5012
J. Phys. Chem. C 2010, 114, 5012–5019
Hydrothermal Stability of Mesostructured Cellular Silica Foams Qiang Li, Zhangxiong Wu, Dan Feng, Bo Tu,* and Dongyuan Zhao* Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and InnoVatiVe Materials, Laboratory of AdVanced Materials, Fudan UniVersity, Shanghai 200433, People’s Republic of China ReceiVed: October 21, 2009; ReVised Manuscript ReceiVed: January 27, 2010
The hydrothermal stability of mesostructured cellular silica foams (MCFs) was studied in detail for the first time, using a variety of techniques including transmission electron microscopy, nitrogen sorption, smallangle X-ray scattering, 29Si solid-state nuclear magnetic resonance, and Fourier transform infrared spectroscopy. It was found that the high aging temperature, greater microporosity, and high calcination temperature contribute to the stability of MCFs in high-temperature steam. The frameworks of MCFs calcined at 550 °C are stable in 100% steam at 600 °C for 12 h, but cannot withstand more critical conditions of 800 °C steam and collapse completely. By elevating the calcination temperature of MCFs to 900 °C, the polymerization degree of the silica frameworks is further enhanced, and the obtained MCF materials exhibit high hydrothermal stability under steam at 800 °C for 12 h. The results indicate that increasing the calcination temperature is an effective method to improve the hydrothermal stability of MCFs. It is concluded that 3-D disordered MCFs show structural variations during the high-temperature steam treatments different from those of 2-D ordered hexagonal SBA-15 materials. The pore size, window size, and wall thickness were unaltered for the steam-treated MCFs, while the pore size decreased and the pore wall thickness became thicker for SBA-15. 1. Introduction The discovery of M41S family mesoporous silicates has triggered enormous research work to develop novel catalysts and supports.1–4 The ordered mesoporous materials are expected to have great advantages in catalysis involving large molecules, such as heavy oil cracking, compared with microporous zeolites, due to their high specific surface areas and large uniform pore sizes. However, mesoporous silicates exhibit poor hydrothermal stability even in boiling water that hinders their industrial applications under extreme conditions. For instance, the collapse of the MCM-41 mesostructure has limited its application in catalytic reactions involving aqueous solution or boiling water.5,6 Nevertheless, the emergence of mesoporous silica SBA-157 with hexagonal pore symmetry, a relatively large and tunable pore size, and a robust framework inspired more studies for applications in molecular catalysis than MCM-41 has.8–10 This could be mainly attributed to its higher hydrothermal stability in boiling water and water vapor.11,12 We have found that the mesostructural regularity of SBA-15 after being heated at 600-800 °C in 100% steam (water vapor) can be wellpreserved.12 Compared to MCM-41 and SBA-15, mesostructured cellular silica foams (MCFs) prepared with the microemulsion templating process by using 1,3,5-trimethylbenzene (TMB) as the organic swelling agent have well-defined ultralarge mesopores (24-42 nm), a narrow pore size distribution, and threedimensional (3-D) pore structures.13,14 Furthermore, MCFs with uniform spherical pores interconnected by windows of around 9-22 nm have high BET surface areas (up to 1000 m2/g).13 Their 3-D mesopores are substantially larger than those of the ordered counterparts SBA-15 and MCM-41. MCFs can provide more favorable conditions for mass diffusion, which is advanta* To whom correspondence should be addressed. Phone: 86-21-51630205. Fax: 86-21-5163-0307. E-mail:
[email protected] (D.Z.); botu@ fudan.edu.cn (B.T.).
geous in the catalytic process, compared to SBA-15. The 1-D channel is usually plugged up by the industrial molding process applied to catalysts, which is unfavorable for mass diffusion. 3-D interconnected pore channels can overcome this shortcoming. On the basis of these considerations, MCFs with 3-D interconnected pore channels may lead to a kind of promising catalyst support, which can open new possibilities for processing catalytic reactions of heavy crude oil as well as heavy residues with a large molecular volume. In recent years, much research has been reported on the development of MCF-type catalysts.15–17 Liu et al.15 have shown that MCF-supported vanadium oxide catalysts exhibit much higher propane conversion and propylene productivity than conventional SBA-15- and MCM-41-supported catalysts in the oxidative dehydrogenation of propane, demonstrating that, besides the active redox sites, internal molecular transport within 3-D mesopores also plays an important role. In practical applications, many industrial processes are involved with steam, such as steam reforming,18 transfer energy,19 and so on. Most important of all, 100% steam at 600-800 °C is usually used to activate and regenerate the catalysts, especially in the fluidized catalytic cracking (FCC) process.20 As a result, the hydrothermal stability of mesoporous silicates as catalysts or catalyst supports is very important under steam. Up to now, much research has been focused on improving the hydrothermal stability of mesoporous silica materials.5,6,11,12,21–26 It has been shown that many factors, such as the salt effect, wall thickness, “zeolite-like” connectivity, and disordered mesostructure, can influence the hydrothermal stability of mesoporous silicate materials. Although many papers have reported the hydrothermal stability of ordered mesoporous silica, these reports are principally limited to the evaluation of the stability of mesoporous materials in boiling water, water vapor saturated nitrogen, or 20% water steam.5,6,11,23 However, the hydrothermal stability of mesoporous silicates in 100% steam at 600-800 °C has rarely been reported.12,25 On and Kaliaguine23
10.1021/jp9100784 2010 American Chemical Society Published on Web 03/01/2010
Hydrothermal Stability of Cellular Silica Foams coated the inner mesopore surface with zeolites to improve the hydrothermal stability of MCFs in 20% steam. To the best of our knowledge, the hydrothermal stability of MCFs, particularly exposed to 100% steam at 600-800 °C, has not been reported until now. In the present work, we systematically investigated the effects of structural parameters on the hydrothermal stability of MCFs treated in 100% steam at 600-800 °C, which is similar to industrial conditions. The results show that the mesostructure of MCFs calcined at 550 °C is stable after exposure to 100% steam at 600 °C for more than 12 h, but collapses gradually at 700 °C steaming. By increasing the calcination temperature of MCFs, the framework becomes more and more “robust” and the mesostructure can withstand the 100% steam at 800 °C for more than 12 h. Furthermore, we found different structural variations between MCFs and SBA-15 during the high-temperature steam treatments. The pore size, window size, and wall thickness were unaltered for the steam-treated MCFs, while the pore size decreased and the pore wall thickness became thicker for SBA15. 2. Experimental Section 2.1. Chemicals. Triblock copolymer poly(ethylene oxide)b-poly(propylene oxide)-b-poly(ethylene oxide), Pluronic P123 (MW ) 5800), was purchased from Aldrich Chemical Inc. Tetraethyl orthosilicate (TEOS), ammonium fluoride, TMB, and potassium chloride were purchased from Shanghai Chemical Co. All chemicals were used as received without any further purification. Millipore water was used in all experiments. 2.2. Synthesis. To investigate the influence of the main structural parameters on the hydrothermal stability of MCFs, six samples were prepared under different conditions, designated as MCF-1, MCF-2, MCF-3, MCF-4, MCF-5, and MCF-6. MCF-1, MCF-2, and MCF-3 were synthesized by the following procedure: A 4.0 g portion of Pluronic P123 was dissolved in 150 mL of 1.6 M HCl solution. Then 4.0 g of TMB and 0.046 g of NH4F were added followed by elevation of the reaction temperature to 40 °C. After that, 8.8 g of TEOS was added and the solution was further stirred at 40 °C for 20 h. The milky solution was transferred into an autoclave and aged in an oven at 70, 100, and 130 °C for 24 h, respectively, corresponding to MCF-1, MCF-2, and MCF-3. The white precipitates were filtered, dried at room temperature for 2 days, and then calcined at 550 °C for 5 h to obtain the MCF materials. The synthesis procedure of MCF-5 and MCF-6 was the same as that of MCF-2, except that the calcination temperature was increased to 700 and 900 °C, respectively. MCF-4 was prepared in the presence of KCl. The preparation process was as follows: A 4.0 g portion of P123 was dissolved in 150 mL of 1.6 M HCl solution. Then 8.0 g of KCl was added. After the solid was dissolved completely, 4.0 g of TMB and 0.046 g of NH4F were added, and the mixture was heated to 40 °C followed by addition of 8.8 g of TEOS. After being stirred at 40 °C for 20 h, the mixture was transferred into an autoclave and aged in an oven at 100 °C for 24 h. Then the white precipitate was collected by filtration, washed until no chloride ion was detected by 0.1 M AgNO3 solution, dried at room temperature for 2 days, and calcined at 550 °C for 5 h to remove the template. 2.3. Hydrothermal Treatments of MCF Materials in 100% Steam. To evaluate the hydrothermal stability of MCF materials, the samples were steamed in a tube furnace. In the steam treatment process, 100 mg of MCF material uniformly paved in a 55 × 15 × 8 mm ceramic boat was set in the center
J. Phys. Chem. C, Vol. 114, No. 11, 2010 5013 of a quartz tube of the furnace. The sample was heated to 600-800 °C with a ramp rate of 1.5 °C/min. Simultaneously, a 100% steam flow produced by heating a flask containing deionized water was passed through the tube furnace. The treatment lasted for 3-12 h under autogenous pressure. The obtained samples are denoted as “St-MCF-x-y-t”. x, y, and t represent the serial number, steaming temperature, and steaming time for the samples, respectively. For instance, St-MCF-2-800-3 indicates an MCF-2 sample treated in 100% steam at 800 °C for 3 h. 2.4. Characterization. Nitrogen sorption isotherms were measured with a Micromeritics Tristar 3000 analyzer at 77 K. Before measurement, the samples were degassed at 180 °C in a vacuum for more than 6 h. The specific surface area (SBET) was calculated using the Brumauer-Emmett-Teller (BET) method. The pore diameter (Dc) and window diameter (Dw) were determined from the adsorption and desorption branches, respectively, according to a modified Broekhoff de Boer method (BdB-FHH). The total pore volume (Vp) was estimated from the amount adsorbed at the relative pressure (P/P0) of 0.99. The micropore volume (Vm) was calculated by using the V-t plot method. The t values were calculated as a function of P/P0 using the de Boer equation t/Å ) [13.99/(log(P0/P) + 0.0340)]1/2. Vm was calculated using the equation Vm/cm3 ) 0.001547I, where I represents the y intercept in the V-t plots. SAXS measurements were carried out on a NanoSTAR small-angle X-ray scattering system (Bruker, Germany) using Cu KR radiation (40 kV, 35 mA). FT-IR spectra of the samples were measured on a Nicolet FT-IR360 spectrometer via the usual KBr pellet technique. These IR spectra were normalized by using the band at 1079 cm-1 as an internal reference band, which is usually assigned to the asymmetric vibration of the Si-O-Si bond. Transmission electron microscopy (TEM) images were recorded on a JEOL 2011 microscope (Japan) operated at 200 kV. Before TEM characterization, the samples were dispersed in ethanol. The suspensions of the samples were dropped on a holey carboncoated copper grid. The pore wall thickness was evaluated from TEM images on the basis of a statistical result of ∼50 pore walls. 29Si solid-state NMR experiments were performed on a Bruker DSX300 spectrometer with a frequency of 59.63 MHz, a recycling delay of 600 s, a radiation frequency intensity of 62.5 kHz, and a reference sample of Q8M8 ([(CH3)3SiO]8Si8O12). 3. Results and Discussion 3.1. Structural and Textural Properties of MCF Materials. The structural and textural parameters of MCFs can be controlled by tuning the aging and calcination temperatures as well as adding KCl in the preparation process (Table 1). The nitrogen sorption isotherms (Figure 1A) of all the MCF materials exhibit type IV curves with steep H1 hysteresis loops at high relative pressure (P/P0), indicating typical mesoporous materials with large pore sizes and narrow pore size distributions (PSDs), which is further confirmed by the corresponding well-resolved SAXS patterns (Figure 1B).13 Aging at 70-130 °C, the capillary condensation steps of the nitrogen sorption isotherms shift gradually to higher P/P0 values, indicating increases of pore sizes and window diameters with increasing of the aging temperature. However, the PSD becomes broader for MCF-3. This is because aging treatment at higher temperature partly destroys the structure of the microemulsion template, which is reflected by the relatively weaker SAXS peaks (Figure 1B). For samples MCF-2, MCF-5, and MCF-6, an increase of the calcination temperature leads to a decrease of the pore size, window size, pore volume, and specific surface area (Table 1). Especially the micropore volume decreases dramatically from
5014
J. Phys. Chem. C, Vol. 114, No. 11, 2010
Li et al.
TABLE 1: Pore Diameter (Dc), Window Diameter (Dw), BET Specific Surface Area (SBET), Total Pore Volume (Vp), Micropore Volume (Vm), and Reduced Surface Area of MCF-1, MCF-2, MCF-3, MCF-4, MCF-5, and MCF-6 before and after Being Treated in 100% Steam at 600 and 800 °C for 3, 6, and 12 h sample
Dc (nm)
Dw (nm)
SBET (m2/g)
Vp (cm3/g)
Vm (cm3/g)
reduced surface areaa (%)
MCF-1 MCF-2 MCF-3 MCF-4 MCF-5 MCF-6 St-MCF-1-600-3 St-MCF-1-600-6 St-MCF-1-600-12 St-MCF-2-600-3 St-MCF-2-600-6 St-MCF-2-600-12 St-MCF-3-600-3 St-MCF-3-600-6 St-MCF-3-600-12 St-MCF-4-600-3 St-MCF-4-600-6 St-MCF-4-600-12 St-MCF-6-800-3 St-MCF-6-800-6 St-MCF-6-800-12
22.6 34.4 43.6b 26.8 27.1 23.0 27.4c 30.2c 34.3c 34.6 36.1 36.5 39.3 40.1 40.1 41.1 d d 22.4 22.4 22.8
9.2 13.3 17.2 10.5 10.5 10.0 11.5 13.5c 13.3c 13.4 13.4 13.4 17.2 17.4 17.2 17.4 22.8 32.8 10.6 10.1 10.5
590 579 338 432 364 303 295 245 211 326 316 249 250 251 268 216 146 85 228 184 162
2.12 2.50 2.55 1.79 1.49 1.38 1.47 1.24 1.16 1.80 1.79 1.37 2.06 1.98 2.08 1.37 1.16 0.63 1.10 0.91 0.82
0.019 0.023 0.012 0.004 0.009 0.010 0.011 0.010 0.008 0.014 0.013 0.011 0.009 0.007 0.009 0.005 0.008 0.003 0.010 0.010 0.008
0 0 0 0 0 0 50.0 58.4 64.2 43.7 45.5 56.9 25.9 25.8 20.8 50.0 66.2 80.3 24.9 39.2 46.5
a The reduced surface area was calculated by the following formula: SBET(presteamed) - SBET(poststeamed)/SBET(presteamed). b The sample exhibits a broad pore size distribution. c More peaks of the pore size distribution of the samples appeared after the steam treatment. d The pore size distribution of the sample is not distinct after the steam treatment.
Figure 1. Nitrogen sorption isotherms (A) and SAXS patterns (B) of MCF-1, MCF-2, MCF-3, MCF-4, MCF-5, and MCF-6. For clarity, the isotherms and SAXS patterns of MCF-1 to MCF-6 are offset on the y axis.
0.023 cm3/g for MCF-2 to negligible values for MCF-5 and MCF-6. Compared to the counterparts calcined at 550 °C, MCF-4 has the lowest specific surface area and pore volume, indicating that the existence of inorganic salt KCl obviously affects the porosities of the MCF materials. TEM images (Figure 2) of the six MCF samples show 3-D interconnected mesostructures with uniform pore sizes. The pore sizes evaluated from the TEM images are consistent with those obtained from N2 sorption isotherms (Table 1). The thickness of the pore walls for all the samples is ∼5 nm regardless of the synthesis conditions. Nevertheless, it is obviously found that the edge of sample MCF-3 (Figure 2c) becomes very smooth under a high aging temperature. This phenomenon is observed in many edge domains of TEM images for sample MCF-3. It results from the recombination of the Si species in the framework at a high aging temperature. That is to say, at a high aging temperature fluoride ions can catalyze the dissolution of Si species in the thermodynamically unfavorable region and then
Figure 2. TEM images of MCF-1 (a), MCF-2 (b), MCF-3 (c), MCF-4 (d), MCF-5 (e), and MCF-6 (f).
the silicate species condense at the thermodynamically favorable region of the framework.13,27 3.2. Hydrothermal Stability of MCFs Calcined at 550 °C. Hydrothermal stability was first evaluated at 600 °C for the MCF samples calcined at 550 °C. After treatment with 100% water vapor for 3-12 h, the hysteresis loops of the nitrogen sorption isotherms for the MCF-1 materials significantly move to higher P/P0 values (Figure 3a), indicating an increase of the pore sizes and window sizes (Table 1). Meanwhile, the PSDs become broad (Figure S1, Supporting Information) and the specific surface area and pore volume decrease (Table 1). This suggests that MCF-1 aged at 70 °C has poor hydrothermal stability. In contrast, the mesostructure of sample MCF-2 aged at 100 °C is well retained after treatment with steam at 600 °C for different durations, which is confirmed by the nitrogen sorption isotherms (Figure 3b) and SAXS patterns (Figure 4A). Even after being
Hydrothermal Stability of Cellular Silica Foams
J. Phys. Chem. C, Vol. 114, No. 11, 2010 5015
Figure 3. Nitrogen sorption isotherms of samples before and after being treated with 100% steam at 600 °C for 3, 6, and 12 h: (a) MCF-1, (b) MCF-2 (c) MCF-3, and (d) MCF-4. For clarity, every isotherm is offset on the y axis.
Figure 4. SAXS patterns of the MCF materials treated in different conditions: (a) MCF-2, (b) St-MCF-2-600-3, (c) St-MCF-2-600-6, (d) StMCF-2-600-12, (e) MCF-5, (f) St-MCF-5-600-3, (g) St-MCF-5-700-3, (h) St-MCF-5-800-3, (i) MCF-6, (j) St-MCF-6-800-3, (k) St-MCF-6-800-6, (l) St-MCF-6-800-12.
treated with steam at 600 °C for 12 h, the MCF-2 materials still exhibit H1-type hysteresis loops in nitrogen sorption isotherms, which are located at the same relative pressure range as those of the untreated sample (Figure 3b). This suggests that the silica framework is stable in the treatment process and the pore sizes and window sizes remain the same (Figure S2, Supporting Information, and Table 1). However, after treatment at 600 °C for 3 h, the BET surface area of this sample is reduced by 43.7%. With increasing treatment time up to 6 and 12 h, further reduction of the BET surface areas of the samples is relatively smaller (Table 1). Furthermore, the well-resolved SAXS patterns of the MCF-2 samples treated by 100% steam also suggest that the mesostructures are well retained after the steam treatment at 600 °C. These results clearly indicate that MCF-2 calcined at 550 °C shows very good hydrothermal stability under pure steam at 600 °C. For the MCF-3 sample, which was treated with 100% steam at 600 °C for 3 h, the pore size distribution becomes narrow (Figure S3, Supporting Information) due to the reconstruction of the silica framework during the process, while the reduction of the micropore
volumes, total pore volumes, and specific surface areas is smaller than that of MCF-2 steamed at the same conditions. With elongation of the steaming duration, changes of the pore size, pore volume, and BET surface area of the samples are not very obvious. For instance, after steam treatment for 12 h, its BET specific surface area, total pore volume, and pore size are nearly the same as those of the St-MCF-3-600-3 sample (Table 1). This suggests that the MCF-3 sample has excellent hydrothermal stability. MCF-4 exhibits even poorer hydrothermal stability compared to MCF-1 (Figure 3d). After steam treatment at 600 °C for 3-12 h, its textural properties deteriorate dramatically. This can be mainly attributed to the effect of KCl on the structure of the microemulsion template. MCF-2 and MCF-3 have good hydrothermal stability at 600 °C among the samples obtained by calcination at 550 °C. However, the N2 sorption isotherms (Figure 5a) of MCF-2 samples treated with steam at 700-800 °C for 3 h show that the mesostructures suffer from severe destruction and collapse completely at 800 °C. After the steam treatment at 800 °C for 3 h, the specific surface area and pore volume decrease
5016
J. Phys. Chem. C, Vol. 114, No. 11, 2010
Li et al.
Figure 5. Nitrogen sorption isotherms of (a) MCF-2, (b) MCF-3, (c) MCF-5, and (d) MCF-6 before and after being treated with steam at 600-800 °C for 3 h. For clarity, every isotherm is offset on the y axis.
Figure 6. FT-IR spectra of (a) as-made and (b) calcined MCF-1, MCF-2, MCF-3, and MCF-4. As-made samples were designated as UN-MCF. (c) FT-IR spectra of MCF-2 samples after being treated with 100% steam at 600 °C for 3, 6, and 12 h.
dramatically from 579 to 81 m2/g and from 2.5 to 0.5 cm3/g, respectively. These results are consistent with the TEM measurements of sample St-MCF-2-800-3 (Figure 8d), indicating that its mesostructure collapses completely. The N2 sorption isotherms (Figure 5b) illustrate that MCF-3 has better hydrothermal stability than MCF-2 at 700-800 °C, but their textural properties still deteriorate a lot, suggesting an unsatisfactory hydrothermal stability. 3.3. Influential Factors of Hydrothermal Stability. The good hydrothermal stability of MCF-2 can be attributed to two aspects. One is micropores formed on the pore wall. Numerous tSisOH bonds are located around these micropores.12,28 The abundant micropores of MCF-2 can be confirmed by the corresponding high micropore volume (Table 1). During the steam treatment process, the condensation of tSisOH bonds takes place mainly around the micropores.12 As a result, many micropores collapse and the micropore volume decreases significantly (Table 1). Correspondingly, the BET surface area of the steamed MCF-2 samples decreases obviously. Another
reason is that the higher aging temperature results in a higher polymerization degree of the silica framework. After steam treatment for a certain duration, the polymerization degree of the amorphous framework is further enhanced and the silica framework becomes more robust. When the steam treatment is prolonged, the structural recombination process almost stops and the change of the structural parameter is very small. The FT-IR spectra (Figure 6) show that the tSisOH band29,30 of MCF-2 at around 970 cm-1 becomes weaker after steam treatment for 3 h and the intensity of the peak is nearly unchanged with the elongation of the steaming time. Furthermore, this can also be confirmed by the 29Si solid-state NMR spectra of MCF-2, St-MCF-2-600-3, and St-MCF-2-600-12 (Figure 7). With an increase of the steaming time, the Q3 signal for Si species in sample MCF-2 becomes very small. By comparing the TEM images of MCF-2 steamed at 600 °C for different times (Figure 8), it is found that, at the edge of sample St-MCF-2-600-12, the TEM image (Figure 8b) is more “smooth” than that of MCF-2 without being steamed (Figure 2b, marked
Hydrothermal Stability of Cellular Silica Foams
Figure 7. 29Si solid-state NMR spectra of samples MCF-2, MCF-6, St-MCF-2-600-3, and St-MCF-2-600-12.
Figure 8. TEM images of (a) St-MCF-2-600-3, (b) St-MCF-2-60012, (c) St-MCF-5-700-3, (d) St-MCF-2-800-3, (e) St-MCF-6-800-3, and (f) St-MCF-6-800-12.
with rectangles in Figures 2b and 8b). This contrast between MCF-2 and St-MCF-2-600-12 can be observed on the most edge domains of their TEM images. This variation characteristic has also been observed previously between the unsteamed and steamed SBA-15.12 This phenomenon is derived from the recombination of tSisOsSit bonds via the simultaneous hydrolysis and condensation processes under high-temperature steam. Similarly, the good hydrothermal stability of the MCF-3 sample mainly derives from a higher degree of tSisOH condensation and a denser silica framework compared to those of MCF-1 and MCF-2, due to the higher aging temperature. This can be confirmed by the low intensity of the tSisOH band in the FT-IR spectrum of the as-made MCF-3 sample (Figure 6a). In contrast with MCF-2 and MCF-3, the poorer hydrothermal stability of MCF-1 may result from its lower polymerization degree of silica frameworks due to the lower aging temperature. This hypothesis is supported by the FT-IR spectra of the asmade MCF materials prepared under different aging temperatures (Figure 6). It reveals that the as-made MCF-1 has more tSisOH groups in the silica framework indicated by the relatively stronger intensity of the IR peak at about 970 cm-1
J. Phys. Chem. C, Vol. 114, No. 11, 2010 5017 (Figure 6a). After calcination at 550 °C, the intensity of the tSisOH band of MCF-1 is still higher than that of the MCF-2 to MCF-4 samples (Figure 6b), which also indicates that MCF-1 has a lower polymerization degree of the silica framework. Another factor related to the poorer hydrothermal stability of MCF-1 is the fewer micropores confirmed by its relatively small pore volume. In this case, the recombination of the tSisOsSit linkages occurs directly in the framework and may result in the severe damage of the mesostructure of MCF-1. Generally, it is believed that the salt effect can improve the hydrothermal stability of mesoporous silicas by increasing the polymerization degree of the silica framework or pore-wall restructuring.6,31 However, in our experiments, after the steam treatment at 600 °C, the mesostructure of the MCF-4 sample synthesized in the presence of KCl is severely destructed. This indicates that the addition of KCl does not obviously increase the polymerization degree of its framework (Figure 6a). Most importantly, MCF-4 has the fewest micropores due to the addition of KCl compared to MCF-1 with MCF-3. According to the observations and analyses mentioned above, it can be concluded that the higher aging temperature and more micropores can result in a higher hydrothermal stability of MCF materials. 3.4. Improving the Hydrothermal Stability of MCFs in 100% Steam at 800 °C. As mentioned above, the mesostructure of MCF materials calcined at 550 °C collapsed in 100% steam at 800 °C. On the basis of the analyses in section 3.3, the hydrothermal stability of MCFs can be improved by enhancing the polymerization degree of the framework. However, the polymerization degree increases with the increase of the calcination temperature. To improve the hydrothermal stability of MCFs in 100% steam at 800 °C, the calcination temperature was increased to 700 °C (for example, sample MCF-5). The N2 sorption isotherms of the MCF-5 sample (Figure 5c) show an H1-type hysteresis loop with a steep capillary condensation step, and the pore size is unaltered after the steam treatment at 700 °C. Furthermore, the TEM images (Figures 2e and 8c) show that the mesopores of the St-MCF-5-700-3 sample become smoother than those of the unsteamed counterpart MCF-5, but the pore wall thickness is approximately unchanged. This clearly indicates that the mesostructure of sample MCF-5 is well preserved after steam treatment at 700 °C for 3 h. However, the mesostructure collapses severely and the specific surface area decreases from 364 to 78 m2/g when the steaming temperature increases to 800 °C (Figure 5c). SAXS patterns (Figure 4B) also confirm that the mesostructure of the MCF materials deteriorates significantly after the steam treatments at 800 °C. In addition, the hydrothermal stability of MCFs was further investigated by increasing the calcination temperature of the samples to 900 °C (sample MCF-6, Figure 5d). After steam treatment at 800 °C for 3 h, the N2 sorption isotherms of StMCF-6-800-3 still show the typical H1 hysteresis loop and the specific surface area decreases only about 24.9% from 303 to 228 m2/g (Table 1 and Figure 10). Its well-resolved mesostructures are also confirmed by the corresponding TEM results (Figure 8e). When the steam treatment time is increased to 12 h, the SAXS peaks of St-MCF-6-800-12 are still well-resolved (Figure 4C). The specific surface area of MCF-6 decreases gradually from 303 to 162 m2/g, and the total pore volume decreases from 1.38 to 0.82 m3/g, while the pore and window sizes are nearly unchanged (Figure S4, Supporting Information, and Table 1). The TEM image of St-MCF-6-800-12 further confirms that it has well-resolved mesostructures (Figure 8f).
5018
J. Phys. Chem. C, Vol. 114, No. 11, 2010
Figure 9. FT-IR spectra of (a) MCF-2 steamed at 600-800 °C and (b) MCF-2, MCF-5, and MCF-6 calcined at different temperatures.
These results clearly indicate that the mesostructure of MCF-6 can be retained with a slight collapse. Therefore, increasing the calcination temperature is an effective way to improve the hydrothermal stability of MCFs in 100% steam at 800 °C. This derives from the enhancement of the polymerization degree of the silica framework. The framework of MCFs becomes more robust by increasing the calcination temperature, and it can withstand the recombination process of tSisOsSit linkages under more critical and severe conditions. 3.5. Variations of the MCF Framework during Steam Treatments. To evaluate the hydrothermal stability, the mesoporous silica materials were usually treated with steam or in boiling water. In boiling water, the destruction of silica materials mainly derives from the continuous hydrolysis of tSisOsSit linkages on the surface of the pore walls. Therefore, more tSisOsSit bonds changed into tSisOH bonds, and many Si species were dissolved into the boiling water. An increase of the Q3 peak intensity for the Si species can be found in the 29 Si solid-state NMR spectra after the treatment with boiling water.12,27,32 In contrast to treatment with boiling water, the treatment in 100% steam at 600-800 °C is a more severe and critical condition. However, the steam in this process cannot take away the silicate species from the materials. The tSisOsSit bonds were hydrolyzed on the surface of the silica wall into tSisOH bonds driven by steam, and then the tSisOH bonds condensed again into tSisOsSit bonds driven by the thermal treatment. The latter variation is dominant for the high-temperature steam treatment. In this way, the polymerization degree of the silica frameworks is largely increased. This can be confirmed by the 29Si solid-state NMR spectra (Figure 7) of MCF-2, St-MCF-2-600-3, and St-MCF2-600-12, which indicate that the Q3 Si species decrease after the steam treatment. During the steam treatment process of MCF-2 at 600-800 °C, the IR spectra of these samples show that the bands at 970 cm-1 related to the tSisOH bond gradually decreased with an increase of the steaming temperature, indicating an increase of the polymerization degree of the framework (Figure 9a). However, under such a critical condition, the recombination of tSisOsSit linkages becomes more rapid and severe, and the silica framework of MCFs calcined at 550 °C cannot withstand the reorganization of the tSisOsSit bonds. As a result, the mesostructure of MCF-2 collapses under the steam at 800 °C.
Li et al.
Figure 10. N2 sorption isotherms of MCF-6 before and after being treated with 100% steam at 800 °C for 3, 6, and 12 h. For clarity, the isotherms of MCF-6 to St-MCF-6-800-12 are offset on the y axis.
It seems that elevating the calcination temperature is an effective method to improve the hydrothermal stability of MCFs although part of the surface area is lost. The decrease of the peak intensity of the tSisOH bond in the FT-IR spectra (Figure 9b) suggests that the polymerization degree of the silica framework is obviously enhanced. This hypothesis can be further proved by the 29Si solid-state NMR of MCF-6 (Figure 7), in which the Q3 signal for the Si species is smaller than that of MCF-2. The framework of MCF-6 can withstand the more severe and critical condition of 100% steam at 800 °C. Interestingly, it is found that the structural variations during the high-temperature steam treatments for MCFs and SBA-15 materials are distinctly different. The pore wall thickness of MCFs remains the same (∼5 nm) during the whole steam treatment process, and the contraction of the MCF framework is not obviously apparent. Furthermore, the pore and window sizes are unaltered. These results are confirmed by the TEM images (Figure 8) and the structural parameters (Table 1). However, for ordered mesoporous silica SBA-15, after being treated with 100% steam, the pore size became smaller, the wall thickness increased, and contraction of the framework was distinctly observed.12 From the viewpoint of thermodynamics, it could be explained that, for reducing the thermodynamic energy and preserving the stability of MCFs, the Si species at a thermodynamically unfavorable region recombine to form a thermodynamically favorable region via simultaneous hydrolysis and condensation processes. This viewpoint can be supported by the change of the edge of the MCF samples from “coarse” to “smooth” (mentioned in section 3.3), clearly indicating that the recombination process in the silica framework occurs. However, for SBA-15, the surface energy is depressed and the mesostructure stabilized through the contraction of the silica framework. It seems that there are two different ways to stabilize the silica frameworks in MCFs and SBA-15. 4. Conclusions The mesostructure of MCF materials with aging temperature higher than 100 °C and calcined at 550 °C displays high hydrothermal stability at 600 °C under 100% steam. In the first 3 h of the pure steam treatment, the BET surface area and pore volume of the MCFs decreased obviously and the polymerization degree of the silica frameworks is enhanced distinctly. With the elongation of the steam treatment, the mesostructure becomes
Hydrothermal Stability of Cellular Silica Foams more robust while the variation of the physicochemical data of the steamed MCFs is slight. However, the mesostructure of the MCFs calcined at 550 °C collapses completely under the steam at 800 °C. By increasing the calcination temperature of the MCFs to 900 °C, the mesostructure of the MCFs is stable under the pure steam at 800 °C. The high aging temperature, more micropores, and high calcination temperature are in favor of the hydrothermal stability of MCFs. However, the salt effect is disadvantageous for the hydrothermal stability. Surprisingly, MCFs and SBA-15 show distinct structural variation characteristics. The pore size, window size, and pore wall thickness are unaltered for steam-treated MCFs, while the pore size decreases and the pore wall becomes thicker for SBA-15 with the same steam treatment. It seems that there are two different ways to depress the thermodynamic energies of the systems and stabilize the mesostructures for MCF and SBA-15 materials. Acknowledgment. This work was supported by the NSF of China (Grants 20721063, 20821140537, and 20890123), the State Key Basic Research Program of the PRC (Grants 2006CB202502 and 2009CB930400), the Science & Technology Commission of Shanghai Municipality (Grant 08DZ2270500), and the Shanghai Leading Academic Discipline Project (Grant B108). Supporting Information Available: Figures showing the pore size distributions of MCF-1, MCF-2, MCF-3, and MCF6. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Beck, J. S.; VartUli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T.-W.; Olson, D. H.; Sheppard, E. W.; McCullen, S. B.; Higgins, J. B.; Schlenker, J. L. J. Am. Chem. Soc. 1992, 114, 10834. (2) Juan, J. C.; Zhang, J. M.; Yarmo, A. J. Mol. Catal. A 2007, 267, 265. (3) Cai, M.; Sha, J.; Xu, Q. Tetrahedron 2007, 63, 4642. (4) Indra, A.; Basu, S.; Kulkarni, D. G.; Gopinath, C. S.; Bhaduri, S.; Lahiri, G. K. Appl. Catal., A 2008, 344, 124. (5) Rhee, C. H.; Lee, J. S. Catal. Lett. 1996, 40, 261.
J. Phys. Chem. C, Vol. 114, No. 11, 2010 5019 (6) Ryoo, R.; Jun, S. J. Phys. Chem. B 1997, 101, 317. (7) Zhao, D.; Feng, J.; Huo, Q.; Melosh, N.; Fredickson, G. H.; Chmelka, B. F.; Stucky, G. D. Science 1998, 279, 548. (8) Xiong, H.; Zhang, Y.; Wang, S.; Liew, K.; Li, J. J. Phys. Chem. C 2008, 112, 9706. (9) Rioux, R. M.; Song, H.; Hoefelmeyer, J. D.; Yang, P.; Somorjai, G. A. J. Phys. Chem. B 2005, 109, 2192. (10) Hicks, J. C.; Mullis, B. A.; Jones, C. W. J. Am. Chem. Soc. 2007, 129, 8426. (11) Li, C.; Wang, Y.; Guo, Y.; Liu, X.; Guo, Y.; Zhang, Z.; Wang, Y.; Lu, G. Chem. Mater. 2007, 19, 173. (12) Zhang, F.; Yan, Y.; Yang, H.; Meng, Y.; Yu, C.; Tu, B.; Zhao, D. J. Phys. Chem. B 2005, 109, 8723. (13) Schmidt-Winkel, P.; Lukens, W. W., Jr.; Yang, P.; Margolese, D. I.; Lettow, J. S.; Ying, J. Y.; Stucky, G. D. Chem. Mater. 2000, 12, 686. (14) Schmidt-Winkel, P.; Lukens, W. W., Jr.; Zhao, D.; Yang, P.; Chmelka, B. F.; Stucky, G. D. J. Am. Chem. Soc. 1999, 121, 254. (15) Liu, Y.-M.; Feng, W.-L.; Li, T.-C.; He, H.-Y.; Dai, W.-L.; Huang, W.; Cao, Y.; Fan, K.-N. J. Catal. 2006, 239, 125. (16) Ungureanu, A.; On, D. T.; Dumitriu, E.; Kaliaguine, S. Appl. Catal., A 2003, 254, 203. (17) Su, Y.; Liu, Y.-M.; Wang, L.-C; Chen, M.; Cao, Y.; Dai, W.-L.; He, H.-Y.; Fan, K.-N. Appl. Catal., A 2006, 315, 91. (18) Eswaramoorthi, I.; Dalai, A. K. Int. J. Hydrogen Energy 2009, 34, 2580. (19) Cavani, F.; Trifiro, F. Appl. Catal., A 1995, 133, 219. (20) Corma, A. Chem. ReV. 1997, 97, 2373. (21) Kim, J. M.; Jun, S.; Ryoo, R. J. Phys. Chem. B 1999, 103, 6200. (22) Ryoo, R.; Kim, J. M.; Ko, C. H.; Shin, C. H. J. Phys. Chem. 1996, 100, 17718. (23) On, D. T.; Kaliaguine, S. J. Am. Chem. Soc. 2003, 125, 618. (24) On, D. T.; Kaliaguine, S. Angew. Chem., Int. Ed. 2002, 41, 1036. (25) Zhang, Z.; Han, Y.; Xiao, F.-S.; Qiu, S.; Zhu, L.; Wang, R.; Yu, Y.; Zhang, Z.; Zou, B.; Wang, Y.; Sun, H.; Zhao, D.; Wei, Y. J. Am. Chem. Soc. 2001, 123, 5014. (26) Shen, S. C.; Kawi, S. J. Phys. Chem. B 1999, 103, 8870. (27) Li, Y.; Zhang, W.; Zhang, L.; Yang, Q.; Wei, Z.; Feng, Z.; Li, C. J. Phys. Chem. B 2004, 108, 9739. (28) Palkovits, R.; Yang, C.-M.; Olejnik, S.; Schu¨th, F. J. Catal. 2006, 243, 93. (29) Tian, B.; Liu, X.; Yu, C.; Gao, F.; Luo, Q.; Xie, S.; Tu, B.; Zhao, D. Chem. Commun. 2002, 11, 1186. (30) Vinu, A.; Satish Kumar, G.; Ariga, K.; Murugesan, V. J. Mol. Catal. A 2005, 235, 57. (31) Jun, S.; Kim, J. M.; Ryoo, R.; Ahn, Y.-S.; Han, M.-H. Microporous Mesoporous Mater. 2000, 41, 119. (32) Kim, J. M.; Ryoo, R. Bull. Korean Chem. Soc. 1996, 17, 66.
JP9100784