Direct hydrothermal synthesis of iron-containing mesoporous silica SBA-15:
potential as support for gold nanoparticles
Ying Li,a Yejun Guan,b Rutger A. van Santen,b Patricia J. Kooyman,c Iulian Dugulan,d Can Li,e and Emiel J. M. Hensena,*
a) Institute of Catalysis, Zhejiang University of Technology, Zhejiang Key Laboratory of Green-chemistry Synthesis Technology, Hangzhou 310014, PR China
b) Schuit Institute of Catalysis, Eindhoven University of Technology, P.O. Box, 513, 5600 MB, Eindhoven, The Netherlands.
c) DCT/TNW, Delft University of Technology, Julianalaan 136, 2628 BL, Delft, The Netherlands.
d) R3 Reactor Institute, Delft University of Technology, Mekelweg 15, 2628 BL, Delft, The Netherlands.
e) State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Zhongshan Road 457, Dalian, China.
* Corresponding Author: Professor Emiel J. M. Hensen Fax: +31-40-2455054 Tel: +31-40-2475178 E-mail: e.j.m.hensen@tue.nl
Abstract
The preparation of mesoporous silica SBA-15 with high iron loadings (14-90 wt% Fe2O3) as a suitable support for gold nanoparticles to be used in CO oxidation catalysis has been investigated. The support materials were prepared by a direct hydrothermal two-step pH adjusting method which consisted of the formation of the silica mesophase at low pH and the inclusion of Fe3+ at varying pH in the range of 2-7. The materials were characterized by XRD, SEM and TEM/EDX, N2 porosimetry and
57
Fe Mössbauer spectroscopy. At relatively low Fe loading, the SBA-15 structure is maintained and iron is predominatly surface grafted to the silica surface. Such a mesoporous silica can accommodate up to 40 wt% Fe2O3 with a surface area of 460 m
2
/g With increasing Fe content, precipitation of iron hydroxides competes with the surface grafting process and the resulting materials are an intimate mixture of hematite particles embedded in an Fe/SBA-15 matrix. A too high Fe3+ content in the synthesis gel results in a high rate of precipitation and impedes the formation of the silica mesophase. The stabilization of the mesophase at pH 7 is proposed to involve interactions of the surfactant with a surface grafted Fe3+ silica phase. The use of an SBA-15 which contains mainly surface grafted Fe3+ as a support for gold nanoparticles results in a more active catalyst for CO oxidation than gold supported by SBA-15 or iron oxide particles.
1. Introduction
Synthesis of nanosized materials has become an important area in materials science in the past decade, because of optical, electronic, magnetic and especially catalytic properties that can differ significantly from their bulk counterparts [1-4]. Ordered mesoporous materials with their intrinsically high surface areas are excellently suited for this purpose. Compared to conventional carrier materials, ordered mesoporous solids have the advantage of stabilizing dispersed metal or metal oxide nanoparticles [5, 6] which especially important when employed as catalysts or catalyst supports. Nanosized iron oxide particles have been reported as active catalysts for various reactions such as methanol decomposition, partial oxidation of methane and methanol and Friedel–Crafts alkylations [7]. Gold nanoparticles dispersed over iron oxide show synergy in the oxidation of CO [8-12]. Hence, new approaches to prepare nanosized iron oxide materials are of particular interest [13, 14].
Most attempts to prepare silica-supported metal oxides or metal catalysts have centered around wet impregnation methods which are somewhat limited by the tendency of the metal or metal oxides to agglomerate on the outer surface and the difficulties to control the growth of desired nanoparticles. High loadings of metal oxides are difficult to obtain and such methods have the risk of clogging of the pores [15-18]. Studies of the direct synthesis of mesostructured iron-containing silicates have focused on routes involving basic conditions [19, 20]. Reports of the direct synthesis of iron-containing SBA-15 mostly deal with relatively low iron loadings [21, 22]. Recently, Liu et al. [13] reported the synthesis of iron oxide nanoparticles within the mesopores of SBA-15 through a conventional hydrothermal approach. The Fe content was limited to a Fe/Si ratio of 0.10 in the synthesis gel. Cannas et al. [14] reported the synthesis of γ-Fe2O3 nanoparticles homogeneously dispersed in an amorphous silica matrix. The highest iron oxide loading reported for amorphous silica was 28.5 wt.% Fe2O3. Kang et al. [25] synthesized monodisperse crystalline manganese ferrite nanocrystals through thermal decomposition of metal-surfactant complexes followed by mild oxidation. γ-Fe2O3 nanocrystallites dispersed in an amorphous silica matrix were also synthesized by mechanical activation of a chemistry-derived precursor at room temperature [26]. Zhao et al. reported the simultaneous formation of hybrid nanocomposites from iron oxides and organosilsesquioxanes [27,28]. Another approach to obtain iron oxide catalysts is to synthesize mesoporous metal oxides with high surface area. Despite prior reports [6,28,29], the beneficial textural properties are most often not maintained when the templating agents
are removed. Recently, Jiao et al reported the synthesis of mesoporous iron oxide materials with a crystallized wall by using mesoporous silica as hard templates [30].
Herein we report a simple hydrothermal approach to obtain highly dispersed iron oxide in mesoporous SBA-15 silica as well as high surface area mesoporous iron oxides stabilized by a small amount of silica. We choose SBA-15 as the silica matrix because it has a higher (hydro)thermal stability than MCM-41 [31]. The synthesis of SBA-15 under acidic conditions precludes the incorporation of transition metals. A solution is a two-step pH adjustment method, initially developed to prepare SBA-15 with improved hydrothermal stability [32]. Briefly, the mesoporous material is synthesized at low pH, while crystallization is promoted at higher pH values by an abrupt pH change. This method has also been successfully employed to incorporate Al and Ti in SBA-15 [33,34]. In the present investigation, we employ this two-step pH strategy to incorporate iron in silica in the presence of a mesoporogen. Maintaining the mesophase requires careful finetuning of the synthesis conditions. A subset of these materials is used as a catalyst support to disperse gold nanoparticles.
2. Experimental Section 2.1. Preparation
Our synthesis approach consists of the addition of the desired amount of Fe3+ ions to the highly acidic solution containing the ingredients to prepare SBA-15. Concentrated ammonia is added to precipitate iron ions in the mesophase precursor. A typical synthesis of FeOx/SBA-15 is as follows: 2 g of Pluronic P123 (BASF triblock copolymer) were dissolved in 75 mL 1.6 M HCl solution to obtain solution A. An amount of Fe2(SO4)3·5H2O (Aldrich, >98%) to arrive at Fe/Si molar ratios of 0.15, 0.5 or 1.0 was added to solution A. Iron nitrates as the Fe precursor should be avoided, because hydrothermal treatment would result in the release of large amounts of nitric oxide. Subsequently, 3.2 mL of TMOS (tetramethoxysilane, Aldrich, 99%) were added drop wise to the resulting mixture. After this, the solution was stirred vigorously for 2 h at 40 °C and ammonia (ca. 16 mL-17 mL) was added very slowly to adjust the pH to 7.0. Usually, this period took 2 h to allow full precipitation of iron. The pH of the mixture was regularly evaluated by indicator paper. After stirring the resulting mixture at 40 °C overnight, it was transferred into an autoclave 100 mL to age for 24 h at 100 °C. The resulting solid was filtered, washed, dried at 60 °C for 15 h and finally calcined at 500 °C for 10 h. The samples with Fe/Si molar ratios in the gel of 0.15, 0.5 and 1.0 are denoted TP0.15, TP0.5 and TP1.0, respectively.
Another set of samples was prepared by changing the pH of the second step. To this end, various amounts of ammonia, i.e., ca. 12 mL, 14 mL, 16 mL, and 17 mL, were added to obtain final pH 3, 4, 6, and 7 in the solution, respectively. The Fe/Si molar ratio was fixed at 2.0. All the other steps were the same as described above. These samples are denoted TP2-x, with x being the final pH.
Mesoporous iron oxides stabilized by a small amount of silica were prepared via a similar procedure. The amount of the iron precursor Fe2(SO4)3.5H2O was fixed at 10.3 g with varying amounts of the silica precursor, that is 0.4, 0.8, and 1.6 mL. The final pH was fixed at 7.0. The other steps were the same as described above. Samples are denoted TP4, TP8, and TP16 for the TMOS amounts of 0.4, 0.8, and 1.6 mL, respectively. For comparison, siliceous SBA-15 and iron oxide were also prepared following the same procedure without adding Fe or silica precursor, respectively. These two materials are denoted TP-SBA-15 and TP-Fe2O3.
Additional samples were prepared for use in the catalysis study. The preparation of samples 1 and 3 from the study of Li et al. [22] were repeated as well as the preparation of a standard SBA-15.
The supported gold catalysts were prepared by a deposition-precipitation method using an aqueous solution of Au(en)2Cl3 [35]. Typically, an amount of Au(en)2Cl3 was dissolved in deionized water. The pH value was adjusted to 12 by addition of 1 M NaOH at room temperature. Then, 1 g of support was added to this solution and the pH of this suspension was again adjusted to 12 by addition of aqueous NaOH. After aging for 2 h, the catalyst was filtered and washed with deionized water until there were no chlorine anions in the filtrate. The catalyst was finally dried at 110 °C overnight and calcined at 400 °C for 4 h. The gold loading was about 1.5 wt% Au(en)2Cl3 (en = ethylenediamine) was synthesized in the following way. A stoichiometric amount of HAuCl4 (Aldrich, 99%) and ethylenediamine (Aldrich, 99%) were dissolved in water and stirred. The precipitate was filtered and dried in vacuum at 40 ºC.
2.2 Characterization
Powder X-ray diffraction patterns were recorded on a Rigaku D/Max-2500/pc powder diffraction system using Cu Kα radiation (40 kV and 100 mA) over the range 0.5° ≤ 2 θ ≤ 10° (low angle) and 10° ≤ 2 θ ≤ 80° (high angle). N2 sorption isotherms were determined at 77 K on a Micromeretics ASAP 2020 system in static measurement mode. The samples were outgassed at 523 K for 10 h prior to sorption measurements. The pore size distribution curves were calculated by analyzing the desorption
branch of the isotherm by the BJH (Barrett-Joyner-Halenda) method. The t-plot method was used for the analysis of the micropores and specifically, the equation t = [13.9900 / (0.0340 - log(p/p0))]0.5 was employed. The thickness range t was chosen from 0.3500 to 0.5000 nm. The Fe and Si content of the calcined materials was determined by Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES on SpectroCiros CCD) after dissolving the sample in a mixture of HF and HNO3.
SEM was performed using a Philips environmental scanning electron microscope XL-30 ESEM FEG (Philips) in high-vacuum mode using acceleration voltages of 3 kV (low-voltage SEM, LVSEM) and a secondary electron (SE) detector. Transmission electron microscopy (TEM) was performed using a FEI Tecnai electron microscope with a field emission gun as the source of electrons operated at 200 kV. Samples were mounted on Quantifoil carbon polymer supported on a copper grid by placing a few droplets of a suspension of ground sample in ethanol on the grid, followed by drying at ambient conditions. EDX elemental analysis was performed using an Oxford instruments EDX system. Some TEM measurements were carried out on a FEI Tecnai electron microscope with a LaB6 filament operated at 200 kV.
For some samples, transmission 57Fe Mössbauer spectra were collected at room temperature with a conventional constant-acceleration spectrometer using a 57Co(Rh) source. Velocity calibration was carried out using an α-Fe foil. The Mössbauer spectra were fitted using the Mosswinn 3.0i program.
2.3 Catalytic activity measurements
The catalytic activity of the gold loaded support materials for the oxidation of carbon monoxide was measured in a single-pass atmospheric gas-phsase reactor. To this end, ca. 0.02 g of the powdered catalyst was diluted with SiC to meet plug flow requirements and kept between quartz wool plugs in a 4 mm quarts glass reactor. A well-calibrated online GC was applied for analysis of the gas phase components. The feed consisted of 1 vol% CO and 0.5 vol% O2 in He. The total gas flow was 100 mL/min. Prior to reaction, the catalysts were calcined in artificial air under heating from room temperature to 200 °C.
3. Results and Discussion Influence of iron content
The physicochemical properties and the composition of the materials are listed in Table 1. In the syntheses involving Fe/Si molar ratios of 2.0 in the initial gel, the calcined materials have pore diameters above 6 nm with surface areas in excess of 280 m2
/g in spite of the relatively high Fe contents. For instance, TP2-7 has a surface area of 335 m2
/g with an iron oxide content over 60 wt%. Fig. 1 shows the low- and high-angle XRD patterns of calcined TP-SBA-15 and Fe-SBA-15 with Fe/Si ratios of 0.15, 0.5, and 1.0. The XRD patterns of TP-SBA-15, TP0.15, and TP0.5 show three well-resolved diffraction lines, corresponding to (100), (110) and (200) reflections of mesoporous materials with the 2D hexagonal structure. Thus, these samples exhibit a highly ordered structure of mesopores. As previously reported [22], the addition of iron salts to the synthesis gel strongly improves the structural order in SBA-15.The hexagonal arrangement of mesopores is maintained up to an iron oxide content of 40 wt% Fe2O3 (TP0.5). The repeating distance a0 from the (100) reflection is slightly higher for TP0.15 than for TP-SBA-15. This may indicate that iron has been incorporated into the silica walls or is dispersed over the surface. Low-angle reflections in the XRD patterns for TP1.0 are absent, indicating that a further increase of the Fe loading results in loss of the ordered mesoporosity.
The high-angle diffraction pattern of TP0.15 (14 wt% Fe2O3) does not exhibit the typical features of iron oxide aggregates, which points to the well-dispersed nature of the iron oxide particles in the SBA-15 material. TP0.5 shows typical diffraction lines of hematite. The size of the iron oxide aggregates determined from line broadening of the diffraction peaks is around 20 nm which agrees reasonably well with the transmission electron micrographs (vide infra). From the XRD intensities , about a quarter of Fe is present as hematite particles in TP0.5. Thus, a substantial amount of the Fe3+ forms iron oxide particles too small to be detected by XRD. No diffraction lines are observed for TP1.0.
Fig. 2 shows the N2 sorption isotherm and the pore size distribution of TP-SBA-15, TP0.15, TP0.5, and TP1. The isotherm of TP-SBA-15 can be classified as type IV with an H1 hysteresis loop, which is typical for SBA-15 materials. The well-defined step that occurs at relatively high pressures (p/p0 = 0.7-0.9) indicates the uniformity of the pores. The hysteresis at higher p/p0 value than for typical SBA-15 [31] points to the presence of larger pores. The pore size distribution centers around 10 nm. The isotherm of TP0.15 is similar to that of SBA-15, but different from the one of TP-SBA-15 in that an obvious hysteresis of the H3 type occurs at p/p0 of 0.8. This hysteresis indicates the condensation and evaporation of N2 within interparticle voids or of some impurity phase, such as lamellar mesostructures
that are sometimes present in liquid-crystal templated materials [36].The very different shape of the isotherm points to the less ordered arrangement of pores in TP1.0.
The micropore surface area for TP-SBA-15 is 35 m2
/g. The micropore surface area increases to 54 and 117 m2
/g for TP0.15 and TP0.5, respectively, whereas TP1.0 does not contain micropores. The presence of micropores in SBA-15 is well-established and usually correlates to the degree of order of the sample [37]. These results also show that such narrow pores are not blocked by iron oxide particles, even for materials with a relatively high iron loading. The absence of micropores in TP1.0 is due to the absence of the ordered hexagonal structure. However, due to the presence of the template, the surface area of this material is still higher than it would be without template (vide infra).
Fig. 3 shows scanning electron micrographs of TP-SBA-15, TP0.15, TP0.5, and TP1.0. TP-SBA-15 exhibits the typical rod-like morphology of SBA-15 silica. This morphology is maintained upon introduction of a small amount of iron in TP0.15. A further increase of the iron content leads to partial loss of this morphology in TP0.5. The morphology of TP1.0 is completely different and made up of agglomerated particles of about 5 µm.
Representative transmission electron micrographs of calcined TP0.15, TP0.5, and TP1.0 are shown in Fig. 4. The micrographs of TP0.15 show well-ordered hexagonal arrays of mesopores. Only a very small number of iron oxide nanoparticles are observed in the electron micrographs in spite of the high Fe loading. EDX analysis of various regions points to a rather homogeneous distribution of iron. Micrographs of TP0.5 show well-ordered hexagonal arrays of mesoporous channels with a large amount of nanosized iron oxide particles dispersed throughout the material. EDX analysis of dark and lighter regions in this material confirm that the former regions contain more Fe than the latter. Thus, besides dispersed iron oxide embedded in SBA-15, iron oxide aggregates are present that are well integrated into the SBA-15. Careful analysis of the electron micrographs and EDX spectra of many areas suggests that iron is homogeneously distributed in TP1.0.
Fig. 5 shows the 57Fe Mössbauer spectra for TP0.15 and TP1.0. The fit parameters are given in Table 2. The spectrum of TP0.15 was fitted with a distribution of hyperfine fields. It contains a considerable contribution (~80 %) around 41 T representing very disordered Fe3+
and a lower contribution with a hyperfine field around 5 T of doublets that are assigned to superparamagnetic nanosized particles. The quadrupole (QS) value of -0.06 mm/s is smaller than that characteristic for α-Fe2O3 (-0.2 mm/s), but this can be caused by an ill-defined electric field gradient in small disordered particles. The spectrum
of TP1.0 is dominated by a symmetrical quadrupole doublet (~73 %) assigned to non-interacting superparamagnetic species [38]. The room temperature Mössbauer spectrum of α-Fe2O3 with particles smaller than 13.5 nm usually contains only quadrupole doublets corresponding to superparamagnetic particles [39]. The presence of a small contribution of a well-defined magnetic sextuplet with hyperfine parameters characteristic of larger α-Fe2O3 indicates the presence of larger iron oxide particles as well. The mechanism of SBA-15 formation involves S0..H+..X-..I+ interactions with S0 the non-ionic surfactant, I+ the protonated silanol group and X- the counteranion. Fe3+ is predominantly present as Fe(H2O)6
3+
ions under the acidic conditions. The amount of Fe3+, which can be incorporated in the mesoporous silica walls, is very low [22], because of the positive charge of the silica surface below pH 2 [40]. At pH values above silica’s point of zero charge of 2, the surface of the mesostructured silica becomes neutral. Appreciable negative charge only developes above pH 6 for silica [40]. With increasing pH, the decrease of S0..H+..I0 interactions impedes the stability the mesophase. With Fe3+, reactions of the neutral surface with iron complexes in the solution need to be considered. Firstly, the iron-aquo complexes hydrolyze with increasing pH. The higher the pH and the higher the iron concentration, the more extensive will be the formation of polymeric forms to such extent that iron hydroxides precipitate out of the solution. Secondly, the silica surface reacts with hydrolyzed metal-aquo complexes through hydrolytic adsorption [41]. At relatively low pH and low Fe3+ concentration, the simplest reaction involves ≡SiOH + Fe(H2O)6OH
2+ → ≡SiOFe(H 2O)5
2+
+ H2O. At higher pH and higher Fe3+ concentration, reactions with Fe(H2O)(OH)2
+
are included as well as precipitation of iron hydroxides. These grafting reactions proceed during ammonia addition. At pH 7, the surfactant-silica interactions in the mesophase are dominated by S0..Fe3+..I- interactions. This provides an explanation for the salt effect in SBA-15 synthesis at pH > 2 [22]. Without Fe3+, the cation concentration becomes too low at pH 7 to synthesize good quality SBA-15, as is exemplified by TP-SBA-15. Stucky and co-workers [42] have studied the effect of pH on the mesophase formation in detail. In comparison, a similar synthesis from a gel containing Fe3+
results in a material with improved ordered mesoporosity. The gel and product Fe/Si ratios for the materials prepared at pH 7 are very similar, because of the grafting and precipitation reactions. The TP0.15 precursor contains Fe3+
ions dispersed over the walls of the mesoporous silica. The iron cations are mainly involved in S0
..Fe3+ ..I
interactions in this case. Calcination, which is required to remove the mesoporogen, results in two competing effects: (i) the diffusion of some Fe into the silica [7,40] and (ii) the aggregation of surface iron to iron oxide particles.
Aggregation is evident from the small contribution around 5 T in the Mössbauer spectrum of TP0.15. The electron micrographs also evidence the presence of nanometer sized iron oxide particles. The main contribution in the Mössbauer spectrum, however, arises from very distorted Fe3+
assigned to surface grafted Fe3+
. An increase of the Fe3+
concentration in the gel leads to increased precipitation of iron hydroxides. Calcination will convert these hydroxides to iron oxide particles. TP0.5 appears to be a composite consisting of SBA-15 containing dispersed Fe, very similar to TP0.15, as well as a small fraction of hematite particles. Electron micrographs evidence that these hematite particles are intimately mixed throughout SBA-15. TP1.0 does not exhibit ordered mesoporosity, which shows that the mesophase is not stable under these conditions. Contrary to our expectation, small iron oxide particles are formed. We explain this by fast precipitation of iron hydroxides at high Fe3+ concentration. This agrees Mössbauer fit results and the absence of diffraction lines of iron oxides in the XRD patterns after calcination. The strong decrease of the Fe3+ concentration upon iron hydroxide precipitation should decrease the ionic strength of the solution to that extent that the mesophase is destabilized. When the synthesis of TP1.0 was repeated in the absence of the mesoporogen, a material with less favorable properties was obtained (surface area 100 m2/g; pore volume 0.09 ml/g). The much higher surface area and pore volume of TP1.0 suggest that interactions exist between the mesoporogen, iron hydroxides and the silica phase, resulting in FeOx/silica with favorable textural properties.
Effect of pH
As the presence of an abundant amount of positive charges in the precursor solution appears pivotal to the formation of well-ordered mesoporous silica structures, the effect of acidity in the second step was studied in more detail. The Fe/Si ratio in the synthesis gel was adjusted to 2. TP2-7 contains a somewhat higher amount of Fe (Fe/Si = 1.2) than TP1.0 (Fe/Si = 0.9). Thus, not all Fe3+ has been integrated into TP2-7. The lower the pH of the second step, the lower is the final Fe content. As the acidity increases, the contribution of precipitation of iron hydroxides becomes smaller. Fig. 6 shows the low- and high-angle XRD patterns of the samples synthesized at various pH values. TP2-3 exhibits typical reflections of the well-ordered hexagonal arrangement of mesopores. The intensities of the reflections decrease with increasing pH. The XRD pattern of TP2-7 shows only weak reflections at low angle. The patterns at high angle show sharp reflections belonging to hematite for all TP2-x materials. Notably, the intensities of these reflections are pronouncedly lower for TP2-7 than for the TP2-x
samples prepared at lower pH. Thus, TP2-7 resembles TP1.0, although the mesostructure of TP2-7 appears to be a bit more preserved. Fig. 7 shows the N2 sorption isotherms of TP2-x. The isotherms show a hysteresis in two steps for the samples prepared at pH 6 and lower. As mentioned before, the hysteresis at p/p0 = 0.6-0.7 is typical for ordered mesoporous silica (pore size 6-7 nm). The second hysteresis at higher relative pressure corresponds to interparticle pores (pore size 30-60 nm). Table 1 shows that the micropore surface area of these materials decreases with increasing pH.
Mesoporous iron oxides stabilized by a small amount of silica
Results thus far indicate that iron oxide/silica composites can be prepared with reasonably high surface areas at Fe2O3 contents up to 60 wt%. We further attempted to increase the Fe
3+
content to arrive at high surface area iron oxides via a similar strategy. To this end, the amount of the iron precursor was kept constant whilst decreasing the amount of the silica precursor. The physicochemical parameters of such samples materials are listed in Table 1. TP4 has a slightly higher Fe content than TP2-7 prepared under similar conditions. A further increase of the Fe/Si ratios results in a relatively small increase of the Fe loading from 67 to 89 wt% accompanied by a decrease of the surface area from 259 to 89 m2/g. The product Fe/Si ratio of TP4 is 1.5 which is only marginally higher than that of TP1.0 and TP2-7. An increase of the gel Fe/Si ratio to 8 or 16 (TP8 and TP16) gives rise to product Fe/Si ratios of 3.2 and 6.1, respectively. This indicates that a substantial part of Fe is not built into the final structure.
Fig. 8 shows the high angle XRD patterns for these samples. No features were present in the low-angle XRD patterns. TP8 shows typical XRD patterns of magnetite (γ-Fe2O3), whereas the diffraction patterns of TP16 and TP-Fe2O3 evidence the formation of hematite (α-Fe2O3). The iron oxide phase is controlled by the calcination temperature [32]. Magnetite γ-Fe2O3 is a technologically important magnetic material that has a wide range of applications in information storage and catalysis [26]. It is very difficult to synthesize a single phase nanocrystalline γ-Fe2O3 via either conventional or chemistry-based processing routes due to the fact that the nanocrystallites tend to aggregate and coarsen during calcination [43]. The other critical obstacle for formation of a single γ-Fe2O3 phase is the conversion from γ-Fe2O3 to α-Fe2O3 (hematite), the latter being thermodynamically stable over a large temperature range of 300-600°C. Here, the only difference between TP8 and the other samples is the amount of silica. This suggests that a small amount of silica can prevent the transformation of magnetite to hematite and thus plays a determining role in the stabilization of magnetite. Fig. 9 shows the N2
sorption isotherms of TP4, TP8 and TP16, and TP-Fe2O3. These isotherms can be classified as well-behaved type IV with an H1 hysteresis loop, which is typical for mesoporous materials with large pores [44]. The pore size of the materials increases from 6 nm to 21 nm when the Fe/Si increases from 4 to 16. In contrast, TP-Fe2O3 without silica does not show such a well-behaved isotherm. The hysteresis at high relative pressure is due to interparticle pores. Fig. 10 shows representative transmission electron micrographs for TP4, TP8, TP16, and TP-Fe2O3. The morphology differs substantially for these samples. TP4 is a composite of iron oxides and silica, similar to TP1.0 and TP2-7. TP8 still resembles TP4, although now some larger iron oxide particles are visible. This may suggest that TP4, TP1.0 and TP2-7 also contain very small magnetite particles and further experiments are necessary to elude on this. TP16 shows a needle-like morphology of hematite particles. TP-Fe2O3 without silica consists of uniform crystals of about 100 nm.
Gold supported on FeOx/SBA-15
A set of gold nanoparticles catalysts was prepared by deposition-precipitation of Au(en)2Cl3. The following materials were chosen: SBA-15, TP1-1.5 and TP5-1.5, TP0.15, TP0.5, and TP-Fe2O3. Fig. 11 shows representative transmission electron micrographs of the gold supported catalysts. Readily apparent are the gold nanoparticles embedded in the well-ordered mesopores of SBA-15, TP1-1.5, TP5-1.5 and TP0.15. The average gold particle size is 4.9 ± 1.3 nm for Au/SBA-15, 5.8 ± 1.9 nm for Au/TP1-1.5, 6.8 ± 1.9 nm for Au/TP5-1.5, and 8.0 ± 1.8 nm for Au/TP0.15. The former three particle sizes correspond quite well with the pore dimensions. The larger gold particle size for Au/TP0.15 indicates that the mesostructure is much less maintained than in the samples prepared at low pH. The average gold particle size for Au/TP0.5 is 10.4 ± 2.2 nm. The electron micrographs show that the gold nanoparticles are dispersed throughout the mesoporous material. Finally, Au/Fe2O3 contains larger gold particles of 15 ± 5.5 nm and they are present at the surface of the hematite crystals. The larger spread is due to the absence of mesopores to limit the growth of the gold particles upon calcination.
Table 3 lists the catalytic activities of this set of supported gold catalysts for CO oxidation. Clearly, synergy occurs between Fe and Au, as all gold catalysts supported on iron containing oxides are more active than Au/SBA-15. The quasi turnover frequencies were corrected for differences in the gold dispersion to give the surface turnover frequencies, because the gold particle size differed quite substantially between the catalysts. Note that this set of catalysts has a much lower activity than
optimized Au/FeOx catalysts [12]. The main reason is that calcination at high temperature is required to remove the mesoporogen. This calcination step dehydrates the iron hydroxides to iron oxides. The extent of the synergy in the Au/FeOx system is very sensitive to the pre-treatment procedure and especially iron hydroxides show very high synergy with gold [9]. This synergy between gold and metal oxides is well-known [8-13], but not understood in detail. Gupta et al. [45] proposed that the oxidation of CO on Au/Fe2O3 involves a redox mechanism using lattice oxygen from iron oxide. Others have stressed that surface hydroxyl groups of the iron oxide may participate in the CO adsorption and explain the synergy [46,47]. The present data suggest the latter to be the more reasonable interpretation, as catalysts such as Au/TP0.15 and Au/TP0.5 hardly contain iron oxide particles, yet are the most active in CO oxidation. The silica in TP1-1.5, TP5-1.5, and TP0.15, contains a surface layer of Fe3+ ions and its surface coverage increases with Fe content. Thus, likely the surface hydroxyl groups associated with this type of iron, as suggested above, promotes the gold catalyzed CO oxidation. Au/TP0.5 has a lower activity. The support material is a composite of an iron-containing SBA-15 and hematite particles. We do not know the Fe content of the SBA-15 part of this material. If this would be quite low, a substantial part of the gold phase may be in contact with a silica surface containing little Fe3+ and this provides a reasonable explanation for the lower activity. Indeed, when gold is supported by hematite as in TP-Fe2O3, a more substantial synergy is observed than for Au/TP0.5.
Thus, the most optimal catalyst is Au/TP0.15. The TOF increases in the order Au/SBA-15 < Au/TP1-1.5 < Au/TP5-Au/TP1-1.5 < Au/TP0.15. This correlates very well with the expected surface density of very well dispersed Fe3+. It is interesting to estimate the monolayer coverage of Fe on a SBA-15 surface. Assuming that Fe3+ is grafted to two nearby hydroxyl groups of SBA-15 and that a conservative estimate for its hydroxyl density is 4 OH/nm2, SBA-15 may accommodate about 19 wt% Fe3+ at a surface area of 1000 m2
/g. Obviously, the materials prepared at low pH maintain such high surface area, but the Fe3+
loading is limited. The Fe content of TP0.15 (14 wt% Fe) is above the monolayer coverage of Fe3+
for a silica with a surface area of 470 m2
/g (9 wt% Fe). However, if one assumes that the surface contains Fe grafted with one bond to the silica, the coverage can be as high as 18 wt%. These estimates suggest that the highest activity of Au/TP0.15 is due to the near complete coverage of its silica surface by Fe3+
.
For gold catalyzed CO oxidation, a good Fe-containing silica support prepared in this manner should therefore have a high surface area that is uniformly covered by Fe3+
high-quality SBA-15 prepared under acidic conditions. These conditions, however, are not conducive to high Fe content. Higher coverages can be achieved by increasing the pH in the synthesis of Fe/SBA-15. A drawback is that this impairs the mesostructure and less favorable textural properties result. The synthesis conditions for TP0.15 form a reasonable compromise to obtain a reasonably high surface area by maintaining the mesophase, to fully cover the silica surface with Fe3+ and to largely avoid hematite particles by precipitation of iron hydroxides. A suggestion from this work is that surface grafting of Fe to a conventional silica may provide the same benefits for synergy in gold catalysis. An additional benefit is that calcination to remove the mesoporogen is not required [9].
4. Conclusions
A two-step pH adjustment method has been investigated to prepare a range of iron oxide/silica nanocomposites over a wide range of iron loadings. In the first step, the alkoxysilane precursor is hydrolyzed under acidic conditions and the initial mesostructure is formed, while an increase of the pH to 7 is employed to deposit the Fe on the surface. At relatively low Fe loading, SBA-15 is formed which contains Fe3+ dispersed over its inner surface. Grafting of hydrolyzed Fe-aquo complexes provides a way to stabilize the silica mesophase up to neutral pH. Further increasing the Fe concentration results in precipitation of iron hydroxides. At intermediate Fe loading, the hematite particles formed upon calcination are embedded in a SBA-15 matrix. Very high concentrations of Fe3+ result in fast precipitation and complete destabilization of the mesophase. Highly dispersed iron oxide particles on silica are formed instead. By further decreasing the silica content, iron oxides can be synthesized with reasonably high surface areas. The use of an SBA-15 which contains mainly surface grafted Fe3+ as a support for gold nanoparticles results in a more active catalyst for CO oxidation than gold supported by SBA-15 or iron oxide particles.
Acknowledgments
This work was financially supported by the Program for Strategic Scientific Alliances between China and Netherlands, funded by the Royal Netherlands Academy of Art and Science and the Chinese Ministry of Science and Technology. Part of the TEM measurements have been carried out with the support of the Soft Matter Cryo-TEM Research Unit, Department of Chemical Engineering and Chemistry, Eindhoven University of Technology.
References
1. Bronstein L. M. Top. Curr. Chem. 2003, 226, 55. 2. Taguchi, A.; Schuth, F. Micro. Meso. Mater. 2005, 77, 1.
3. Antonelli, D. M.; Ying, J. Y. Angew. Chem., Int. Ed. Engl. 1995, 34, 2014.
4. Yang, P.; Deng, T.; Zhao, D.; Feng, P.; Pine, D.; Chmelka, B. F.; Whitesides, G. M.; Stucky, G. D.
Science 1998, 282, 2244.
5. Lin, Y.; Wu, S.; Hung, Y.; Chou, Y.; Chang, C.; Lin, M.; Tsai, C.; Mou, C. Chem. Mater. 2006, 18, 5170.
6. Jiao, F.; Jumas, J.; Womes, M.; Chadwick, A. V.; Harrison, A.; Bruce, P. G. J. Am. Chem. Soc. 2006, 128, 12905.
7. Guan, Y.; Li, Y.; Hensen, E. J. M.; Santen, R. A.; Li, C. Catal. Lett. 2007, 117, 18. 8. Haruta, M. Catal. Today 1997, 36, 153.
9. Hodge, N.A.; Kiely, C. J.; Whyman, R.; Siddiqui, M. R. H.; Hutchings, G. J.; Pankhurst, Q. A.; Wagner, F. E.; Rajaram, R. R.; Golunski, S. E. Catal. Today 2002, 72, 133.
10. Chen, M. S.; Goodman, D. W. Catal. Today 2006, 111, 22.
11. Hashmi, A. S. K.; Hutchings, G. J. Angew. Chem. Int. Ed. 2006, 45, 7896.
12. Herzing, A. A.; Kiely, C. J.; Carley, A. F.; Landon, P.; Hutchings, G. J. Science 2008, 321, 1331. 13. Köhna, R.; Fröba, M. Anorg. Allg. Chem. 2003, 629, 1673.
14. Fabrizioli, P.; Bürgi, T.; Baiker, A. J. Catal. 2002, 206, 143.
15. Cannas, C.; Musu, E.; Musinu, A.; Piccaluga, G.; Spano, G. J. Non-Crystal. Solids 2004, 345 &
346, 653.
16. Abe, T.; Tachibana, Y.; Uematsu, T.; Iwamoto, M.; J. Chem. Soc., Chem. Commun. 1995, 1617. 17. Kohn, R.; Paneva, D.; Dimitrov, M.; Tsoncheva, T.; Mitov, I.; Minchev, C.; Froba, M. Micropor.
Mesopor. Mater. 2003, 63, 125.
18. Kohn, R.; Froba, M. Catal. Today 2001, 68 , 227.
19. Zhang, Q. H.; Yang, W.; Wang, X. X.; Wang, Y.; Shishido, T.; Takehira, K. Micropor. Mesopor.
Mater. 2005, 77, 223.
20. Hamdy, M. S.; Mul, G.; Wei, W. Anand, R.; Hanefeld, U.; Jansen, J. C.; Moulijn, J. A. Catal.
Today 2005, 110, 264.
21. Vinu, A.; Sawant, D. P.; Ariga, K.; Hossain, K. Z.; Halligudi, S. B.; Hartmann, M.; Nomura, M.
Chem. Mater. 2005, 17, 5339.
22. Li, Y.; Feng, Z.; Lian, Y.; Sun, K; Zhang, L.; Jia, G.; Yang, Q.; Li, C. Microporous Mesoporous
Mater. 2005, 84, 41.
23. Liu, C. Y.; Chen, C. F.; ·Leu, J. P.; Lin,· Y. C. J. Sol-Gel Sci. Techn. 2007, 43, 47.
24. Cannas, C.; Concas, G.; Gatteschi, D.; Falqui, A.; Musinu, A.; Piccaluga, G.; Sangregorio, C.; Spano, G. Phys. Chem. Chem. Phys. 2001, 3, 832.
26. Xue, J. M.; Zhou, Z.; Wang, H. J. Mater. Chem. Phys. 2002, 75, 81.
27. Tian, B.; Liu, X.; Yang, H.; Xie, S.; Yu, C.; Tu, B.; Zhao, D. Adv. Mater. 2003, 15, 1370.
28. Zhao, L.; Clapsaddle, B. J.; Satcher, J. H., Jr.; Schaefer, D. W.; Shea, K. J. Chem. Mater. 2005 17, 1358
29. Srivastava, D. N.; Perkas, N.; Gedanken, A.; Felner, I. J. Phys. Chem. B 2002, 106, 1878.
30. Jiao, F.; Harrison, A.; Jumas, J.-C.; Chadwick, A. V.; Kockelmann, W.; Bruce, P. G. J. Am. Chem.
Soc. 2006, 128, 5468
31. Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredrickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (b) Zhao, D.; Huo, Q.; Feng, J.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1998, 120, 6024.
32. Choi, D. G.; Yang, S. M. J. Colloid Interface Sci. 2003, 261, 127.
33. Li, Y.; Zhang, W. H.; Zhang, L.; Wei, Z.; Yang, Q.; Feng, Z.; Li, C. J. Phys. Chem. B 2004, 108, 9739.
34. Wu, S.; Han, Y.; Zou, Y. C.; Song, J. W.; Zhao, L.; Di, Y.; Liu, S. Z.; Xiao, F. S. Chem. Mater. 2004, 16, 486.
35. Zanella, R.; Sandoval, A.; Santiago, P.; Basiuk, V. A.; Saniger, J. M. J. Phys. Chem. B 2006, 110, 8559.
36. Sindorf, D. W.; Maciel, G. E. J. Am. Chem. Soc. 1981, 103, 4263.
37. Hoang, V. T.; Huang, Q. L.; Eic, M.; Do, T.; Kaliaguine, S. Langmuir 2005, 21, 2051. 38. Vasquez-Mansilla et al., J. Magn. Magn. Mater. 1999, 204, 29.
39. T.; Guo, L.; Tao, Y.; Hu, T. D.; Xie, Y. N.; Zhang, J. Nanostruct. Mat. 1999, 11, 487. 40. Iler, R. K. The Chemistry of Silica; Wiley-VCH: Weinheim, 1979.
41. Samoson, A.; Lippmaa, E.; Engelhardt, G.; Lohse, U.; Jerschkewitz, H.-G. Chem. Phys. Lett. 1987,
134, 589.
42. Man Kim, J.; Han, Y.-J.; Chmelka, B.F.; Stucky, G.D. Chem. Comm. 2000, 2437. 43. Sharrock, M. P. Trans. Magn. 1989, 25, 4374.
44. Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscow, L.; Pierotti, R. A.; Ronquerol, J.; Siemieniewska, Pure Appl. Chem. 1985, 57, 603.
45. Gupta, N. M.; Tripathi, A. K. Gold Bull. 2001, 34, 120.
46. Smit, G.; Zrncevic, S.; Lazar K. J. Mol. Catal. A. 2006, 252, 103.
47. Neri, G.; Bonaviat, A.; Rizzo, G.; Galvagno, S.; Donato, N.; Caputi, L. S. Sens. Actuat. B 2004,
101, 90.
Figure captions
Figure 1. (a) Low-angle XRD patterns of calcined TP-SBA-15, TP 0.15, TP0.5, and TP1.0 and (b) high-angle XRD patterns of calcined TP 0.15, TP0.5, and TP1.0.
Figure 2. N2 adsorption-desorption isotherms and pore size distributions (inset) of calcined TP-SBA-15 (a), TP0.15 (b), TP0.5 (c), and TP1.0 (d).
Figure 3. Representative scanning electron micrographs for calcined TP-SBA-15 (a), TP0.15 (b), TP0.5 (c), and TP1.0 (d).
Figure 4. Representative transmission electron micrographs of calcined TP0.15 (a), TP0.5 (b) and TP1.0 (c). EDX analysis indicates Fe2O3 contents of 12 wt% for TP0.15 (a, left), of 90 wt% for TP0.50 (b, left, dark particles in white circle) and of 13 wt% for TP0.50 (b, left, next to white circle).
Figure 5. 57Fe Mössbauer absorption spectra and fits of TP0.15 (left) and TP1.0 (right).
Figure 6. (a) Low- and (b) high-angle XRD patterns of calcined TP2-3, TP2-4, TP2-6, and TP2-7. Figure 7. N2 adsorption-desorption isotherms and pore size distributions (inset) of TP2-3 (a),
TP2-4 (b), TP2-6 (c), and TP2-7 (d).
Figure 8. High-angle XRD patterns of calcined TP4, TP8, TP16, and TP-Fe2O3.
Figure 9. N2 adsorption-desorption isotherms and pore size distribution (inset) for calcined TP4 (a), TP8 (b), TP16(c), and TP-Fe2O3 (d).
Figure 10. Transmission electron micrgraphs of calcined TP4 (a), TP8 (b), TP16 (c), and TP-Fe2O3 (d).
Figure 11. Transmission electron micrgraphs of Au/SBA-15 (a), Au/TP1-1.5 (b), Au/TP5-1.5 (c), Au/TP0.15 (d), Au/TP0.5 (e), and Au/TP-Fe2O3 (f).
18
Table 1. Physicochemical properties of the FeO
x/SBA-15 and SBA-15 materials.
Sample
Fe/Si
1pH
2Fe
(wt%)
Fe
2O
3(wt%)
d
crystal(nm)
3a
0(nm)
S
total(m
2/g)
S
micro(m
2/g)
PV
(mL/g)
d
p4(nm)
d
wall(nm)
TP-SBA-15
0
7
-
-
-
11.3
354
35
1.2
10
1.3
TP0.15
0.15
7
10
14
amorph.
13.2
470
54
1.3
9
4.2
TP0.5
0.5
7
28
40
21
12.4
460
117
0.7
7
5.4
TP1.0
1.0
7
38
54
amorph.
-
283
0
0.6
5
-
TP2-3
2.0
3
37
53
32
11.6
312
119
0.6
5; 35
6.6
TP2-4
2.0
4
41
58
58
12.1
267
93
0.5
6; 44
6.1
TP2-6
2.0
6
43
61
49
11.9
290
83
0.6
6; 30
5.9
TP2-7
2.0
7
44
63
amorph.
-
335
21
0.7
6
-
TP4
4.0
7
47
67
amorph.
-
259
0
0.6
13
-
TP8
8.0
7
57
81
30
-
100
12
0.4
21
-
TP16
16.0
7
62
89
18
-
89
16
0.4
28
-
TP-Fe
2O
3∞
7
70
100
47
-
41
0
0.1
-
-
SBA-15
0
1.5
0
0
-
12.1
840
105
1.3
7
5.1
TP1-1.5
1
1.5
0.2
0.3
amorph.
12.1
940
n.d.
0.65
6.6
5.5
TP5-1.5
5
1.5
2.8
4
amorph.
12.7
930
n.d.
0.7
6.8
5.9
1Fe/Si molar ratio in the gel.
2pH value after addition of ammonia.
3Crystal size of iron oxide phase calculated by use of the Scherrer equation from relevant XRD reflections.
419
Table 2. Fit parameters1 for the 57Fe Mössbauer spectra.
Sample IS (mm·s-1) QS (mm·s-1) Γ (mm·s-1) Hyperfine field (T) Spectral contribution (%) TP 0.15 0.35 -0.06 0.50 31.4 100 TP 1 0.35 0.37 0.94 -0.21 0.67 0.40 - 51.7 73 27
1 Experimental uncertainties: Isomer shift: IS ± 0.03 mm/s; Quadrupole splitting: QS ± 0.03 mm/s; Linewidth: Γ ± 0.05 mm/s; Hyperfine field ± 0.02 T; Spectral contribution: ± 5%.
Table 3. Gold nanoparticle size and catalytic activity of supported gold catalysts in the oxidation of CO (0.02 g catalyst; total flow 100 mL/min containing 1 vol% CO and 0.5 vol% O2 in He; reaction temperature 200 °C).
Catalyst dAu (nm) D1 (%) qTOF2 (mol CO/molAu.s) TOF3 (mol CO/molAu.s) Au/SBA-15 4.9 ± 1.3 26 0.3 1.1 Au/TP1-1.5 5.8 ± 1.9 22 0.7 3.3 Au/TP5-1.5 6.8 ± 1.9 19 1.8 9.4 Au/TP0.15 8.0 ± 1.8 16 1.6 9.8 Au/TP0.5 10.4 ± 2.2 12.5 0.6 5.2 Au/TP-Fe2O3 15 ± 5.5 8.7 0.7 8.1 1 Dispersion defined as D = 1.3/dAu 100%.
2 Quasi turnover frequency based on total gold loading. 3
20
1
2
3
4
5
6
TP-SBA-15Int
ens
ity
(
a.
u.
)
Two Theta (degree)
TP 1.0 TP 0.15 TP 0.5
a
10
20
30
40
50
60
70
80
*
*
*
*
*
*
*
*
*
*
b
Two Theta (degree)
Int
ens
ity
(
a.
u.
)
TP 1.0 TP 0.15 TP 0.5*
*: hematite Fe2O3 Figure 121 0.0 0.2 0.4 0.6 0.8 1.0 0 200 400 600 800 1 10 100 10 Pore diameter (nm) d V /d log (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po a 0.0 0.2 0.4 0.6 0.8 1.0 0 200 400 600 800 1000 1 10 100 9 Pore diameter (nm) d V /dl og (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po b 0.0 0.2 0.4 0.6 0.8 1.0 0 200 400 600 1 10 100 24 7 Pore diameter (nm) d V /d log (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po c 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 400 1 10 100 35 5 Pore diameter (nm) d V /dl og (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po
d
Figure 222
23
24 Figure 5 -12 -8 -4 0 4 8 12 3.90 3.96 4.02 Int ens ity ( 10 6 c ount s ) Doppler velocity (mm/s) -12 -8 -4 0 4 8 12 3.0 3.2 3.4 Int ens ity ( 10 5 c ount s ) Doppler velocity (mm/s)
25 1 2 3 4 5 6 pH = 7 pH = 6 pH = 4 pH = 3 Int ens ity ( a. u. ) 2 Theta (degree) 10 20 30 40 50 60 70 80 * * * * * * pH = 3 pH = 4 pH = 6 Int ens ity ( a. u. ) 2 Theta (degree) pH = 7 *: Hematite * * * * * * * * Figure 6
26 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 400 1 35 5 Pore diameter (nm) d V /dl o g (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po
a
0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 1 10 100 44 6 Pore diameter (nm) d V /d log (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Pob
0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 400 1 10 100 30 6 Pore diameter (nm) d V /dl o g (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po C 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 400 500 1 10 100 6 Pore diameter (nm) d V /dl o g (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Pod
Figure 727
10
20
30
40
50
60
70
80
H H H H H H H H H H M M M MM M MTP16
TP8
TP4
TP-Fe
2O
3Int
ens
ity
(
a.
u.
)
2 Theta (degree)
M H: hematite M: magnetite Figure 828 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 400 1 10 100 6 Pore diameter (nm) d V /d log (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po a 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 1 10 100 13 Pore diameter (nm) dV /dl og (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po b 0.0 0.2 0.4 0.6 0.8 1.0 0 100 200 300 10 100 21 Pore diameter (nm) d V /d log (D ) Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po c 0.0 0.2 0.4 0.6 0.8 1.0 0 50 100 1 10 100 3.3 43 Pore diameter (nm) dV /d log (D ) d Adsorption Desorption V ol ads or bed ( cc /g S T P ) P/Po Figure 9
29
30
31
SYNOPSIS TOC
Fe/SBA-15 was prepared by a two-step pH adjustment method. At low Fe content, iron is dispersed over the silica walls of SBA-15. At intermediate Fe content, precipitation of iron hydroxides occurs which are included in the silica mesophase. Higher iron contents lead to fast precipitation and complete deterioration of the SBA-15 structure. Gold nanoparticles show synergy with the surface grafted Fe in the oxidation of carbon monoxide.
