SBA2 P63mmc Fm3m

SBA-2 was first synthesized by Stucky's group in 1995 by using Gemini surfactant C„H2„+1N+(CH3)2(CH2)sN+ (CH3)3 (Cn_s_i) [10]. Based on powder XRD results, the material was believed to consist of discrete supercages in a hexagonal close-packed (hcp) arrangement and the space group was determined to be P63/mmc. The unit cell dimensions are a ~ 6 nm and c ~ 10 nm which can be tuned by changing

Figure 5. TEM image of 3D porous single crystal of Cr2O3 synthesized using SBA-15 as a template which has been completely removed. It is shown that all the nanorods of oxide are linked by smaller bridges. The structure is a negative replication of SBA-15. The inset is the corresponding SAED pattern from the whole particle.

Figure 5. TEM image of 3D porous single crystal of Cr2O3 synthesized using SBA-15 as a template which has been completely removed. It is shown that all the nanorods of oxide are linked by smaller bridges. The structure is a negative replication of SBA-15. The inset is the corresponding SAED pattern from the whole particle.

the chain length of the surfactants. It was also mentioned that the shape of the mesopores in SBA-2 is bottlelike. The pore diameter was about 3.5 nm according to the BJH method [25] and no micropore was detectable. Such pore size measurement gave rise to a hypothesis that there must be some mesopores connecting the spherical supercages, which are separated by the silica wall with a minimal thickness of 2.5 nm. Based on TEM observation of some fringes, Zhou et al. proposed a 2D network of mesopores in SBA-2 [88]. This model may not be correct according to the recent adsorption-desorption studies.

Ravikovitch believes that the conventional BJH method for adsorption-desorption measurement becomes extremely inaccurate for spherical cavities in mesoporous materials such as SBA-2. They developed a NLDFT and remeasured the diameter of the cages in SBA-2 with the unit cell dimension of 4.9 nm. The diameter became 4.4 nm, corresponding to a minimal wall thickness of 0.5 nm [26]. In a more recent work, Garcia-Bennett et al. performed adsorption and desorption experiments on several SBA-2 specimens using three different gases, nitrogen, cyclopentane, and mesitylene. It was found that the windows of the supercages in SBA-2 were actually micropores which could be as small as below 0.4 nm so that only small molecules can be adsorbed [89]. The above two experiments indicated that the touch points in between two spherical cages have a very thin wall and some micropores may form without templating during the hydrothermal treatment and/or calcinations. The volume of these micropores is too small to be detected by adsorption-desorption experiments.

The C„_ surfactant molecules have a high charge density and large head groups. Globular micelles instead of cylindrical rods form in the solution. Formation of the SBA-2 structure mainly depends on the close packing of the globular surfactant-silicate arrays. Therefore, both hcp (Fig. 6a) and cubic close-packed (ccp) (Fig. 6b) structures are possible and indeed have often been observed in the SBA-2 specimens. Another possible packing structure is body-centered cubic as shown in Figure 6c, which, however, has not been observed in this material.

From analysis of the TEM images of SBA-2, it was found that intergrowth of the hcp and ccp phases is very common. The latter was designated STAC-1 and the highest symmetry would be Fm3m [88] (Fig. 7). Such intergrowth can be layered (one-dimensional) along the [001] axis of the hcp phase or the [111] direction of the ccp phase. The intergrowth can also be three-dimensional, forming a domain structure in the particles. Such a domain structure was found in SBA-2 with

Figure 6. Graphic representations of possible packed structures, (a) hcp, (b) ccp, and (c) body-centered cubic packing.

Figure 6. Graphic representations of possible packed structures, (a) hcp, (b) ccp, and (c) body-centered cubic packing.

Figure 7. TEM image from SBA-2 showing layered intergrowth of the hexagonal closed packed (ABA) and cubic closed packed (ABCA). The ccp structure dominates the particle.

three different morphologies. One morphology is a large hollow sphere of about 50 to 150 ¡m in diameter and the thickness of shells is only about 1 to 2 ¡m. The second is small solid sphere with the diameter in the range of 2 to 3 ¡m. The third morphology is a sheetlike plate [90, 91].

To date, a pure phase of either hcp or ccp phase SBA-2 has not been synthesized. In the recent report by Garcia-Bennett et al. [89], it was presented that the morphology of the SBA-2 particles could be controlled by changing the acidic/alkali conditions and the concentration of the surfactant. When the surfactant concentration was 40%, large hexagonal prisms were observed. The XRD pattern of this sample can be indexed onto a hexagonal unit cell with a = 4.3 and c = 6.8 nm. However, the monophasic nature of this so-called hcp SBA-2 has not been confirmed by other techniques.

SBA-12 was synthesized using the oligomeric alkyethylene oxide surfactant Brij 76 (C18EO10) [65]. The space group was originally determined to be P63/mmc, similar to SBA-2. The difference between these two phases is that different surfactants are used, cationic Gemini surfactants for SBA-2 and nonionic oligomeric surfactants for SBA-12. Nevertheless, no significant structural difference was found in between these two phases. For example, a typical calcined SBA-12 has a 3D hexagonal unit cell with a = 6.34 and c = 10.2 nm. The BET surface area is 1150 m2/g and the pore volume is 0.83 cm3/g [66].

It was soon revealed by using TEM that the SBA-12 specimens usually contain mixed hcp and ccp phases, exactly same as we have seen from SBA-2. On the other hand, a functionalized SBA-12, SH-SBA-12, synthesized by Spanish scientists, showed large domains of the ccp phase. Based on good TEM images of single domains, electron crystallo-graphic studies gave a face centered cubic structural model with cell dimension a = 8.2 nm [21]. The structure forms with each cavity connected to its 12 nearest-neighbor cavities through openings parallel to the (110) directions. The ideal space group is therefore Fm3m (Fig. 6b). To date, syntheses of the pure hcp and ccp SBA-12 have not been achieved.

It is noted that the diameter of the pores in SBA-12 was measured to be 6.3 nm when the pore center to pore center distance was about 7.4 nm according to Ravikovitch's nonlocal density functional theory [26]. The corresponding minimum wall thickness is over 1 nm which is larger than that in SBA-2.

Another mesoporous phase with the symmetry of P63/mmc is named HMM-2, which consists of an organic/ inorganic hybrid framework, the same as the composition of HMM-1 [55, 56].

MCM-48 was first reported as one of the members in the M41S series using the same surfactant as that for MCM-41 with a different ratio of surfactant to silicate source [3]. Although MCM-48 can be also obtained by phase transformation from MCM-41 [92], the pore system in the former is significantly different from the latter. Its structure can be described using a model of minimal surface of gyroid (Fig. 8), first suggested in 1993 [8] and confirmed a few years later using HRTEM [93, 94] and a newly developed method of electron crystallography [19]. This minimal surface divides space into two identical but separate compartments, forming a so-called bicontinuous cubic with space group Ia3d. The unit cell parameter is about 8.5 nm. The mean pore size of MCM-48 has been determined according to the nonlocal density functional theory to be 3.1 to 3.3 nm with the wall thickness of 0.8 to 1 nm [95]. The minimal surface can be approximated by the simple analytical function cos x sin y + cos y sin 2 + sin x cos 2 = 0

MCM-48 is one of few mesoporous silicas which can be produced as relatively large single crystals. For example, MCM-48 particles of a few micrometers in diameter with a unique crystalline morphology of a cube truncated by rhomb dodecahedron was synthesized by Kim et al. [96].

Incorporation of metal cations in MCM-48 can be achieved as successfully as that in MCM-41. The guest cations include Ti, Cr, V [97], Cr [98], Fe [99], Mn, Al [21, 100], Zr [101], B [36], Cu, Zn [102], etc. The degree of

Figure 8. Schematic drawing of the mesopores in MCM-48.

order of the mesopores in MCM-48 usually decreases when the doping level increases.

Encapsulation of metals such as Pd inside the MCM-48 mesoporous system resulted in some nanoball domains [80]. In this case, the metal did not fill completely the pores. Nevertheless, after removing the silicate matrix by a dilute HF solution, pure porous Pd nanoballs were recovered which actually are negative replicas of the MCM-48 structure. If the mesopores of MCM-48 are fully filled by carbon and the silica framework is removed by HF, a 3D carbon negative replica of MCM-48 can be fabricated [103]. The final product is also mesoporous with a quite narrow pore size distribution centered at 2.4 nm and specific surface area of 1200 m2/g. These carbon replicas of MCM-48 can retain the parent symmetry of Ia3d, designated CMK-1, or changed to 14j/a, designated CMK-4, [24] depending on the silicate source used. The wall structure of the mesoporous carbon can contain graphite domains instead of being completely amorphous. It is interesting to see that micropores of 0.5 to 0.8 nm in diameter exist in the silica wall of MCM-48 so that a 3D framework of carbon becomes possible after removal of the silica framework [104]. This structural feature is similar to that of SBA-15. Three-dimensional porous negative replicas of MCM-48 with other materials, such as crystalline Pt, have also been investigated [105].

SBA-16 is also a cubic phase with a symmetry of Im3m and the cell parameter a = 16.6 nm when triblock copoly-mer with large PEO segments, EO106PO70EO106, was used [66]. The structure of this phase can be regarded as body-centered packing of globular cages (Fig. 6c) as proved by HRTEM [20]. The calcined specimen has a pore size of 5.4 nm, a pore volume of 0.45 cm3/g, and a BET surface area of 740 m2/g. In this case, the corresponding minimal wall thickness is about 9 nm. According to the NLDFT method, on the other hand, the pore size of SBA-16 with the same unit cell dimension is about 8.5 nm and the minimal wall thickness is about 5.9 nm [26]. Both values of the minimal wall thickness are much larger than those in SBA-2 and SBA-12, indicating a weaker interplay between the globular micelles during the assembly of organic and inorganic species. Consequently, the formation of close packed structures does not take place in this synthetic system.

On the other hand, a relatively thicker wall must contain some small channels connecting the spherical cages. Otherwise, gas adsorption on the inner surface would be impossible. As determined by Voort et al., a large amount of micropores exist in SBA-16. The porosity is greatly dependent on the synthetic conditions. When a high ratio of silicate source to surfactant is applied, many microporous nanocapsules will be created [106]. The details of these channel bridges have not been revealed by HRTEM or by other techniques.

In Zhao's first report about SBA-16, they also detected two other cubic phases [66]. One was SBA-11 with symmetry of Pm3m and the unit cell dimension of 10.6 nm, prepared using nonionic alkyl PEO oligomeric surfactants,

C16H33(OCH2CH2)10OH (C16EO10). A typical calcined

SBA-11 specimen contains uniform mesopores of 2.5 nm in diameter and has a BET surface area of 1070 m2/g and a pore volume of 0.68 cm3/g. The second phase, designated SBA-14, was obtained when the surfactants with shorter chains and smaller EO segments were used, such as C12EO4. The thermally stable SBA-14 contains mesopores with pore diameter of 2.2 nm and a BET surface area of 670 m2/g. However, the space group of this phase was not determined. Both SBA-11 and SBA-14 have not been extensively studied since Zhao's report in 1998.

SBA-1 has a complicated packing structure of two types of globular cages and was first synthesized by Huo et al. in 1994 [9, 107] using tetraethyl orthosilicate (TEOS) as a silica source and hexadecyltriethylammonium bromide (HTEABr) as the surfactant under strongly acidic conditions. The synthetic conditions were refined by Kim and Ryoo in 1999 to produce an excellent specimen by using a reaction mixture of 1 HTESBr : 5 TEOS : xHCl : 12.5xH2O, where x can be varied from 240 to 400 and applied stirring of 4 hours at 273 K [108]. The pore size together with the unit cell dimensions can be varied by choosing different chain lengths in the surfactant. The product using C16H33N(C2H5)3Br as a template has a unit cell parameter a = 7.6 nm, pore size of 2.1 nm in diameter by the BJH method, and surface area of 1355 m2/g. However, according to the NLDFT method, the pore size of the similar specimen is about 4.1 nm and the minimal wall thickness is about 0.3 nm [26].

The structure of SBA-1 contains large cages (4 nm in diameter) at positions of (1/2, 0, 1/4), (1/2, 0, 3/4), (0, 1/4, 1/2), (0, 3/4, 1/2), (1/4, 1/2, 0), and (3/4, 1/2, 0), and small cages (3.3 nm in diameter) at the (0, 0, 0) and (1/2, 1/2, 1/2) positions (Fig. 9) [20]. The windows of the cages are also two different sizes. If the synthetic conditions are well controlled, large particles of SBA-1 with decaoctahedral crystallike shape can be fabricated as demonstrated by Che et al. (Fig. 10) [109].

In the same article, Sakamoto et al. also described for the first time another cubic phase with the same symmetry Pm3n but prepared under basic conditions, designated SBA-6 [20]. The unit cell dimension of SBA-6 (14.6 nm) is about double

Pm3n Cubic Structure Sba

Figure 9. Schematic drawing of the structures of SBA-1 and SBA-6. The large cages and small cages are indicated by A and B respectively. Reprinted with permission from [20], Y. Sakamoto et al., Nature 408, 449 (2000). © 2000, Macmillan Magazines Ltd.

Figure 10. SEM image showing decaoctahedral crystal-like morphology of SBA-1. Reprinted with permission from [109], S. Che et al., Chem. Mater. 13, 2237 (2001). © 2001, American Chemical Society.

Figure 9. Schematic drawing of the structures of SBA-1 and SBA-6. The large cages and small cages are indicated by A and B respectively. Reprinted with permission from [20], Y. Sakamoto et al., Nature 408, 449 (2000). © 2000, Macmillan Magazines Ltd.

Figure 10. SEM image showing decaoctahedral crystal-like morphology of SBA-1. Reprinted with permission from [109], S. Che et al., Chem. Mater. 13, 2237 (2001). © 2001, American Chemical Society.

that of SBA-1, while the wall thickness of the former is much thinner. The diameter of the large cages is 8.5 nm and that of the small cages is 7.3 nm.

Chemical doping in SBA-1 has not been investigated as extensively as in MCM-41. But its 3D porous network is very attractive for potential application in catalysis and it is indeed possible to introduce various metal cations in order to modify the physicochemical properties. Up to date, V, Ti, Mo, and Co have been incorporated into the SBA-1 framework and the original structure with space group of Pm3n is maintained very well [110-114]. Among them, the incorporation of vanadium is particularly interesting, not only because the doping level could be quite high (e.g. Si/V < 20) but also because the incorporation of V leads to the formation of secondary mesopores [111].

Synthesis of the Pm3n phase with an ethane-silica hybrid framework proved to be successful [115]. The structure is similar to SBA-1 with the unit cell parameter a = 11.1 nm, the pore size about 2.9 nm by the BJH calculation, and the BET surface area about 770 m2/g. On the other hand, the pore size was determined to be 5.1 nm by the NLDFT method with a minimum wall thickness of 1.1 nm [26]. The latter method is believed to be more accurate for the packing structures of globular cages.

We discussed the most popular ordered mesoporous silicas in the above sections. There are indeed many other phases which are either difficult to synthesize or have been discovered recently. Two examples are FDU-1 and FDU-2.

FDU-1 has a cubic unit cell with the cell parameter a = 22 nm, space group /m3m, synthesized by Zhao's group [116] using triblock poly(ethylene oxide)-poly(butylenes oxide)-poly(ethylene oxide) copolymer, such as EO39BO47-EO39. The pore size is about 12 nm, pore volume about 0.77 cm3/g, and BET surface area about 740 m2/g. The structure is similar to SBA-16 (Fig. 6c) but has a significantly larger pore size and unit cell dimensions, which are the largest among all the known cubic mesoporous silicas.

FDU-2, also reported by Zhao's group recently, has sym metry of Fd3m. The material was prepared by using tri head group quaternary ammonium surfactants, [CmH2m+j N+(CH3)2CH2CH2N+(CH3)2CH2CH2CH2N+(CH3)3-3Br] (m = 14, 16, 18), as the structure-directing agents under basic conditions at low temperature [117]. It was assumed that FDU-2, similar to FDU-1, is also a packed structure of globular cages, but the arrangement of the cages is a novel diamond type. Although the proposed structure was supported by XRD and TEM images, the details of the structure have yet to be determined.

0 0

Post a comment