Sie sind auf Seite 1von 14

Applied Catalysis A: General 187 (1999) 255268

Preparation and characterization of SiC microtubes


Nicolas Keller a , Cuong Pham-Huu a , Marc J. Ledoux a, , Claude Estournes b , Gabrielle Ehret c
a

Laboratoire de Chimie des Matriaux Catalytiques, ECPM, Universit, Louis Pasteur, 25, rue Becquerel 67087 Strasbourg Cedex 2, France b Groupe des Matriaux Inorganiques, IPCMS, UMR 7504 du CNRS, 23 rue de Loess, 67037 Strasbourg Cedex, France c Groupe Surface & Interface, IPCMS, UMR 7504 du CNRS, 23 rue de Loess, 67037 Strasbourg Cedex, France Received 24 March 1999; received in revised form 2 June 1999; accepted 2 June 1999

Abstract Silicon carbide microtubes with medium surface area, 3070 m2 g1 , were successfully prepared by shape memory synthesis involving the reaction between SiO vapors and low surface area, 12 m2 g1 , carbon microbers. The gross morphology of the carbon microbers was conserved during the carburization process. After calcination at 600 C in order to eliminate the unreacted carbon, hollow SiC microtubes were obtained. The pore size distribution of the material was centered around 10 nm, allowing a high accessibility of potential reactants to an active phase during catalytic reactions. The surface was covered by an amorphous layer 3 nm thick. XPS measurements revealed that this amorphous phase was composed of a mixture of SiO2 and SiOx Cy . Soda treatment at 80 C allowed complete removal of this surface phase without any change in the material morphology. A similar result was also obtained after treatment with an aqua regia medium. 1999 Elsevier Science B.V. All rights reserved.
Keywords: Silicon carbide; Microtubes; High surface area; Catalyst support

1. Introduction Activated carbon has been widely used as a catalyst support for active phases in liquid-phase processes because of its high resistance towards aggressive environments, its possible separation from the reaction media and also because of the relatively simple recovery of the metals in the active phase at the end of the catalysts life. However, attrition problems under vigorously agitated liquid can lead to the loss of active components and to a decrease in the catalyst particle size, which renders more difcult the separation of the catalyst from the liquid phase. For these reasons,
Corresponding author. Tel.: +33-3-88-13-68-81; fax: +33-3-8813-68-80 E-mail address: ledoux@cournot.u-strasbg-fr (M.J. Ledoux)

a growing interest has appeared for the use of a new type of catalyst support based on carbon bers [1,2]. However, the high afnity of carbon for oxygen at relatively low temperatures also renders the use of carbon as a catalyst support less attractive, i.e. support loss during oxidative regeneration. It was then of interest to search for substitutes for carbon having the same positive properties but without the drawbacks. Silicon carbide (SiC) exhibits a high thermal conductivity and is chemically inert until 800 C. The high thermal conductivity avoids the formation of hot spots during regeneration and the chemical inertness allows easy recovery of the active phase by acidic or basic washing. In addition, when used as catalyst support, silicon carbide offers a unique advantage over traditional materials such as alumina or silica since it is also an excellent electrical conductor. Therefore,

0926-860X/99/$ see front matter 1999 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 6 - 8 6 0 X ( 9 9 ) 0 0 2 2 3 - 9

256

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268 Table 1 Characteristics of the carbon microbers Carbon Ash Volatile compounds Sulfur Average diameter BET surface area 9497 wt.% 0.10.3 wt.% 1.54.0 wt.% 0.30.5 wt.% 1015 m 12 m2 g1

there is the possibility that metal crystallites introduced onto silicon carbide will form an interaction with the surface, and as a consequence, modications in the morphological characteristics of the catalyst particles could give rise to unexpected activity and selectivity patterns. Similar results were reported by Kaneko [3] for carbon nanober catalyst support. However, for silicon carbide to be useful as catalyst support, it must be prepared in a medium surface area form (20100 m2 g1 ) with an appropriate pore size distribution and the inability to do this has been the limiting factor in its application as support for heterogeneous catalysis. For this reason, considerable attention has been focused on developing methods for preparation, either directly via thermal treatment of the precursors or assisted by a gasication catalyst, that will yield high surface area materials [410]. The high surface area SiC synthesized from these techniques was successfully used as catalyst support for many reactions such as the hydrodesulfurization reaction [6], CH4 reforming [9], automotive exhaust pipe reactions [10,11], isomerization of linear saturated hydrocarbons [12] and selective oxidation of hydrogen sulde for Claus tail-gas treatment [13]. More detailed studies concerning the synthesis, characterization and catalytic use of silicon carbide as catalyst support have been published by Lednor and Chorley [14], by Ledoux and Pham-Huu [15] and by Ledoux, Pham-Huu and Chianelli [16]. Finally, due to the low sinterability of silicon carbide, it must be convenient to nd a method for preparation which can directly yield the material in its nal form (grain, cylinder or honeycomb) without an additional need for shaping. The aim of the present article is to report the preparation of high surface area silicon carbide microtubes from low surface area carbon bers (12 m2 g1 ) according to the shape memory synthesis method developed by Ledoux et al. [5,6] for use as a heterogeneous catalyst support material. The silicon carbide microtubes (after synthesis and after air calcination in order to burn off the remaining carbon) were characterized by different techniques such as powder X-ray diffraction (XRD), thermal gravimetry analysis (TGA), and surface area and pore distribution measurements by nitrogen adsorption. The morphology and microstructure of the different solid materials were observed by scanning electron microscopy (SEM) and high-resolution transmission electron mi-

croscopy (HRTEM). The nature of the phases which composed the uppermost layers of the material was investigated using the XPS technique. The resistance of the sample in an aggressive environment (strongly acidic) was also investigated. The morphology of the sample after such a treatment was observed by SEM.

2. Experimental section 2.1. Raw materials The carbon microbers were supplied by Carbone Lorraine Com. (purity 97.5%) and their characteristics are summarized in Table 1. Silicon (Janssen, 99.9%) and silica (Merck, 99.5%) were used in a powder form. 2.2. SiC synthesis apparatus A sketch of the apparatus is given in Fig. 1(a). The carbon microbers and the Si and SiO2 mixture were located in two volumes separated by 3 cm in a densied alumina crucible (inner diameter 60 mm, length 100 mm), which was not active towards the reactants and the material formed (Fig. 1(b)). The SiO vapor was generated by heating the mixture of silicon and silica powder, located in the lower part of the alumina crucible. The SiO vapor was then pumped through the carbon ber bed maintained at ca. 1250 C where silicon carbide was formed according to the following reaction: SiO(gas) + 2C(solid) SiC(solid) + CO(gas) (1)

The alumina crucible was introduced into an impermeable densied silicon carbide tube (inner diameter 130 mm, length 800 mm) in which a dynamic vacuum (p = 0.05 Torr) could be maintained. The dynamic vacuum was achieved via a turbo pump, the CO formed

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

257

Fig. 1. (a) Schematic diagram of the apparatus used in the synthesis of SiC microtubes. (b) Schematic diagram of the crucible used in the synthesis of SiC microtubes.

was pumped out of the reactor and allowed the displacement of the reaction in favor of SiC formation. Before the reaction, the carbon microbers were evacuated at 1000 C (heating rate = 10 C min1 ) for 2 h in order to desorb the impurities from its surface. The temperature was increased from 1000 C to the synthesis temperature and kept at this value for 15 h. After synthesis, the sample was cooled to room temperature under vacuum and then stored in a closed vessel before characterization. It has been reported in our previous publications that, depending on the reaction conditions, a part of the unreacted carbon remained in the core of the material. In order to obtain clean pure SiC microtubes, it was necessary to burn off this unreacted carbon. The sample was separated into two batches: one was char-

acterized without any treatment and the second was calcined in air at 600 C for 2 h. 2.3. Characterization techniques Structural characterization of the samples was done by powder XRD measurements, carried out with a Siemens Diffractometer Model D-5000, using a Cu K radiation. The measurements were made with long duration scan (10 s) and a small step scan (0.02 2 ). The mean crystallite sizes were determined from the Scherrer equation with the normal assumption of spherical crystallites. The nature of the crystalline phases present in the different samples was checked using the data base of the Joint Committe on Powder Diffraction Standards (JCPDS).

258

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

TGA experiments were performed using a Balzer thermo-analyser to determine the total amount of carbon remaining in the material. The sample was placed in a platinum crucible and heated from room temperature to 1000 C (heating rate = 10 C min1 ) using a 20% (v/v) O2 /N2 mixture at a ow rate of 5 cm3 min1 . During the heating process, the weight change of the catalyst was monitored continuously. The pore size and surface area measurements were performed on a Coulter SA-3100 porosimeter using N2 as adsorbant. The sample, after treatment, was transferred to the BET cell via a glove-box under dry nitrogen. Before each measurement, the sample was evacuated at 300 C for 3 h in order to desorb the impurities adsorbed on its surface. The cell was equipped with a greaseless valve in order to avoid air exposure of the sample during the transfer to the porosimeter. SBET is the surface area of the sample calculated from the nitrogen isotherm using the BET method. SBJH is the surface area of all the pores except micropores calculated from the N2 desorption isotherm. The micropore surface area and volume were calculated using the t-plot method developed by de Boer and co-workers [17]. A more detailed study has been published by Mikhail et al. [18] concerning the correctness of the different parameters used in the method. The t-plot consists of the analysis of the vl t plot curve where vl is the volume of nitrogen adsorbed as liquid at a given pressure P/P0 by the BET surface and t is the statistical thickness obtained by dividing the volume of nitrogen adsorbed as liquid at a given pressure P/P0 by the BET surface. The combination of the t-plot and the BJH method for narrow and larger pores allows the analysis of a nearly complete pore volume and pore surface distributions in the sample studied. The morphology of the material was observed by SEM using a Jeol microscope Model JMS-840 operated with an accelerating voltage of 20 kV. The samples were pre-covered by a thin layer of gold in order to avoid the charging effect problem during analysis. Surface characterization was carried out by XPS. XPS spectra were recorded on a Cameca Nanoscan 50 using the Al K source at 1486.6 eV. The energy scale of the instrument was calibrated using Ag 3d5/2 = 368.0 eV and Au 4f7/2 = 84.0 eV. The C 1s at 283.3 eV and the Si 2p at 103.0 eV, corresponding, respectively, to the carbon engaged in the carbidic phase and the silicon in the silica phase, were both used as

reference, by comparison with the published literature. The resolution in energy was 0.7 eV and the vacuum was maintained at below 5 109 atm. The sample was pressed on an indium foil and then transferred under air into the XPS preparation chamber. Only C 1s and Si 2p were recorded, and for a given emission, the number of accumulations was selected in order to achieve a good signal-to-noise ratio. The samples were analyzed without any pretreatment, i.e. ion etching, in order to avoid any surface modication such as selective reduction or composition modication. Indeed, it has been reported by several authors in the literature that Ar+ ion bombardment induced preferential elemental sputtering thus leading to an inaccurate analysis. TEM and EDS were carried out in a Topcon Model EM200B operating at 200 kV with a point to point resolution of 0.17 nm. To prevent artifacts due to contamination, no solvents were used at any stage and samples were prepared by grinding the catalysts between glass plates and bringing the powder into contact with a holey carbon-coated copper grid. Great care was taken during the TEM experiments in order to avoid heating effects from the incident beam. EDS analysis was carried out with a 5 nm and 100 s life time electron beam. The detector was equipped with beryllium windows, and thus, it allowed the detection of light elements such as carbon or oxygen.

3. Results and discussion 3.1. Synthesis of cubic -SiC microtubes Typically, 4.0 0.1 g of carbon microbers was used for each experiment. The reaction temperature was set between 1200 and 1300 C and the duration xed at 15 h. The reaction conditions and the corresponding C SiC conversion, given as examples, are summarized in Table 2. The total amount of carbon was never transformed into SiC even with a long duration of reaction, meaning that SiC conversion could be inuenced by modifying the reaction conditions. This was explained by several factors: (i) as a function of the transformation, the SiC layers formed on the carbon surface acted as a barrier for the diffusion of SiO and CO vapors, consequently decreasing the rate

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

259

Table 2 Inuence of the synthesis parameters on the C SiC conversion and the surface area of the SiC microtubes synthesized by the gassolid reaction between SiO vapor and carbon bers Temperature, Si/SiO2 molar ratio, (Si + SiO2 )/C mass ratio 1300 C/1/12 1300 C/1/8 1200 C/1/8
a

Conversion (wt.%)a 85 60 18

SBET after synthesis (m2 g1 /sample) (m2 g1 /SiC) 18 (21) 18 (30) 25 (138)

SBET after calcination (m2 g1 ) 30 45 66

The conversion was calculated based on the starting weight of the carbon bers.

of carburization; and (ii) the carbon bers used in this preparation have a very low surface area (few square meters which justied the corrected value of the surface area in Table 2) compared to the high surface area activated charcoal used in other works [6]. Such an ordered and non-porous structure could mitigate SiC nucleation and also limit the vapor diffusion in the material. Under similar reaction conditions, the total conversion C SiC was lower when starting from the low surface area carbon bers than when starting from high surface area activated charcoal [19]. Similar results have already been reported by Moene et al. [9] during his investigation on SiC synthesis. He has observed that the number of nucleation sites on activated charcoal was higher compared to that observed on the ordered graphite material, resulting in a more grained structure of the nal SiC. In addition, it has been reported by Benaissa et al. [20] that SiC formation was initiated in several domains throughout the activated charcoal matrix and especially on the defect points and micropores of the activated carbon surface which had a high adsorption energy. The higher rate of SiC formation observed on the activated charcoal was also attributed to the high porosity which could favor SiO vapor diffusion. 3.2. Crystalline structure and oxidation behavior The XRD patterns of the SiC microtubes prepared at 1300 C for 15 h are presented in Fig. 2. Whatever the reaction temperature and duration, XRD showed the presence of only silicon carbide as the sole crystalline reaction product. The synthesized SiC exhibited diffraction lines corresponding to -SiC crystallized in a face centered cubic structure. In addition to the -SiC diffraction lines, a broad peak located at a low two-theta angle was observed for the sample synFig. 2. XRD patterns of the SiC microtubes after synthesis (a) and after calcination in air at 600 C for 2 h (b). The diagrams were recorded with a step of 0.02 (2 ) and a scan time of 10 sec per step.

thesized at 1300 C, which could be attributed to the unreacted remaining carbon. The reaction temperature used (12001400 C) allowed the formation of small amounts of stable -SiC along with -SiC, evidenced by a peak located at around diffraction angles of 33.7 corresponding to -SiC in the 4H and 15R structures.

260

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

This phenomenon was attributed to the formation of stacking faults along the [111] direction in the SiC particles, as will be shown by TEM below, and according to the SiC microstructure published previously using HRTEM [20]. No traces of SiO2 or silicon were detected along with the carbide XRD pattern, which means that such species, if present, were only in an amorphous form or in very small amounts. The XRD patterns of silicon carbide synthesized at 1300 C for different durations (not reported) showed the increase in the silicon carbide diffraction lines at the expense of those corresponding to the starting carbon, meaning that a relatively long time was needed to transform more of carbon microbers into SiC microtubes. It was signicant to note that carburization was realized by direct exchange between carbon and oxygen atoms and no traces of silicon were detected on the intermediate samples. The relatively long time needed in order to obtain a high C SiC conversion was attributed to the nature of the transformation, i.e. a Shrinking Core Model [21,22], where SiC formation depended on the diffusion of the SiO vapor through the SiC mantle and also to the relatively low reaction temperature used in this study. The Shrinking Core Model states that the reactional interface is maximum at the beginning of the transformation and then decreases regularly as the reaction takes place. At the beginning of carburization, all of the carbon surface was accessible and available to the SiO vapor; as soon as SiC formation started, the transformation rate decreased due to the diffusion limitation of the SiO and CO vapors through the rst SiC layers. The nal material structure could be described as silicon carbide microtubes containing a carbon core. The behavior of a support under an oxidizing atmosphere is important in order to check its resistance during oxidative regeneration. The TGA diagrams of the carbon bers and of the as-synthesized material under a 20% (v/v) O2 /N2 ow are shown in Fig. 3. The carbon microbers remained stable up to around 500 C and then rapidly burned completely (Fig. 3(a)). The SiC microtubes (containing carbon) were stable up to 600 C (Fig. 3(b)). At around 600 C, a weight loss was observed due to complete combustion of the remaining carbon located in the core of the tubes. For higher temperatures, the SiC microtubes remained stable and no weight loss was observed up to 800 C where the oxidation of SiC by air could begin to result in a weight

Fig. 3. Thermogravimetry analysis of the carbon microbers (a) and SiC microtubes after synthesis (b).

gain. Similar observations have already been reported by Stegenga et al. [23] during their investigation of the thermal resistance of a support for automotive exhaust pipe reactions. In conclusion, it seemed that carbon covered by a layer of SiC (as shown by the SEM observation below) hindered oxygen penetration, meaning that the SiC coating acted as a diffusion barrier for carbon oxidation. The small increase in weight observed at high temperatures was attributed to the oxidation of the SiC surface according to Eq. (2) and as per the results reported by Moene [41].
3 SiC(s) + 2 O2(g) SiO2(s) + CO(g)

(2)

The SiC after calcination in air at 600 C for 2 h exhibited an XRD pattern (Fig. 2(b)) similar to that observed before calcination. The broad peak located at the low two-theta angle disappeared, meaning that all of the remaining low surface area carbon was eliminated by the oxidative treatment. In addition, no trace

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

261

Fig. 4. Nitrogen adsorptiondesorption isotherms of carbon microbers (1), SiC microtubes after synthesis (2) and SiC microtubes after calcination in air at 600 C for 2 h (3). Filled points correspond to adsorption curves and empty points to desorption curves.

Fig. 5. Pore size distribution of SiC microtubes after synthesis and after calcination in air at 600 C for 2 h.

of silica or other oxides were observed. Such a phenomenon could be explained by the fact that silicon dioxide, if formed, was located mainly on the material surface (see XPS results below) or in an amorphous form, neither of which is clearly shown by the XRD technique. The relatively low oxidation temperature used in the present work also explained the lack of bulk oxidation of the sample. 3.3. Surface area and porous measurements To be used as catalyst supports, materials must be prepared with a relatively high surface area, i.e. 10 m2 g1 , which allows the exposure of the active phase to the reactant mixture. The material must also contain as few micropores as possible, which could hinder the access to the reactants. In this section, investigation of the evolution of the surface area of the SiC microtubes as a function of the calcination treatment along with the pore size distribution modication is described. The nitrogen isotherm adsorptiondesorption curves measured for the initial carbon bers, the as-synthesized SiC microtubes and the material after calcination in air are shown in Fig. 4. The adsorptiondesorption isotherms of the starting carbon bers were found to be of Type III according to the well-accepted isotherm classication [24,25],

and this shape indicated the existence of macropores. The initial specic surface area of the carbon bers, calculated using the standard BET method, was very low (12 m2 g1 ). The nitrogen isotherm of the SiC tubes was drastically modied compared to that obtained on the starting carbon ber, with an increase in the porous volume from 0.010 to 0.058 cm3 g1 . The Type IV isotherms observed were typical of a solid containing mesopores ranging from 2.5 to 50 nm. The t-plot method performed on the material showed a straight line crossing the origin, meaning that almost no micropores (2.5 nm) were present in the material. The pore size distributions of the SiC tubes before and after calcination in air are presented in Fig. 5. After synthesis, an important fraction of the macropores of the carbon bers precursor had disappeared, leading to the formation of a large number of mesopores located around 1011 nm and to an increase in the surface area from 12 m2 g1 to around 12 m2 g1 . The increase in the material surface area when going from C to SiC was probably due to the formation of several nuclei at the beginning of the synthesis, followed by the formation of small domains of SiC, and thus, creation of pores which increased the surface area of the material. Carbon consumption during the course of the transformation also created voids inside the material and contributed to the increase in the sample surface area. Such a phenomenon has already been reported by Benaissa et al. [20] during their investigation on the mechanism of SiC growth by HRTEM. This

262

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

explanation will be conrmed by the SEM micrographs reported below. However, it is signicant to note that the value of the surface area given here was underestimated due to the presence of unreacted carbon with low surface area inside the material. The remaining carbon was eliminated from the SiC material by calcination of the sample in air at 600 C (according to the results obtained by the TGA technique) for 2 h. The surface area of the sample after calcination was signicantly increased from 12 to 45 m2 g1 , and thus, it highlighted the advantage of the gassolid synthesis method employed in this work to produce high surface area SiC tubes from the starting low surface area carbon bers. The average distribution of the pores was left unchanged and centered around 1011 nm which could allow a good dispersion of an active phase and its accessibility without microdiffusion phenomena. The absence of modication of the pore size distribution after carbon removal strongly supported the fact that such a treatment only removed carbon inside the ber, without sintering or additional formation of other compounds with a different pore size distribution. The higher nal surface area of the calcined sample was also due to the disappearance of carbon, leading to the formation of empty bers or tubes, as will be observed by SEM later. Such a transformation is sketched in Fig. 6. However, one could also suggest that the increase in surface area after air treatment could also be attributed to the formation, on the SiC surface, of a layer of amorphous silica with high surface area. This hypothesis was rejected as the calcined material surface area, 25 m2 g1 in one sample, remained unchanged after soda treatment at 80 C which removed the silica or oxycarbide layers present on the surface, 24 m2 g1 (see XPS results below). These results were in line with those reported by Moene [9] who observed that almost no SiO2 appeared on the SiC surface at a temperature around 800 C even after more than 2 h of oxidation. It is signicant to note that the material surface area was strongly dependent on the reaction temperature, 66 m2 g1 at 1200 C compared to 45 m2 g1 at 1300 C. Such a phenomenon was attributed to the surface diffusion process, as already reported by different authors in the literature [26,27]. Elder and Kristic [26] and Hase et al. [27] have shown that SiC started to sinter by surface diffusion at a temperature lower than the theoretical sintering temperature. The corrected

values of the BET surface in Table 2 clearly illustrate the sintering phenomenon. At low conversion (short reaction time), SiC was not sintered and the surface area was as high as 139 m2 g1 , while for a longer reaction time, this surface fell to 30 m2 g1 . The rate of the loss of surface area by a surface diffusion process was proportional to the initial surface area of the treated sample. This explained the relatively small decrease in the SiC surface area with increasing reaction temperature. This behavior was completely different from that observed when the starting carbon material was a high surface area activated charcoal, for which one could notice an important drop in the SiC surface area with increasing reaction temperature [19].

3.4. Morphology and microstructure Figs. 79 show the SEM micrographs of the carbon precursor and the resulting SiC (after synthesis and after air calcination at 600 C). The SEM micrographs of the carbon bers showed an average diameter of 1015 m with a smooth surface (Fig. 7(a) and (b)) in agreement with the low surface area obtained by the BET technique. Low magnication SEM micrograph of the corresponding SiC microtubes synthesized at 1300 C is presented in Fig. 8(a) and it shows that the gross morphology of the carbon microbers was retained. Similar results have already been reported in a previous article [6] during the silicon carbide synthesis from other carbon shapes. Such a result was explained by the fact that the transformation reaction between the SiO vapors and carbon proceeded via the step-by-step replacement of C atoms by Si atoms, with the release of CO, leading to the formation of SiC. Such a transformation mechanism provides a nal material with the same gross morphology as that of the precursor. High magnication SEM images of the resulting SiC material are presented in Fig. 8(b) and (c). The microstructure of the SiC microtubes was strongly modied when compared to that observed on the starting carbon microbers: the smooth surface carbon bers were transformed into an homogeneous porous SiC, with an average network structure of around 50 nm. The morphology, in different magnications, of the SiC material after calcination in air at 600 C is presented in Fig. 9 and it shows that the gross morphology of the sample was retained. Cut calcined SiC

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

263

Fig. 6. Formation of SiC microtubes during calcination in air, starting from the SiCcarbon composite.

material revealed the presence of empty SiC microtubes, formed by combustion of the unreacted carbon inside the core of the material (Fig. 9(c)). Similar results have already been observed during the synthesis of vanadium carbide, where the carbide bers obtained were empty [28]. The thickness of the SiC wall and the tube internal diameter could be modied by acting on the C SiC conversion, i.e. synthesis parameters (Fig. 10). The TEM image of the SiC material is presented in Fig. 11. The lattice image of the as-synthesized SiC tube is reported in Fig. 11(a), showing a high density of stacking faults along the 111 direction. The stacking faults can be viewed in the TEM micrograph as black stripes in the grains. The high resolution TEM image showed that the interplanar distance along this 111 direction was constant and equal to 0.25 nm. No long range periodicity was observed, which led to the conclusion that the SiC formed was a one-dimensionally disordered polytype, which grew preferentially along the 111 direction, according to the literature and previous work [20,29]. Stacking faults are often classied as growth faults during the layer-by-layer growth of a crystal with close-packed structure when a new layer is stacked incorrectly. This conrmed the step-by-step mechanism of synthesis highlighted above. Similar results have already been reported by Wang et al. [30] and by Koumoto et al. [31] during their investigations of SiC formation and structure. The high magnication TEM image also showed that the SiC surface was covered by an amorphous phase, 3 nm thick (Fig. 11(b)), which contained Si,

Fig. 7. SEM micrographs of the carbon microbers.

O and C, as identied by EDS. The presence of an amorphous phase, attributed to a silicon oxidic phase on the SiC surface, has already been reported by many authors in the literature [3236]. Zhu et al. [37] have shown, by HRTEM, the presence of an amorphous layer which seems to be an inherent part of the SiC whisker. The nature of the amorphous phase observed on the SiC microtubes will be inves-

264

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

GaussianLorentzian spectra, each component being representative of a chemical entity. The XPS spectra (Figs. 12 and 13) showed that the SiC surface was covered by an amorphous phase composed of at least two silicon based components and not simply by an SiO2 layer. The binding energy of these different phases which composed the amorphous layer are reported in Table 3, in close agreement with those reported in the literature. The existence of silicon suboxide along with SiO2 was highly improbable since, according to thermodynamic considerations, silicon suboxide (Si<4+ ) should be unstable in the solid state

Fig. 8. SEM micrographs of the SiC microtubes after synthesis.

tigated and discussed in the next section using XPS analysis. 3.5. Surface investigation In order to obtain further insight about the nature of the different components present on the SiC microtubes, surface XPS was performed. Dissymmetries and line widths of the main C 1s and Si 2p peaks showed that these signals do not correspond to a unique chemical form. In order to resolve the different chemical components, theoretical C 1s and Si 2p spectra were constructed by addition of the

Fig. 9. SEM micrographs of the SiC microtubes after calcination in air at 600 C for 2 h.

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

265

Table 3 XPS of the C 1s and Si 2p binding energy and attribution of the C 1s and Si 2p components recorded from SiC microtubes (comparison with the published literature) C 1s (eV) I SiC SiC after synthesis SiC after synthesis followed by soda treatment Literature Portre et al. [33] Ramahan et al. [32] Bouillon et al. [34,35] 283.3 283.3 283.3 283.3 II SiOx Cy 284.7 284.2 III CO 287.0 Si 2p (eV) I SiC 100.5 100.5 100.5 100.2 101.1 II SiOx Cy 101.6 101.5 102.1 III SiO2 103.0 103.2 102.8 103.2

Fig. 10. SEM micrograph of the SiC microtubes synthesized at 1200 C for 15 h, and calcined in air at 600 C for 2 h, corresponding to a C SiC conversion of 18%.

[33]. The second peak could be assigned to a silicon based compound containing both carbon and oxygen, i.e. a silicon oxycarbide phase, SiOx Cy : the presence of oxygen provides an electron charge transfer from silicon to oxygen, which increases the Si 2p binding energy (Fig. 12(a)). In Fig. 13(a), the XPS spectrum of C 1s is reported. Along with the carbon engaged in the carbidic phase (BE = 283.3 eV), two other C 1s peaks can be resolved unambiguously and the second one attributed to the oxycarbide species. Similar results have been reported by several authors working on silicon based materials such as SiC and Si3 N4 . Benaissa et al. [36], using a combination of HRTEM and EDS analysis, have shown the presence of a silicon containing amorphous layer with a thickness of 13 nm. Mozdrierz et al. [38] have reported the stoichiometry of SiO1.52 C0.61.05 for the amorphous

phase present on the SiC surface using a combination of several techniques, i.e. EELS, HRTEM and EDS. This result has also been conrmed by Moene [42] during his investigations on Mo and Ni supported on SiC, using the temperature-programmed technique. He observed that oxides supported on SiC display a lower reduction temperature compared to those obtained with the same catalysts supported on bulk SiO2 prepared and treated under similar conditions. Rahaman et al. [32], using XPS analyses of the amorphous layer covering a silicon nitride or silicon carbide surface, have suggested that such a layer was probably made up of a silicon oxynitride or oxycarbide rather than pure SiO2 . XPS measurements have allowed Porte and Sartre [33] to state the existence of a silicon oxycarbide phase in silicon carbide bers. Similar results were also reported by Bouillon et al. [34] using the same technique, concerning silicon carbide bers prepared by pyrolysis of a polycarbosilane precursor, or by Pampuch [39] during the oxidation of SiC. The amorphous mixture of SiO2 and SiOx Cy present on the SiC surface could be removed by soda (20% in an aqueous solution) treatment at 80 C for 1 h. The material, after treatment and water washing, was examined by XPS (Fig. 12(b) and Fig. 13(b)). After soda treatment, the Si 2p peaks in SiO2 and SiOx Cy disappeared totally and Si 2p engaged in SiC was the only form observed (Fig. 12(b)). It must be pointed out that the material after soda treatment at 80 C was exposed to air for a relatively long period, and that XPS analysis revealed almost no additional supercial oxidation. Such a result could be explained by the fact that a large part of the oxidic amorphous phases present on the SiC surface was formed during the course of

266

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

Fig. 12. XPS spectra of Si 2p of SiC microtubes after synthesis (a) and followed by soda treatment (b).

Fig. 11. TEM micrographs of the SiC microtubes after synthesis, showing the presence of a high density of stacking faults along the [111] axis and an amorphous layer on the surface.

material synthesis, probably by the reaction between C, SiO and oxygen in the reaction zone. The C 1s XPS spectrum of the soda treated sample presented in Fig. 13(b) only showed the carbon engaged in the SiC form. The vanishing of the third peak observed in the C 1s spectrum could be explained by the fact that this peak was due to carbonate species dissolved during the soda treatment, as already reported in the literature [33]. In summary, the SiC surface synthesized by the gassolid reaction was not pure but covered by an

amorphous layer with a thickness around 3 nm. Such a layer was not only composed of SiO2 but probably also a mixture of SiO2 and SiOx Cy phases which strongly affected the surface properties of the material, as compared to those observed on a bulk silica material resulting in a higher ability to disperse metal species deposited on it. 3.6. Stability of the SiC bers in an aggressive environment As already pointed out in Section 1, the best advantage offered by the usage of SiC as a catalyst support is its total chemical inertness which allows its use for operations in aggressive media. The stability of SiC in such an aggressive environment was investigated by

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268

267

por at 12001300 C under dynamic pumping according to the shape memory synthesis. The results observed underline the high potential of such a method for preparing high surface area support material from low surface area precursors. The material displayed a high surface area without the presence of microposity, a good resistance towards attrition and a suitable pore distribution (10 nm) for catalytic applications, especially in liquid-phase reactions. Furthermore, the sample exhibited a high resistance towards oxidation compared to the corresponding carbon bers and an excellent stability in an aggressive environment. The catalytic applications of materials based on SiC microtubes, especially for the liquid slurry dewaxing process and selective oxidation, will be presented elsewhere.

Acknowledgements The TGA and SEM experiments were performed at the Groupe des Matriaux Inorganiques of the IPCMS (UMR 7504 of the CNRS). Helen Lamprell is gratefully acknowledged for having performed part of the experiments. J. Ohmet (Groupe Surfaces et Interfaces) and Prof. J. Guille (Groupe des Matriaux Inorganiques) of the IPCMS-UMR 7504 of the CNRS are gratefully acknowledged for XPS experiments and helpful discussions.

Fig. 13. XPS spectra of C 1s of SiC microtubes after synthesis (a) and followed by soda treatment (b).

References
[1] M.S. Hoogenraad, R.A.G.M.M. Van Leeuwarden, G.J.B. Van Breda Vriesman, A. Broersma, A.J. Van Dillen, J.W. Geus, Preparation of catalysts VI, in: G. Poncelet, J. Martens, B. Delmon, P.A. Jacobs, P. Grange (Eds.), Studies in Surface Science and Catalysis, vol. 91, Elsevier, Amsterdam, 1995, p. 263. [2] V.B. Fenelonov, L.B. Avdeeva, O.V. Goncharova, L.G. Okkel, P.A. Simonov, A.Y. Derevyankin, V.A. Likholobov, Preparation of catalysts VI, in: G. Poncelet, J. Martens, B. Delmon, P.A. Jacobs, P. Grange (Eds.), Studies in Surface Science and Catalysis, vol. 91, Elsevier, Amsterdam, 1995, p. 825. [3] K. Kaneko, Langmuir 3 (1987) 357. [4] M.A. Vannice, Y.L. Chao, R.M. Friedman, Appl. Catal. 20 (1986) 91. [5] M.J. Ledoux, S. Hantzer, J. Guille, D. Dubots, US Patent No. 4 914 070. [6] M.J. Ledoux, S. Hantzer, C. Pham-Huu, J. Guille, M.P. Desaneaux, J. Catal. 114 (1988) 176.

treating SiC in a boiling aqua regia solution for 2 h. The sample was then washed with demineralized water until the pH equaled 7, ltered and dried in an oven at 100 C for 2 h. SEM observations of the morphology of the sample, before and after treatment, show that this morphology was kept unchanged. Similar results have been reported by Rubio et al. [40] and by Moene [41] with boiling HNO3 solution.

4. Conclusion Low surface area carbon bers (12 m2 g1 ) were readily converted to high specic surface area SiC microtubes (3070 m2 g1 ) after reaction with SiO va-

268

N. Keller et al. / Applied Catalysis A: General 187 (1999) 255268 [25] S.J. Gregg, K.S.W. Sing, Adsorption, Surface Area and Porosity, Academic Press, 1982. [26] P. Elder, V.D. Kristic, J. Mater. Sci. Lett. 8 (1989) 941. [27] T. Hase, B.W. Lin, T. Iseki, H. Suzuki, J. Mater. Sci. Lett. 5 (1986) 69. [28] F. Meunier, P. Delporte, B. Heinrich, C. Bouchy, C. Crouzet, C. Pham-Huu, P. Panissod, J.J. Lerou, P.L. Mills, M.J. Ledoux, J. Catal. 169 (1997) 33. [29] S. Shinozaki, K.R. Kinsman, Acta Metallurgica 26 (1978) 769. [30] L. Wang, H. Vada, J. Mater. Res. 7 (1992) 148. [31] K. Koumoto, S. Takeda, C.H. Pai, T. Sato, H. Yanagida, J. Am. Ceram. Soc. 72(10) (1989) 1985. [32] M.N. Rahaman, Y. Boiteux, L.C. de Jonghe, Am. Ceram. Soc. Bull. 65 (1986) 1171. [33] L. Porte, A. Sartre, J. Mater. Sci. 24 (1989) 271. [34] E. Bouillon, E.D. Mocaer, J.F. Villeneuve, R. Pailler, R. Naslain, M. Monthioux, A. Oberlin, C. Guimon, G. Pster, J. Mater. Sci. 26 (1991) 1517. [35] E. Bouillon, F. Langlais, R. Pailler, R. Naslain, F. Cruege, J.C. Sarthou, A. Delpuech, C. Laffon, P. Lagarde, M. Monthioux, A. Oberlin, J. Mater. Sci. 26 (1991) 1333. [36] M. Benaissa, C. Pham-Huu, J. Werckmann, C. Crouzet, M.J. Ledoux, Catal. Today 23 (1995) 283. [37] J. Zhu, X.G. Ning, H.G. Xu, K.Y. Hu, W. Cao, H.Q. Ye, J. Mater. Sci. 26 (1991) 3202. [38] N. Mozdrierz, M. Backhaus-Ricoult, D. Imhoff, in: B. Jouffrey, C. Colliex (Eds.), Proc. 13th Int. Congr. on Electron Microscopy, 1722 July, Paris, Les Editions de Physiques B, vol. 2A, 1993, p. 251. [39] R. Pampuch, W. Ptak, S. Jonas, J. Stoch, Mater. Sci. Monographs 6 (1980) 435. [40] J. Rubio, F. Rubio, J.L. Oteo, J. Mater. Sci. 26 (1991) 2841. [41] R. Moene, M. Makkee, J.A. Moulijn, Appl. Catal. A: Gen. 167 (1998) 321. [42] R. Moene, E.P.A.M. Tijser, M. Makkee, J.A. Moujlin, Appl. Catal. A: Gen., 184 (1999) 127.

[7] P.W. Lednor, Catal. Today 15 (1992) 241. [8] R. Moene, L.F. Kramer, J. Schooman, M. Makkee, J.A. Moulijn, Preparation of catalysts VI, in: G. Poncelet et al. (Eds.), Studies in Surface Science and Catalysis, vol. 91, 1995, p. 371. [9] R. Moene, H.T. Boon, J. Schoonman, M. Makkee, J.A. Moulijn, Carbon 34 (1996) 567. [10] M. Boutonet-Kizling, P. Stenius, S. Andersson, A. Frestad, Appl. Catal. B: Environ. 1 (1992) 149. [11] C. Pham-Huu, S. Marin, M.J. Ledoux, M. Weibel, G. Ehret, M. Benaissa, E. Peschiera, J. Guille, Appl. Catal. B: Environ. 4 (1994) 45. [12] C. Pham-Huu, P. Del Gallo, E. Peschiera, M.J. Ledoux, Appl. Catal. A: Gen. 132 (1995) 77. [13] N. Keller, C. Pham-Huu, C. Crouzet, M.J. Ledoux, S. Savin-Poncet, J.-B. Nougayrde, J. Bousquet, Catal. Today, in press. [14] R.W. Chorley, P.W. Lednor, Adv. Mater. 3 (1991) 474. [15] M.J. Ledoux, C. Pham-Huu, Catal. Today 15 (1992) 263. [16] M.J. Ledoux, C. Pham-Huu, R.R. Chianelli, Curr. Opinion Solid State, Mater. Sci. 1 (1996) 96. [17] J.H. De Boer, in: Everett D.H. (Ed.), The Structure and Properties of Porous Materials, 1958. [18] R.S.H. Mikhail, S. Brunauer, E.E. Bodor, J. Coll. Int. Sci. 26 (1968) 45. [19] N. Keller, C. Pham-Huu, S. Roy, M.J. Ledoux, C. Estourns, J. Guille, J. Mater. Sci., in press. [20] M. Benaissa, J. Werckmann, J.L. Hutchison, E. Peschiera, J. Guille, M.J. Ledoux, J. Crys. Growth 131 (1993) 5. [21] N.W. Hurst, S.J. Gentry, A. Jones, B.D. McNicol, Catal. Rev.-Sci. Eng. 24 (1982) 233. [22] J.L. Falconer, K.A. Schwartz, Catal. Rev.-Sci. Eng. 25 (1983) 141. [23] S. Stegenga, M. van Waveren, F. Kapteijn, J.A. Moulijn, Carbon 30 (1992) 577. [24] S. Brunauer, P.H. Emmet, E. Teller, J. Am. Chem. Soc. 60 (1938) 309.

Das könnte Ihnen auch gefallen