experiment under 5% H2 in Argon ow (20
mL minꢁ1) from 20 C to 140 C.
ꢀ
ꢀ
The reduction equation involves the formation of two
hydrochloric acids in the case of the palladium salt and four
hydrochloric acids in the case of the platinum salt, which can be
removed from the material or trapped by the amines contained
in the material. In addition, metallic palladium is known for
trapping hydrogen. Therefore, the quantitative measurement of
the reduction ratio is thus not available by this technique.
We can observe an abrupt weight loss between 60 ꢀC and
80 ꢀC in the TGA curve (Fig. 8) of the M1-Pd(II) sample and
The intensity of the rst structure peak is weaker than that of
the material before reduction, but comparable to that of the
initial M1 material. This denitely indicates that the electronic
contrast between silica planes and the interface of bilayers has
come back close to that found in the original lamellar packing
of M1.
ꢀ
between 120 and 150 C for the M1-Pt(IV) sample. Considering
the composition of M1-Pd(II) as (OxSi–(CH2)11–NH2)2.6$PdCl2.2
,
the loss of two chloride ions during the removal of hydrochloric
acid is calculated to 9.3% of the total mass of impregnated
M1-Pd(II) material.
It could also signicate that the palladium salt is not at the
interface of bilayers anymore.
While the M1-Pd(0) SAXS prole seems to present the orig-
inal lamellar structure, the M1-Pt(0) sample shows a very
distinct prole. The structure peak located around 3.9 nmꢁ1 is
the only remaining peak of the initial lamellar structure.
Nevertheless, a broad lump around 1.2 nmꢁ1 appears.
This emergence could be due either to a form factor of
objects dispersed in the material, to a structure factor caused by
the interactions of nano-sized objects, or to a combination of
these two possibilities.
Fig. 9 SAXS profile of samples M1-Pd(0) (grey) and M1-Pt(0) (black).
Fig. 8 In situ thermal reduction analysis of (a) M1-Pd II and (b) The grey square and the black disk correspond respectively to the first
M1-Pt(IV).
Bragg peak of Pd0 and Pt0.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 77619–77628 | 77625

View Article Online
RSC Advances
Paper
In all cases, the presence of nano-sized objects has to be
considered. By focusing on the wider angles on SAXS pattern, we
conrm the hypothesis of nano-sized objects. First, the struc-
ture peaks characteristic of platinoid salts (Fig. 2) are missing
on both SAXS proles (Fig. 9).
In comparison, only one structure peak is observable at
around 28.0 nmꢁ1 (marked by a square and a ring respectively
for the M1-Pd(0) and the M1-Pt(0) materials in the SAXS
patterns displayed in Fig. 9). Each structure peak is character-
istic of metallic palladium and platinum, conrming the
effective reduction of platinoid salts.
In order to estimate the platinoid particle size, we applied
the Scherrer equation (eqn (1)).
kl
FWHM2 ꢁ s2 cos q
where k ¼ 0.89 is a standard correction factor, l ¼ 0.71 A is the
used wavelength, FWHM is the Full Width at Half Maximum of
the peak, s is the instrumental resolution and q is the peak
position in radian. q is obtained from the scattering vector q
according to the eqn (2):
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
t ¼
(1)
˚
4p sin q
q ¼
(2)
l
Fig. 10 SEM images of samples of (a) M1-Pd(0) and (b) M1-Pt(0).
The FWHM have been calculated from the Voigt tting
equation shown in Fig. S2.†
Table 3 includes the parameters and calculation results for
both materials containing the metallic species.
The TEM images of M1-Pd(0) (Fig. 11a) shows nanoparticles
of different shapes. The Fig. 11b is a focus on one of them. The
particles looks attened and composed of multi-facets from
several merged smaller particles.
The lattice spacing between the (111) planes of metallic Pd is
observed in Fig. 11c at 0.22 nm, which is in good agreement
with the distance found in bulk Pd crystal and SAXS results.
The calculation indicates that palladium particles are larger
than platinum particles. The palladium particle size is about
43 nm, while the platinum particles are estimated to 13 nm
(Table 3). The above-discussed contribution in the SAXS prole
at small angles (Fig. 9) of the M1-Pt(0) sample could thus be
explained by the form factor of such particles, and by the
structure factor contribution coming from interactions between
these particles. Indeed, the range of observable scattering vector
q in the SAXS measurement conditions (0.25 nmꢁ1 < q <
30 nmꢁ1) allows observing the contribution of particles whose
size is smaller than 25 nm.
SEM images of both materials are also very distinct (Fig. 10).
The M1-Pd(0) image exhibits a layered dappled powder, testi-
fying to the presence of palladium nanoparticles very dispersed
inside the material (Fig. 10a). The typical size of visible particles
is in agreement with that calculated by the Scherrer equation.
The M1-Pt(0) image (Fig. 10b) shows inhomogeneous disper-
sion of objects whose size is included between 50 and 500 nm.
Table 3 Parameters and calculated size of platinoid particles in
impregnated materials
Pd
Pt
q/nmꢁ1
28.1
0.45
0.37
28.0
0.92
0.36
FWHM/nmꢁ1
s/nmꢁ1
Size/nm
43.1 ꢄ 0.3
13.4 ꢄ 0.2
Fig. 11 TEM images of sample M1-Pd(0).
77626 | RSC Adv., 2015, 5, 77619–77628
This journal is © The Royal Society of Chemistry 2015

View Article Online
Paper
RSC Advances
materials was obtained thanks to bis-silica-alkylammonium
carbamate dimers acting as a layered template for the immobili-
zation of crystalline platinoid salt complexes. These complexes
were changed into metallic nanoparticles aer reduction treat-
ment. The layered template and sandwich structures, which we
called OSS materials, were investigated by SAXS, TGA-MS and
Raman spectroscopy. It is showed that the meso-structure of the
initial template material is not affected by the different elaboration
steps. Due to the combined effects of the connement and of the
crystallographic structure of platinoid salts, the thermal reduction
leads to the formation of well-dispersed palladium nanoparticles
or to platinum aggregated nanorods. This approach therefore gives
access to supported platinoids salts and nanoparticles which can
be useful for catalytic applications. Beyond their use as precursors
for the preparation of nanoparticles, such ionic hybrid materials
could have applications as ionic conductors at low temperature.
Fig. 12 TEM image of sample M1-Pt(0).
The TEM image in Fig. 12 conrms the presence of aggre-
gated nanoparticles in the M1-Pt(0) material.
The particles appear elongated, like rods aggregated into the
interlayer space of the lamellar structure of the M1-Pt(IV)
material.
The metallic species seem to have migrated at the extremity
of the lamellar network and outside of the material during the
reduction (Fig. S3†). Unfortunately, even if the TEM images
illustrated in Fig. S3† exhibit lattice fringes, we could not
determine the growth direction of the nanorods due to the
limited resolution of the images.
Acknowledgements
The authors acknowledge C. Rey for her help on thermal anal-
ysis aspects, R. Turgis for assistance in organic synthesis and
characterization of organic compounds, J. Ravaux for SEM and
X-EDS measurement and H.P. Brau for their help on TEM
measurements and assistance in the image processing.
The palladium and the platinum atoms have similar ionic
and atomic radius. Furthermore, their crystallographic struc-
ture is a face-centered cubic structure (FCC). The differences of
size and shape between palladium and platinum nanoparticles
is thus likely to arise from the initial oxidation state of platinoid
salt precursors (Pd2+ and Pt4+). The latter parameter inuences
the geometry of the initial nanocrystal growing at the interface
between the two alkylaminosilica bilayers of the material. Due
to this localization, the platinoid salts undergo a connement
effect before the thermal reduction. The square plans composed
of PdCl42ꢁ units result in a reduction toward two directions that
further leads to at aggregates. Oppositely, the structure
Notes and references
1 A. P. Alivisatos, Science, 1996, 271, 933–937.
2 G. Arrachart, I. Karatchevtseva, D. J. Cassidy, G. Triani,
J. R. Bartlett and M. W. C. Man, J. Mater. Chem., 2008, 18,
3643–3649.
`
3 F. Colbeau-Justin, C. Boissiere, A. Chaumonnot,
2ꢁ
composed of octahedral PtCl6 units could favor the growing
A. Bonduelle and C. Sanchez, Adv. Funct. Mater., 2014, 24,
233–239.
up of nanoparticles toward one direction leading to nanorods.
Other examples of platinum nanorod growth can be found in
the literature. Some syntheses involve SBA15 or MCM-41 silica
as cylindrical templates,21–33 and other so templating
approaches using surfactants and organic solvents result in the
¨
4 A. Zamboulis, N. Moitra, J. J. E. Moreau, X. Cattoen and
M. W. C. Man, J. Mater. Chem., 2010, 20, 9322–9338.
5 W. Otani, K. Kinbara, Q. Zhang, K. Ariga and T. Aida, Chem.–
Eur. J., 2007, 13, 1731–1736.
synthesis of platinum nanorods from
a lamellar meso-
´
6 R. Turgis, G. Arrachart, C. Delchet, C. Rey, Y. Barre, S. Pellet-
phase.15,18,19 Therefore, the shape of the nal metallic nano-
particles is not only determined by the geometry of the
connement matrix. Starting from the lamellar structure found
in both Pd- and Pt-bearing materials, we demonstrated in this
study that the local structure near metallic atoms is determi-
nant for the geometry of the nal metal nanoparticles, which
could be of planar or rod-like shape.
Rostaing, Y. Guari, J. Larionova and A. Grandjean, Chem.
Mater., 2013, 25, 4447–4453.
7 A. Charlot, S. E. Mourabit, F. Goettmann, G. Arrachart,
R. Turgis and A. Grandjean, RSC Adv., 2014, 4, 64138–64141.
8 W. Huang, J. Guo, Y. Xiao, M. Zhu, G. Zou and J. Tang,
Tetrahedron, 2005, 61, 9783–9790.
9 L. N. Lewis and N. Lewis, J. Am. Chem. Soc., 1986, 108, 7228–
7231.
10 J. Stein, L. N. Lewis, Y. Gao and R. A. Scott, J. Am. Chem. Soc.,
1999, 121, 3693–3703.
4. Conclusion
The self-assembly of amino-undecyl-triethoxysilane molecules with 11 J. L. Speier, Catal. Org. Synth., Academic Press, 1979, 407–
carbon dioxide as removable assembly agent is an excellent route
447–.
to produce hybrid mesostructured lamellar silica. In this article, we 12 M. Chauhan, B. J. Hauck, L. P. Keller and P. Boudjouk, J.
develop stable sandwich-layered structures which consist of Organomet. Chem., 2002, 645, 1–13.
stacked metal layers between organosilica layers. The formation of 13 C. C. C. Johansson Seechurn, M. O. Kitching, T. J. Colacot
such
sandwich-like
organosilica@platinoid@organosilica
and V. Snieckus, Angew. Chem., Int. Ed., 2012, 51, 5062–5085.
This journal is © The Royal Society of Chemistry 2015
RSC Adv., 2015, 5, 77619–77628 | 77627

View Article Online
RSC Advances
Paper
´
´
14 Q. M. Kainz, R. Linhardt, R. N. Grass, G. Vile, J. Perez- 35 M. H. Lim and A. Stein, Chem. Mater., 1999, 11, 3285–3295.
´
Ramırez, W. J. Stark and O. Reiser, Adv. Funct. Mater., 36 T. Asefa, M. J. MacLachlan, N. Coombs and G. A. Ozin,
2014, 24, 2020–2027. Nature, 1999, 402, 867–871.
15 R. Krishnaswamy, H. Remita, M. Imperor-Clerc, C. Even, 37 W. J. Hunks and G. A. Ozin, J. Mater. Chem., 2005, 15, 3716–
´
P. Davidson and B. Pansu, ChemPhysChem, 2006, 7, 1510–
1513.
16 G. Surendran, L. Ramos, B. Pansu, E. Prouzet, P. Beaunier,
F. Audonnet and H. Remita, Chem. Mater., 2007, 19, 5045–
5048.
3724.
38 P. V. D. Voort, D. Esquivel, E. D. Canck, F. Goethals,
I. V. Driessche and F. J. Romero-Salguero, Chem. Soc. Rev.,
2013, 42, 3913–3955.
¨
´
39 E. Ruiz-Hitzky, S. Letaıef and V. Prevot, Adv. Mater., 2002, 14,
17 G. Surendran, F. Ksar, L. Ramos, B. Keita, L. Nadjo,
439–443.
´
E. Prouzet, P. Beaunier, P. Dieudonne, F. Audonnet and 40 R. Voss, A. Thomas, M. Antonietti and G. A. Ozin, J. Mater.
H. Remita, J. Phys. Chem. C, 2008, 112, 10740–10744. Chem., 2005, 4010–4014.
18 F. Ksar, G. Surendran, L. Ramos, B. Keita, L. Nadjo, 41 Q. Zhang, K. Ariga, A. Okabe and T. Aida, J. Am. Chem. Soc.,
`
E. Prouzet, P. Beaunier, A. Hagege, F. Audonnet and
2004, 126, 988–989.
H. Remita, Chem. Mater., 2009, 21, 1612–1617.
42 A. Shimojima and K. Kuroda, Chem. Rec., 2006, 6, 53–63.
19 Y. Song, R. M. Garcia, R. M. Dorin, H. Wang, Y. Qiu, 43 (a) J. J. E. Moreau, B. P. Pichon, G. Arrachart, C. M. Wong Chi
E. N. Coker, W. A. Steen, J. E. Miller and J. A. Shelnutt,
Nano Lett., 2007, 7, 3650–3655.
20 Y.-M. Lu, H.-Z. Zhu, W.-G. Li, B. Hu and S.-H. Yu, J. Mater.
Chem. A, 2013, 1, 3783–3788.
Man and C. Bied, New J. Chem., 2005, 29, 653–658; (b)
G. Arrachart, A. Bendjerriou, C. Carcel, J. J. E. Moreau and
M. Wong Chi Man, New J. Chem., 2010, 34, 1436–1440.
´
ˇ ´
¨
44 S. C. Nunes, K. Burglova, J. Hodacova, R. A. S. Ferreira,
¨
21 C. H. Ko and R. Ryoo, Chem. Commun., 1996, 2467–2468.
22 Z. Liu, O. Terasaki, T. Ohsuna, K. Hiraga, H. J. Shin and
R. Ryoo, ChemPhysChem, 2001, 2, 229–231.
L. D. Carlos, P. Almeida, X. Cattoen, M. Wong Chi Man
and V. de Zea Bermudez, Eur. J. Inorg. Chem., 2015, 1218–
1225.
´
23 F. Kleitz, S. H. Choi and R. Ryoo, Chem. Commun., 2003, 45 A. Chemtob, L. Ni, C. Croutxe-Barghorn and B. Boury,
2136–2137. Chem.–Eur. J., 2014, 20, 1790–1806.
24 Z. Liu, Y. Sakamoto, T. Ohsuna, K. Hiraga, O. Terasaki, 46 J. Alauzun, E. Besson, A. Mehdi, C. Reye and R. J. P. Corriu,
C. H. Ko, H. J. Shin and R. Ryoo, Angew. Chem., Int. Ed.,
2000, 39, 3107–3110.
25 Y.-J. Han, J. M. Kim and G. D. Stucky, Chem. Mater., 2000, 12,
2068–2069.
26 L.-X. Zhang, J.-L. Shi, J. Yu, Z.-L. Hua, X.-G. Zhao and
M.-L. Ruan, Adv. Mater., 2002, 14, 1510–1513.
27 C. Yang, P. Liu, Y. Ho, C. Chiu and K. Chao, Chem. Mater.,
2003, 15, 275–280.
Chem. Mater., 2008, 20, 503–513.
47 R. Besnard, J. Cambedouzou, G. Arrachart, O. Diat and
S. Pellet-Rostaing, Langmuir, 2013, 29, 10368–10375.
48 R. Besnard, G. Arrachart, J. Cambedouzou and S. Pellet-
Rostaing, RSC Adv., 2015, 5, 57521–57531.
49 Y. Li, R. Beck, T. Huang, M. C. Choi and M. Divinagracia, J.
Appl. Crystallogr., 2008, 41, 1134–1139.
50 S. Lambert, K. Y. Tran, G. Arrachart, F. Noville, C. Henrist,
C. Bied, J. J. E. Moreau, M. W. Chi Man and B. Heinrichs,
Microporous Mesoporous Mater., 2008, 115, 609–617.
51 B. P. Pichon, M. Wong Chi Man, C. Bied and J. J. E. Moreau,
J. Organomet. Chem., 2005, 1126–1130.
´
28 J. Zhu, Z. Konya, V. F. Puntes, I. Kiricsi, C. X. Miao, J. W. Ager,
A. P. Alivisatos and G. A. Somorjai, Langmuir, 2003, 19, 4396–
4401.
29 D. Wang, H. Luo, R. Kou, M. P. Gil, S. Xiao, V. O. Golub,
Z. Yang, C. J. Brinker and Y. Lu, Angew. Chem., Int. Ed., 52 P. J. Hendra, Nature, 1966, 212, 179.
2004, 43, 6169–6173.
53 L. A. Woodward and J. A. Creighton, Spectrochim. Acta, 1961,
30 S.-H. Liu, R.-F. Lu, S.-J. Huang, A.-Y. Lo, S.-H. Chien and
S.-B. Liu, Chem. Commun., 2006, 3435–3437.
31 N. C. King, R. A. Blackley, W. Zhou and D. W. Bruce, Chem.
Commun., 2006, 3411–3413.
17, 594–599.
54 N. M. Aghatabay, M. Somer, M. Senel, B. Dulger and
F. Gucin, Eur. J. Med. Chem., 2007, 42(8), 1069–1075.
55 H. Stammreich and R. Forneris, Spectrochim. Acta, 1960, 16,
363–367.
32 N. C. King, R. A. Blackley, M. L. Wears, D. M. Newman,
W. Zhou and D. W. Bruce, Chem. Commun., 2006, 3414–3416. 56 J. Pei and X. Li, J. Solid State Electrochem., 2000, 4, 131–140.
33 H. Liu, D. Ma and X. Bao, Dalton Trans., 2009, 1894–1896.
57 D. Michalska and R. Wysokinski, Chem. Phys. Lett., 2005,
403, 211–218.
´
`
34 E. D. Martınez, C. Boissiere, D. Grosso, C. Sanchez,
H. Troiani and G. J. A. A. Soler-Illia, J. Phys. Chem. C, 2014,
118, 13137–13151.
77628 | RSC Adv., 2015, 5, 77619–77628
This journal is © The Royal Society of Chemistry 2015