See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/47560899
Synthesis of composite ruthenium-containing silica nanomaterials from amine-stabilized ruthenium nanoparticles... Article in Journal of Materials Chemistry · October 2010 DOI: 10.1039/c0jm01680b · Source: OAI
CITATIONS
READS
8
18
6 authors, including: Karine Philippot
Yannick Guari
French National Centre for Scientific Research
French National Centre for Scientific Research
165 PUBLICATIONS 3,813 CITATIONS
143 PUBLICATIONS 2,669 CITATIONS
SEE PROFILE
SEE PROFILE
Pierre Lecante
Bruno Chaudret
Centre National de la Recherche Scientifique,…
French National Centre for Scientific Research
139 PUBLICATIONS 3,360 CITATIONS
524 PUBLICATIONS 16,088 CITATIONS
SEE PROFILE
All content following this page was uploaded by Mar Tristany on 09 January 2014.
The user has requested enhancement of the downloaded file.
SEE PROFILE
PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Synthesis of composite ruthenium-containing silica nanomaterials from amine-stabilized ruthenium nanoparticles as elemental bricks†‡ Mar Tristany,ab Karine Philippot,*ab Yannick Guari,c Vincent Colliere,ab Pierre Lecanted and Bruno Chaudretab Received 31st May 2010, Accepted 28th July 2010 DOI: 10.1039/c0jm01680b A facile synthesis of composite ruthenium-containing silica nanomaterials from amine stabilizedruthenium nanoparticles as elemental bricks is described. This route takes advantage of an organometallic approach and the use of bifunctional H2N–(CH2)x–Si(OEt)3 amines as stabilizing ligands (x ¼ 3, 11) for the synthesis of ruthenium nanoparticles, from [Ru(COD)(COT)] (COD ¼ 1,3cyclooctadiene, COT ¼ 1,3,5-cyclooctatriene) as the metal precursor. Classical hydrolysis and polycondensation steps of the sol–gel approach further lead to the formation of the silica matrix around the ruthenium nanoparticles. [RuO2]@SiO2 nanocomposites are then obtained through a calcination step in air. The final composite nanomaterials were characterized by different techniques and they were found to possess a high specific surface area making them attractive materials for catalysis.
Intoduction The synthesis of nanocomposite materials has attracted much attention because of their potential applications in numerous areas.1 In this respect, metal nanoparticles included inside mesoporous membranes are interesting for applications in catalysis2 and were shown to be active catalytic filters to improve the gas sensor selectivity.3 However, the interest in composite nanomaterials as catalysts depends on their efficiency, in terms of selectivity, reproducibility and long time stability. This requires homogeneous size and dispersion as well as high surface area of the active species in the mesoporous support, together with their immobilization inside the inorganic host material to prevent any migration and coalescence.4 Hybrid metal@SiO2 materials are, in general, prepared using the pores/channels of mesoporous silica as templates either for the inclusion of preformed metal nanoparticles5 or for controlling the in situ growth of nanoparticles from metal precursors in the presence of reducing agents.6 Other works describe the grafting of ruthenium species to silica supports using a previously treated silica with an N-(2-aminoethyl)-3-aminopropyl trimethoxysilane7 or a ruthenium complex containing a pendant siloxy group.8 In these works, a siloxane linkage present on the support or as ligand in the metal precursor is used as grafting agent. a CNRS, LCC (Laboratoire de Chimie de Coordination), 205 Route de Narbonne, F-31077 Toulouse, France b Universit e de Toulouse, UPS, INPT, LCC, F-31077 Toulouse, France. E-mail:
[email protected] c Institut Charles Gerhardt Montpellier, Chimie Mol eculaire et Organisation du Solide, UMR 5253 CNRS-UM2.Universit e de Montpellier II, cc1701, Place E. Bataillon, F-34095 Montpellier Cedex 05, France d CNRS, CEMES (Centre d’Elaboration de Mat eriaux et d’Etudes Structurales), 29 rue J. Marvig, F-31055 Toulouse, France † This paper is part of a Journal of Materials Chemistry themed issue on Advanced Hybrid Materials, inspired by the symposium on Advanced Hybrid Materials: Stakes and Concepts, E-MRS 2010 meeting in Strasbourg. Guest editors: Pierre Rabu and Andreas Taubert. ‡ Electronic supplementary information (ESI) available: Supplementary Figs SI1–6. See DOI: 10.1039/c0jm01680b
This journal is ª The Royal Society of Chemistry 2010
These methods allow a good dispersion of the metal nanoparticles, but they necessitate several steps and sometimes harsh conditions (strong reducing agents, high temperatures.). In addition, it is difficult to maintain real control of the metal particles size distribution, a parameter that is critical in determining catalytic activity and selectivity,9 and of their dispersion in the solid support. This is due to small pore size or to the presence of functional groups that can limit a total filling of the matrices on one hand, or nanoparticles sintering that can appear at high temperature on the other hand. To circumvent these difficulties, some groups tried to develop novel preparation methods, in a few steps and in mild conditions, allowing the synthesis of both the metal particles and the silica. For instance, Dai and co-workers succeeded in the preparation of silica materials containing gold nanoparticles via a cosynthesis sol–gel process, leading to the simultaneous formation of mesostructures and metal-ion doping.10 Somorjai and Yang and co-workers reported the solvothermal growth of mesoporous SBA-15 silica in the presence of PVP-stabilized platinum nanoparticles by the nanoencapsulation method.11 The colloidal platinum solution was mixed with an aqueous solution of pluronic acid used as an organizing mediator to build the silica matrix. As another example, Budroni and Corma described the preparation of gold organic–inorganic materials following a twostep procedure.12 In this case, two surfactants were employed (1-dodecane thiol and 3-mercaptopropyltrimethoxysilane) for capping the gold nanoparticles during their synthesis and a second step consisting in hydrolysis/polycondensation of the alkoxysilane with tetraethyl ortho silicate (TEOS), which was necessary to obtain the nanomaterials. In our case, we have recently reported a relevant methodology, based on the use of metal nanoparticles as elemental bricks for the preparation of well-controlled silica nanocomposite materials thanks to a unique but bifunctional ligand that acts simultaneously as a nanoparticle stabilizer and a sol–gel promoter.13 Despite the simplicity of our methodology, a weak point could be pointed out, the use of a complex ligand for the stabilization of the metal particles. As this could limit the interest of this methodology, we J. Mater. Chem., 2010, 20, 9523–9530 | 9523
decided to evaluate the generality of this approach using more simple amines. We report in this paper, new results on the development of our method for the access to silica materials containing ruthenium oxide nanoparticles using commercial amines.
Results and discussion Our methodology for the synthesis of composite rutheniumcontaining silica nanomaterials takes advantage of the mild conditions of the organometallic approach to form metal nanoparticles.14 The key-point is the use of a bifunctional ligand for the stabilisation of ruthenium nanoparticles playing a double role. First, the amine group present in the ligand L (either 3aminopropyltriethoxysilane, H2N(CH2)3Si(OEt)3; L1 or 11aminoundecyltriethoxysilane, H2N(CH2)11Si(OEt)3; L2) (see Scheme 1), can interact with the surface of the nanoparticles and stabilize them. Secondly, it contains a triethoxysilyl group that can undergo the hydrolysis and polycondensation steps of a sol– gel approach,15 and consequently lead to the formation of the silica materials containing the metal nanoparticles. In this way, the ruthenium nanoparticles are first stabilized by the bifunctional ligand and further used as elemental bricks for the formation of the silica matrix around them. As the metal nanoparticles are embedded in the silica matrix thanks to their own stabilizing ligand, they are uniformly dispersed in silica. A further calcination step leads to the oxidation of the Ru nanoparticles into RuO2. This methodology was previously applied using N-(phenylmethyl)-N-[11-[3-(triethoxysilyl)propoxy]undecyl]–(PhCH 2)2N(CH 2)11O(CH2)3Si(OEt)3.13 As
shown in Scheme 1, the procedure involves first exposure to a dihydrogen atmosphere (3 bar) of a yellow THF solution of the organometallic complex Ru(COD)(COT) (COD ¼ 1,3cyclooctadiene; COT ¼ 1,3,5-cyclooctatriene), which is a well-known, efficient precursor for the synthesis of sizecontrolled Ru nanoparticles,16 in the presence of ligand L1 or L2. While in our previous study,13 a large series of [L]/ [Ru(COD)(COT)] molar ratios had been studied, we here focus on three [L]/[Ru(COD)(COT)] systems (also called in this work x/y ratios), namely 0.5, 1 and 2. Dark-brown and stable colloidal solutions are obtained containing small-sized and well-dispersed ruthenium nanoparticles as observed by electronic microscopy analysis. Using L1 (3-aminopropyltriethoxysilane) as ligand with a [L1]/ [Ru(COD)(COT)] molar ratio of 0.5, small and spherical nanoparticles are observed by transmission electronic microscopy (TEM) (Fig. 1a). These nanoparticles are well-dispersed on the grid and display a size in the range 1.0–2.6 nm with a mean diameter of 1.7(0.3) nm (see corresponding size histogram, Fig.1b). In addition, the high resolution-TEM image reveals the highly crystalline feature of these nanoparticles (Fig. 1c). These nanoparticles can be isolated by precipitation upon addition of pentane, purified by several washings and further dried under vacuum, giving rise to a black powder. Wide-angle X-ray scattering (WAXS) measurements on the purified solid confirmed the hcp crystalline structure of the ruthenium nanoparticles as for bulk ruthenium and revealed a coherence length in good agreement with the mean diameter measured by TEM (Fig. 2). The EDS analysis during the TEM observation from many different areas confirms
Scheme 1 A schematic representation of the synthesis of [Ru(0)]@L nanoparticles, [Ru(0)]@L0 and [RuO2]@SiO2 nanocomposites.
Fig. 1 A TEM image (a) with a corresponding size histogram (b) and HREM image (c) of ruthenium nanoparticles synthesized in the presence of ligand L1 with [L1]/[Ru] ¼ 0.5.
9524 | J. Mater. Chem., 2010, 20, 9523–9530
This journal is ª The Royal Society of Chemistry 2010
Fig. 2 WAXS measurement on ruthenium nanoparticles synthesized in the presence of ligand L1 with [L1]/[Ru] ¼ 0.5 (dark grey) and comparison with Ru hcp (light grey).
the presence of elements of the 3-aminopropyltriethoxysilane stabilizer (N and Si, Cu from TEM grid). Finally, purification of the particles gives rise to a black powder from which the coordination of L1 at the ruthenium(0) nanoclusters’ surface was corroborated by further investigation using elemental analysis, ATR-IR, and CP-MAS 13C and 29Si NMR spectroscopies (see Fig. SI1-SI3‡). When a [L1]/[Ru(COD)(COT)] molar ratio of 1 is used, similar nanoparticles are formed. TEM and HREM images show small and spherical nanoparticles that are also well-dispersed on the grid and well-crystallized (Fig. 3a and c). Their mean diameter is estimated to be 1.6(0.3) nm. Finally, for the ratio [L1]/[Ru] ¼ 2, slightly smaller ruthenium nanoparticles are formed, with a mean diameter estimated to be 1.5(0.3) nm (Fig. SI4). Even if the decrease in size observed here when the quantity of ligand increases is not so marked, these results are in agreement with our previous results using (PhCH2)2N(CH2)11O(CH2)3Si(OEt)3 as ligand. They thus confirm the possibility of controlling the size of the nanoparticles by varying the [ligand]/[Ru] ratio. The colloidal solutions obtained with L1 are very stable whatever the [L]/[Ru] ratio, since no precipitation of bulk ruthenium from the solution was observed after several weeks of storage under argon, hence demonstrating the efficiency of 3aminopropyltriethoxysilane in stabilizing the Ru(0) nanoparticles. Nevertheless, the purified [Ru(0)]@L solids are difficult to handle, as they have a tendency to burn in open air, most particularly at the smaller [L]/[Ru] ratio of 0.5. With a [L]/[Ru] ¼ 1 ratio, the solids displayed a lower tendency to burn in open air, while keeping a metal content sufficiently high for easier
characterization of the materials. Thus, the ratio [L]/[Ru] ¼ 1 appeared to be the best compromise of stability against burning of the [Ru(0)]@L solids and the metal content. This work has thus been pursued considering only this ratio. The second step of our synthesis methodology consists in the hydrolytic polycondensation of the triethoxysilyl groups borne by the ligand L1 and leads to the hybrid materials [Ru(0)]@L10 with L10 ¼ H2N(CH2)SiO1,5. This is achieved through the addition of a catalytic amount of tetra-n-butylammonium fluoride (TBAF) ([L1]/[TBAF] ¼ 1/0.02) in THF and a stoechiometric amount of water per Si-OR function ([L1]/[H2O] ¼ 1/1.5) as usually performed to obtain silica materials. This treatment was carried out for the [L1]/[Ru(COD)(COT)] ratio of 1 over 5 days at room temperature, leading to a dark suspension. Ligand excess was eliminated by successive washings with pentane leading to pure [Ru(0)]@L10 hybrid material as a black solid. TEM analysis performed from this solid after dispersion in THF reveals the presence of ruthenium nanoparticles embedded in a silica matrix (Fig. SI5‡) the size of which is difficult to estimate as the TEM images reveal very thick areas. Powder X-ray diffraction (XRD) in the range of 2q ¼ 10–100 on the isolated solid confirms the formation of amorphous silica (characteristic broad peak around 20 ), as well as the presence of bulk ruthenium in the hexagonal phase (characteristic peaks at 44 ). The third step of the procedure corresponds to an annealing treatment of the [Ru(0)]@L10 gels in air to eliminate the organic part and to transform the ruthenium nanoparticles into ruthenium oxide. We have previously described an annealing procedure involving heating the materials from room temperature to 400 C with a temperature rate of 2 C min1 and a final heating at 400 C for 5 h, giving rise to [RuO2]@SiO2 solid material. In the present case, such a calcination treatment did not lead to homogeneous materials, as revealed by TEM analysis of the resulting solids for which large objects of ruthenium oxide were observed beside small nanoparticles well-dispersed in the silica matrix (Fig. SI6‡). This is likely to result from the high reactivity of the [Ru(0)]@L10 materials in air due to the small size of the particles. Thus, a partial oxidation of the samples in mild conditions was performed by bubbling air directly in the argon kept THF suspension at room temperature or under 2 bar of synthetic air. Such a pre-treatment gives rise to non-pyrophoric samples on which a classical calcination step can be carried out. In these conditions, TEM observations after calcination reveal a majority of small RuO2 nanoparticles with a mean size of 2.0(0.4) nm that are well-dispersed in the silica matrix (Fig. 4b), but a few large objects of RuO2 are also present in some areas of the grid. An annealing treatment at only 300 C but with a longer
Fig. 3 A TEM image (a) with a corresponding size histogram (b) and HREM image (c) of ruthenium nanoparticles synthesized in the presence of ligand L1 with [L1]/[Ru] ¼ 1.
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 9523–9530 | 9525
Fig. 4 TEM micrographs of passivated [Ru(0)]@L10 hybrid gel (a) and of [RuO2]@SiO2 (b) materials for [Ru]/[L] ¼ 1 and corresponding size histograms.
time (15 h) did not avoid the formation of large RuO2 objects. The increase in size observed for the particles comparing TEM images collected after nanoparticles synthesis (1.6 nm) or passivation (1.7 nm) steps and the ones collected after the calcination step (2.0 nm) is in the same range as previously observed and is expected, as oxygen atoms are incorporated to form RuO2. Despite the presence of a few large objects of RuO2 in some areas of the TEM grid, the increase of the mean size of the particles from 1.6 to 2.0 nm is in agreement with the absence of important coalescence of the particles. The specific surface area of the [RuO2]SiO2 material was determined by BET measurement and found to be of 32 m2 g1. The ruthenium and silicon contents determined by elemental analysis are 43.7 wt% and 15.3 wt%, respectively. Finally, to try to circumvent the lack of homogeneity of the samples due to the presence of large objects of RuO2 after
Fig. 5 A HREM image and size histogram of ruthenium nanoparticles synthesized in the presence of ligand L1 and TEOS with [Ru]/[L1]/[TEOS] ¼ 1/1/1 (a); a TEM micrograph and size histogram of [RuO2]@SiO2 material (b).
9526 | J. Mater. Chem., 2010, 20, 9523–9530
calcination, some syntheses were performed in the presence of tetraethylorthosilicate (TEOS). The idea was to increase the thickness of the silica walls and thus to better take away the ruthenium particles from each other and to limit their coalescence. All the other reaction parameters were kept similar. In the presence of TEOS ([Ru]/[L1]/[TEOS] ¼ 1/1/1), nanoparticles displaying a mean size of 1.6(0.3) nm are formed (Fig. 5a). In these conditions, and as previously observed, after the polycondensation, passivation (through air injection) and calcination (400 C) steps, a solid material with higher specific surface area is obtained (237 m2 g1) with only a slightly lower Ru content (31.2 wt%). It is clear that the surface area of this material is low in comparison with other known silica composite materials, which can present more than 1000 m2 g1 as surface area.8 Nevertheless, the addition of TEOS allows one to get a solid with a higher porosity. This solid material was characterized by powder-XRD and 29Si NMR showing intensive peaks at 28 , 35 , 40 and 54 for RuO2 and signals at 102 and 110 ppm, respectively attributed to Q3 and Q4 silicon atoms, respectively. However, the results were not completely satisfactory since some large RuO2 objects are always present in some areas of the grid (Fig. 5b). A lack of stability for the [Ru(0)]@L10 hybrid materials during the calcination process appeared to us as the best explanation for such unsatisfactory results. The capping shell formed from L1 is not thick enough to allow an efficient separation of the ruthenium nanoparticles, which can thus easily migrate in the solid at higher temperatures and lead to large RuO2 objects. This led us to change the ligand used in step 1 for the synthesis of the ruthenium nanoparticles. As long alkylchain amines are already known in our group to be better stabilizers than short alkylchain ones (since they better embed the nanoparticles),17 we considered another bifunctional amine having a longer alkylchain than L1, namely, 11-aminoundecyltriethoxysilane (H2N(CH2)11Si(OEt)3; L2). Syntheses performed with this new ligand, following the same reaction conditions as for L1, with a [L2]/[Ru] ratio of 0.5, gave
Fig. 6 TEM images of ruthenium nanoparticles synthesized in the presence of ligand L2 with [L2]/[Ru] ¼ 1 with size histogram (a) and of corresponding [RuO2]@SiO2 material (b).
This journal is ª The Royal Society of Chemistry 2010
Fig. 7 TEM micrographs with size histograms of ruthenium nanoparticles synthesized in the presence of ligand L2 and TEOS ([Ru]/[L2]/ [TEOS] ¼ 1/1/1) (a) and of the corresponding [RuO2]@SiO2 material (b).
rise to small nanoparticles with a mean diameter around 1.6(0.5) nm, but did not allow the absence of large RuO2 objects at the end of the procedure. Thus, we pursued our study considering only a [L2]/[Ru] ratio of 1. In these conditions, the decomposition of Ru(COD(COT) in THF at room temperature and under 3 bar of dihydrogen in the presence of L2 gave rise to a dark brown colloidal suspension containing ruthenium nanoparticles with a narrow size distribution and having a mean diameter of 1.5(0.3) nm, as seen by TEM (Fig. 6a). As observed for L1, the obtained Ru nanoparticles tend to be smaller in size when the [Ru]/[L] increases. With L2, the mean size of the particles is similar to ligand L1, indicating that there is no obvious influence of the ligand chain length on the size of the particles. After hydrolytic polycondensation, giving rise to a [Ru(0)]@L20 intermediate, partial oxidation in mild conditions and finally calcination (400 C; 2 C min1; 5 h) to form [RuO2@SiO2] nanomaterial, ruthenium oxide nanoparticles that are well-dispersed in the silica matrix were observed by TEM with a very small quantity of large RuO2 objects in some areas of the grid (Fig. 6b). BET measurement of [RuO2]@SiO2 nanomaterial revealed a specific surface area of 143 m2 g1 and elemental analysis a Ru and Si content of 41 wt% and 15 wt%. When the same protocol was carried out adding TEOS in the first step of the procedure ([Ru]/[L2]/[TEOS] ¼ 1/1/1), similar
ruthenium nanoparticles (1.6(0.3) nm as mean diameter) are formed, but better results are obtained after polycondensation, passivation (through air injection) and calcination (400 C) steps (Fig. 7), as only a few large RuO2 objects are visible beside welldispersed RuO2 nanoparticles in the silica matrix. We did not observe a significant difference when performing the passivation step by exposure of the solid [Ru(0)]@L20 under 2 bar of synthetic air or by carrying out the calcination at 300 C. The soobtained RuO2 nanoparticles in the silica matrix display a mean size of 1.8(0.3) nm, which means a slight increase in size in comparison with the Ru nanoparticles used as elemental bricks (1.6 nm), as previously observed. BET measurement of [RuO2]@SiO2 nanomaterial revealed a specific surface area of 265 m2 g1 and elemental analysis a Ru and Si content of 33 wt% and 19 wt%, respectively. The better results obtained with ligand L2 support the theory that the combination between a longer alkyl chain bifunctional amine and an increasing of the thickness of the silica walls limits the migration of the particles inside the silica gel and further their coalescence. To evaluate the interest of the so-obtained [RuO2]@SiO2 nanomaterials, we used them for a test catalytic reaction, the aerobic oxidation of benzyl alcohol.18 The results obtained have to be considered as preliminary, as they were not optimized. In a typical experiment, the chosen [RuO2]@SiO2 hybrid material (in a quantity determined to get 0.071 mmol of Ru) was introduced into a Fischer–Porter reactor. Then the bottle was heated to 80 C and a toluene solution (5 mL) containing 100 mL of benzyl alcohol (1 mmol) was added to the catalyst. Except for the catalytic test of entry 1, which was carried out with ambient air, after vacuum treatment, the system was pressurized under 1 bar O2 and the reaction was started by stirring (t ¼ 0 min). The evolution of the reaction was followed by GC analysis of the aliquots taken from the reaction solution. The results are reported in Table 1. [RuO2]@SiO2 nanomaterial obtained with L1 ([Ru]/[L1] ¼ 1/ 1) gave rise to 60% conversion of benzyl alcohol into benzaldehyde after 24 h at 80 C under an air atmosphere but gave 100% conversion in the same time when the reaction was performed under 1 bar of pure oxygen (see entry 2 versus 1). The reaction goes faster under dioxygen pressure. [RuO2]@SiO2 nanomaterial prepared with the same ligand and in the presence of TEOS ([Ru]/[L1]/[TEOS] ¼ 1/1/1) led to a lower conversion of 91% after 24 h (entry 3 versus 2) in the same reaction conditions (80 C; 1 bar O2) and appeared less effective at short reaction times. This
Table 1 Preliminary tests of catalytic oxidation of benzyl alcohol with [RuO2]SiO2 nanomaterials Conversion (%) Entry a
1 2 3 4 5
Synthesis conditions of [RuO2]@SiO2
Catalytic conditionsa
Ru content (%)b
Ssc m2 g1
2h
5h
10 h
24 h
48 h
[Ru]/[L1] ¼ 1 [Ru]/[L1] ¼ 1 [Ru]/[L1]/[TEOS] ¼ 1/1/1 [Ru]/[L2] ¼ 1 [Ru]/[L2]/[TEOS] ¼ 1/1/1
air 1 bar O2 1 bar O2 1 bar O2 1 bar O2
43.72 43.72 31.25 41.21 32.41
32 32 237 143 265
10.8 45.2 17.7 20.0 34.9
20.9 82.5 29.8 45.8 56.3
34.2 94.5 52.1 79.6 73.1
60.1 100 90.9 100 100
87.6 — — — —
a Solvent: toluene; T ¼ 80 C; [Ru] ¼ 0.07 mmol; [benzyl alcohol] ¼ 1 mmol; [substrate]/[metal] ¼ 14 b Ru content determined by elemental analysis c Ss ¼ specific surface determined by BET
This journal is ª The Royal Society of Chemistry 2010
J. Mater. Chem., 2010, 20, 9523–9530 | 9527
difference may be explained by the lower quantity of ruthenium in the catalyst (Ru content of 31% against 43%) or by a lower diffusion of the substrate toward the metal. The two samples prepared from ligand L2 with and without TEOS gave rise to 100% conversion after 24 h. Nevertheless, [Ru]/[L2]/[TEOS] catalyst appears more efficient than the [Ru]/[L2] one at shorter reaction times (entry 5 versus 4). This behavior is different from the one observed for [Ru]/[L1] and [Ru]/[L1]/[TEOS]. Such differences of catalytic activity both regarding conversions at short reactions times and conversion after 24 h may be explained in terms of accessibility of the metal surface. From the catalytic results obtained for materials prepared using ligand L1, adding TEOS leads to a decrease in the catalytic activity, which is not the case for materials prepared using ligand L2. The metal surface seems more accessible even in the presence of TEOS for materials prepared with L2. Finally, sample [Ru]/[L1] appeared to be the best catalyst (entry 1). From these preliminary results, which show similar catalytic activities even if some differences have been observed, it is difficult to draw a clear conclusion. A good compromise between the ruthenium content and the specific area is necessary to get a better catalytic behavior but this will also depend on the accessibility of the metal surface. If we compare these preliminary results with other ones recently obtained with Ru-based materials described in the literature,18f,g the catalytic activity observed may appear lower. Nevertheless, this is a non-optimized study, which was carried out only to validate our methodology of synthesis and to show the interest of the so-obtained composite nanomaterials. In summary of this part, despite a moderate activity, the preliminary tests performed show that the prepared hybrid nanomaterials may have an interest as catalysts and would merit development. As a key point, the accessibility of the metal surface would require better control.
Conclusion A facile method based on the use of nanoparticles as elemental bricks to prepare [RuO2]@SiO2 nanocomposite materials has been developed. This procedure profits from the combination of both an organometallic approach for the synthesis of metal nanoparticles and a sol–gel one for the formation of silica solids. The key point is the use of ligands having a double role, namely as stabilizing agents for the nanoparticle synthesis and as precursors for the silica matrix. The polycondensation step is reached through the usual conditions for sol–gel synthesis that affords hybrid materials displaying small-sized and well-dispersed ruthenium nanoparticles in the silica gel. A final annealing treatment under air leads to porous solid materials containing ruthenium oxide nanoparticles. The extension of this methodology to 3-aminopropyltriethoxysilane and 11-aminoundecyltriethoxysilane, gives rise to reactive nanoparticles that have the tendency to burn in open air, thus making the annealing treatment difficult to control and the final materials not homogeneous. In these conditions, large ruthenium oxide nano-objects resulting from coalescence of the particles are indeed observed. To circumvent this inconvenience, a mild passivation can be applied to the materials obtained after the polycondensation step and before the 9528 | J. Mater. Chem., 2010, 20, 9523–9530
calcination one. This mild pre-treatment allows one to obtain better results, in particular with 11-aminoundecyltriethoxysilane, which has the longest alkylchain. Nevertheless, the addition of TEOS at the beginning of the reaction, thus allowing a higher thickness of the silica walls and limiting the migration and the coalescence of the particles, is necessary to get the best results. In summary, a sufficient alkyl chain length for the ligand and a sufficient thickness for the silica walls appear as the best compromise to get composite [RuO2]@SiO2 nanomaterials displaying a homogeneous dispersion of the nanoparticles inside the silica matrix and interesting surface specific area making them attractive materials for catalytic applications. Preliminary tests using such nanomaterials in the aerobic oxidation of benzyl alcohol have shown their interest even if no optimization was performed.
Experimental General procedures and reagents All operations concerning nanoparticle syntheses were carried out in Schlenck or Fischer–Porter glassware under argon or in a glove-box. (1,5-cyclooctadiene)(1,3,5-cyclooctatriene)ruthenium (0) complex was purchased from Nanomeps-Toulouse, 3-aminopropyltriethoxysilane (L1) and tetrabutylammoniumfluoride (1 M in THF) from Aldrich, 11-aminoundecyltriethoxysilane (L2) from Gelest, tetraethylorthosilicate (2 M in THF) from Aldrich. All these reactants were used without any further purification. Solvents (from SDS) were dried and distilled before use; THF over sodium-benzophenone and pentane over calcium hydride. All reagents and solvents were degassed before use by means of three freeze-pump-thaw cycles. Calcinations of the composite nanomaterials were performed on a Netzsch STA 409PC LUXX instrument. Solid samples for TEM/HREM observations were prepared using ultramicrotomy techniques and then deposited on carbon covered-copper grids. For HREM analyses, another thin layer of carbon was evaporated at the surface of the material slice resulting from ultramicrotomy. Colloidal solutions were observed after deposition of a drop of the crude solution on a covered holey copper grid. Microscopy analyses were performed at the ‘‘Service Commun de Microscopie Electronique de l’Universit e Paul Sabatier’’ (TEMSCAN-UPS). TEM images were obtained using a JEOL 1011 electron microscope operating HREM observations at 100 kV with a resolution point of 4.5 A. were carried out with a JEOL JEM 2010 electron microscope The size working at 200 kV with a resolution point of 2.35 A. distributions were determined through a manual analysis of the enlarged micrographs by measuring ca. 300 particles on a given grid to obtain a statistical size distribution and a mean diameter. Powder X-ray diffraction patterns were measured on a Panalytical MPDPro powder diffractometer equipped with a fast linear detector and carried out at the ‘‘Service de Diffraction des Rayons X du LCC’’. Specific surface areas were determined by the Brunauer– Emmett–Teller (BET) method on Micromeritics Gemini 2360 and Micromeritics ASAP 2010 analyzers at CIRIMAT-UPS. This journal is ª The Royal Society of Chemistry 2010
IR specta were recorded on a Perkin-Elmer GX 2000 spectrophotometer at the LCC with a 4 cm1 resolution. CP MAS solid-state NMR spectra were recorded on a Bruker AV400WB (at 100.5 MHz for 13C and at 79.4 MHz for 29Si) with a rotation of 12 kHz and a D1 of 5s. Elemental analyses were performed at the ‘‘Service de Microanalyses du LCC’’ for C, H, N contents and at the ‘‘Service Central d’Analyses du CNRS’’ (Vernaison, France) for Ru content. Data collection for WAXS was performed at the CEMESCNRS (Toulouse) on small amounts of powder. The samples were sealed in 1 mm diameter Lindemann glass capillaries. The measurements of the X-ray intensity scattered by the samples irradiated with graphite monochromatized Mo Ka (0.071069 nm) radiation were performed using a dedicated two-axis diffractometer. Measurement time was 15 h for each sample. Scattering data were corrected for polarization and absorption effects, then normalized to one Ru atom and Fourier transformed to obtain the RDFs. 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.
Sample synthesis Synthesis of colloidal Ru(0) nanoparticles Ruthenium nanoparticles (step 1 of Scheme 1) were prepared through the following general procedure, which is described for a [L1]/[Ru(COD)(COT)] molar ratio of 1: Ru(COD)(COT) (400 mg, 1.27 mmol) was introduced in a Fischer–Porter bottle reactor under an inert atmosphere. A solution of the chosen ligand L1 (296 mL, 1.27 mmol) in freshly distilled and degassed THF (40 mL) was added to the ruthenium precursor under argon using a Teflon canula. The obtained yellow solution was exposed to a dihydrogen atmosphere (3 bars) for 20 h at room temperature, leading to a dark-brown colloidal solution containing Ru nanoparticles. After elimination of dihydrogen pressure, a drop of the colloidal solution was deposited onto a holey carbon grid for TEM observation. The protocol is here described for the [L1]/[Ru] but the same type of experiment was performed with ligand L1 in other ratios [L]/[Ru] namely 0.5 and 2 as well as with ligand L2 for [L2]/[Ru] ratios 0.5 and 1. Preparation of [Ru(0)L0 ] hybrid materials [Ru(0)]@L0 hybrid materials were obtained from ruthenium colloidal solutions through catalytic condensation. Then, the hydrolytic polycondensation was performed by nucleophilic catalysis as follows: after dihydrogen pressure elimination, TBAF (1 M in THF solution; ratio [L]/[TBAF] ¼ 1/0.02) and 0.5 equivalent per Si–OR group of permuted water ([L]/ [H2O] ¼ 1/1.5) were added to the Ru colloidal solution and the mixture was further stirred for 5 days at room temperature giving rise to a dark black solid. Afterwards, the solvent was evaporated and the black residue was purified by This journal is ª The Royal Society of Chemistry 2010
filtration and pentane washings and finally dried under vacuum overnight. Passivation of [Ru(0)]L0 hybrid materials [Ru(0)]L0 hybrid purified materials were dispersed in freshly distilled and degassed THF under an argon atmosphere. As specified in the results part, the passivation of the materials was done following two different procedures; either air was injected through a syringe into the suspension or the mixture was pressurized under 2 bar of synthetic air. The solvent was then evaporated and the passivated solids dried under vacuum overnight. Preparation of [RuO2]@SiO2 hybrid materials [RuO2]@SiO2 hybrid materials were prepared by calcination of the corresponding [Ru(0)]@L0 nanomaterials from 20 C to 400 C under air (rate: 2 C min1, then 5 h at 400 C). In some cases this calcinations step was performed at 300 C for 15 h. [RuO2]@SiO2 hybrid materials obtained in the presence of TEOS [RuO2]@SiO2 hybrid materials obtained in the presence of TEOS were prepared following the same procedure but tetraethylorthosilicate (TEOS) was added in the THF solution at the beginning of the process with Ru(COD)(COT) and ligand L (with [Ru(COD)(COT)]/[L]/[TEOS] ¼ 1/1/1). After that, the treatments were the same as the ones previously reported for [RuO2]@SiO2 hybrid materials. Catalytic experiments In a typical experiment, the chosen [RuO2]@SiO2 hybrid material (the quantity of which was determined to get 0.071 mmol of Ru) was introduced into a 20 mL Fischer–Porter bottle. Then the bottle was heated to 80 C and a toluene solution (5 mL) containing 100 mL of benzyl alcohol (1 mmol) was added to the catalyst. After vacuum treatment, the system was pressurized under 1 bar O2 and the reaction was started by stirring (t ¼ 0 min). The evolution of the reaction was followed by GC analysis of the aliquots taken from the reaction solution.
Acknowledgements The authors are indebted to I. Fourqaux (CMEAB-UPS) for ultramicrotomy preparation of samples and to M. Zahmakiran for helpful discussions on catalysis. The authors gratefully acknowledge the financial support of this work by CNRS (Project PICS no. 2428).
References 1 (a) T. Harada, S. Ikeda, H. Ng, Y. T. Sakata, H. Mori, T. Torimoto and M. Matsumura, Adv. Mater., 2008, 18, 2190–2196; (b) R. M. Rioux, R. Komor, H. Song, J. D. Hoefelmeyer, M. Grass, K. Niesz, P. Yang and G. A. Somorjai, J. Catal., 2008, 254(1), 1– 11; (c) S. Ikeda, S. Ishino, T. Harada, N. Okamoto, T. Sakata, H. Mori, S. Kuwabata, T. Torimoto and M. Matsumura, Angew. Chem., Int. Ed., 2006, 45, 7063–7066; (d) J. Okal, Catal. Commun., 2010, 11, 508–512.
J. Mater. Chem., 2010, 20, 9523–9530 | 9529
2 (a) A. Taguchi and F. Sch€ uth, Microporous Mesoporous Mater., 2005, 77, 1–45; (b) F. Su, L. Lv, F. Y. Lee, T. Liu, A. I. Cooper and X. S. Zhao, J. Am. Chem. Soc., 2007, 129(46), 14213–14223; (c) G. Budroni, A. Corma, H. Garcia and A. Primo, J. Catal., 2007, 251(2), 345–353. 3 (a) C. H. Kwon, D. H. Yun, H. –K. Hong, S.-R. Kim, K. Lee, H. Y. Lim and K. H. Yoon, Sens. Actuators, B, 2000, 65, 327–330; (b) R. K. Sharma, P. C. Chan, Z. Tang, G. Yan, I.-M. Hsing and J. K.-O. Sin, Sens. Actuators, B, 2001, 72, 160–166; (c) P. Montmeat, C. Pijolat, G. Tournier and J. P. Viricelle, Sens. Actuators, B, 2002, 84, 148–159; (d) A. Cabot, J. Arbiol, A. Cornet, J. R. Morante, F. Chen and M. Liu, Thin Solid Films, 2003, 436, 64–69; (e) J. Arbiol, A. Cabot, J. R. Morante, F. Chen and M. Liu, Appl. Phys. Lett., 2002, 81, 3449–3451. 4 S. Jansat, K. Pelzer, J. Garcıa-Ant on, R. Raucoules, K. Philippot, A. Maisonnat, B. Chaudret, Y. Guari, A. Medhi, C. Reye and R. J. P. Corriu, Adv. Funct. Mater., 2007, 17, 3339–3347. 5 (a) K. Pelzer, K. Philippot, B. Chaudret, W. Meyer-Zaika and G. Schmid, Z. Anorg. Allg. Chem., 2003, 629, 1217–1222; (b) M. Boutros, A. Denicourt-Nowicki, A. Roucoux, L. Gengembre, P. Beaunier, A. Gedeon and F. Launay, Chem. Commun., 2008, (25), 2920–2922. 6 (a) Y. Guari, C. Thieuleux, A. Mehdi, C. Reye, R. J. P. Corriu, S. Gomez-Gallardo, K. Philippot, B. Chaudret and R. Dutartre, Chem. Commun., 2001, 1374–1375; (b) Y. Guari, C. Thieuleux, A. Mehdi, C. Reye, R. J. P. Corriu, S. Gomez-Gallardo, K. Philippot and B. Chaudret, Chem. Mater., 2003, 15, 2017–2024; (c) Y. Guari, K. Soulantica, K. Philippot, C. Thieuleux, A. Mehdi, C. Reye, B. Chaudret and R. J. P. Corriu, New J. Chem., 2003, 27, 1029–1031; (d) V. Hulea, D. Brunel, A. Galarneau, K. Philippot, B. Chaudret, P. J. Kooyman and F. Fajula, Microporous Mesoporous Mater., 2005, 79, 185–194. 7 T. Oonak, K. Hashimoto, H. Kominanmi, Y. Kear and Y. Matsubara, Catal. Today, 2006, 111, 354–360; D. J. Mihalcik and W. Lin, Chem. Cat. Chem., 2009, 1, 406–413. 8 D. J. Mihalcik and W. Lin, Chem. Cat. Chem., 2009, 1, 406–413. 9 G. A. Somorjai, F. Tao and J. Y. Park, Top. Catal., 2008, 47, 1–14. 10 B. Lee, H. Zhu, Z. Zhang, S. H. Overbury and S. Dai, Microporous Mesoporous Mater., 2004, 70, 71–80. 11 H. Song, R. M. Rioux, J. D. Hoefelmeyer, R. Komor, K. Niesz, M. Grass, P. Yang and G. A. Somojai, J. Am. Chem. Soc., 2006, 128, 3027–3037.
9530 | J. Mater. Chem., 2010, 20, 9523–9530
12 G. Budroni and A. Corma, Angew. Chem., Int. Ed., 2006, 45, 3328–3331. 13 (a) A. Maisonnat, K. Philippot, B. Chaudret, S. Jansat, J. GarciaAnt on, F. Turpin, Y. Guari, R. Corriu, V. Matsura, PCT/FR2007/ 000905; (b) V. Matsura, Y. Guari, C. Reye, J. P. Corriu, M. Tristany, S. Jansat, K. Philippot, A. Maisonnat and B. Chaudret, Adv. Funct. Mater., 2009, 19, 3781–3787. 14 K. Philippot and B. Chaudret, in Comprehensive Organometallic Chemistry III, ed. R. H. Crabtree and M. P. Mingos, Elsevier, Amsterdam area, The Netherlands, Vol. 12 – Applications III: Functional Materials, Environmental and Biological Applications, ed. Dermot O’Hare, 2007, Ch. 03, pp. 71–99. 15 (a) F. Lerouge, G. Cerveau and R. J. P. Corriu, New J. Chem., 2006, 30, 1364–1376; (b) R. J. P. Corriu, A. Mehdi and C. Reye, J. Mater. Chem., 2005, 15, 4285–4294; (c) B. Boury and R. J. P. Corriu, Chem. Rec., 2003, 3, 120–132. 16 (a) C. Pan, K. Pelzer, K. Philippot, B. Chaudret, F. Dassenoy, P. Lecante and M.-J. Casanove, J. Am. Chem. Soc., 2001, 123(31), 7584–7593; (b) K. Pelzer, O. Vidoni, K. Philippot, B. Chaudret and V. Colliere, Adv. Funct. Mater., 2003, 13(2), 118–126; (c) K. Pelzer, K. Philippot and B. Chaudret, Z. Phys. Chem., 2003, 217, 1539– 1547; (d) K. Pelzer, B. Laleu, F. Lefebvre, K. Philippot, B. Chaudret, J.-P. Candy and J.-M. Basset, Chem. Mater., 2004, 16, 4937–4941; (e) S. Jansat, D. Picurelli, K. Pelzer, K. Philippot, M. G omez, G. Muller, P. Lecante and B. Chaudret, New J. Chem., 2006, 30, 115–122; (f) M. Tristany, B. Chaudret, P. Dieudonne, Y. Guari, V. Matsura, M. Moreno-Ma~ nas, K. Philippot and R. Pleixats, Adv. Funct. Mater., 2006, 16, 2008–2015. 17 (a) E. Ramirez, L. Erades, K. Philippot, P. Lecante and B. Chaudret, Adv. Funct. Mater., 2007, 17, 2219–2228; (b) C. Pan, K. Pelzer, K. Philippot, B. Chaudret, F. Dassenoy, P. Lecante and M.J. Casanove, J. Am. Chem. Soc., 2001, 123, 7584–7593. 18 (a) A. Abad, P. Concepci on and A. Corma, Angew. Chem., Int. Ed., 2005, 44, 4066–4069; (b) P. Korovchenko, C. Donze, P. Gallezot and M. Besson, Catal. Today, 2007, 121, 13–21; (c) P. Maity, C. S. Gopinath and G. K. Lahiri, Green Chem., 2009, 11, 554–561; (d) A. Villa, D. Wang, N. Dimitratos, D. Su, V. Trevisan and L. Prati, Catal. Today, 2010, 150, 8–15; (e) Y. Chen, H. Lim, Q. Tang, Y. Gao, T. Sun, Q. Yan and Y. Yang, Appl. Catal., A, 2010, 380, 55–65; (f) M. Zahmakiran and S. Ozkar, J. Mater. Chem., 2009, 19(38), 7112–7118; (g) X. Yang, X. Wang and J. Qiu, Appl. Catal. A, 2010, 382, 131–137.
This journal is ª The Royal Society of Chemistry 2010
